<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xml:base="https://www.dbdump.org"  xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
 <title>dbdump.org</title>
 <link>https://www.dbdump.org</link>
 <description></description>
 <language>en</language>
<item>
 <title>Baking-in bibliographic references directly into RDF / OWL ontologies</title>
 <link>https://www.dbdump.org/node/56</link>
 <description>&lt;div class=&quot;field field-name-field-image field-type-image field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;og:image rdfs:seeAlso&quot; resource=&quot;https://www.dbdump.org/sites/www.dbdump.org/files/styles/large/public/field/image/IMG_20170914_160235.jpg?itok=stK0TAi-&quot;&gt;&lt;img typeof=&quot;foaf:Image&quot; src=&quot;https://www.dbdump.org/sites/www.dbdump.org/files/styles/large/public/field/image/IMG_20170914_160235.jpg?itok=stK0TAi-&quot; width=&quot;480&quot; height=&quot;270&quot; alt=&quot;&quot; /&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;Not every part of an ontology is novel. Often enough we reuse existing works and standards to populate them. At other times the ontology implements policies and processes that are outlined in a document narrative but never before formalised into a data structure. To give credit where credit is due, and to ensure policy compliance, it can be worthwhile to integrate bibliographic citations and references directly within the ontology. Ontologies already exist to record bibliographic data. &lt;a href=&quot;https://github.com/structureddynamics/Bibliographic-Ontology-BIBO&quot;&gt;BIBO&lt;/a&gt; is one which has the added benefit that it provides a direct mapping for bibtex-style bibliography managers. &lt;a href=&quot;https://www.loc.gov/bibframe/&quot;&gt;Bibframe&lt;/a&gt; provides a comprehensive solution if you have a corporate document repository, but avoid using Dublin Core’s bibliographicCitation term as it was never standardised.&lt;/p&gt;
&lt;h2&gt;
	Bibliography as ontological objects, but why?&lt;/h2&gt;
&lt;p&gt;When the bibliographic reference for the documents used to create the ontology are themselves available as ontological terms, they can be directly referenced by other terms or queried directly. This isn’t an exercise in minutiae; it is a change management tool and an additional asset for application developers.&lt;/p&gt;
&lt;p&gt;When a term represents a concept defined by a standard, you can directly link it to the specific edition of the document that defines it authoritatively. Besides avoiding the need for those pesky double quotes in the definition property, this also tracks which edition of the standard this term represents. This can easily be done with various levels of verbosity (hadPrimarySource) using the &lt;a href=&quot;https://www.w3.org/TR/prov-o/&quot;&gt;PROV&lt;/a&gt; ontology or by using a simple Dublin Core source annotation.&lt;/p&gt;
&lt;p&gt;This is important not only for change management: it helps to differentiate terms that have the same labels and similar descriptions but whose context differs from one edition to another. Thus, datasets that record events using older specifications retain their integrity and newly generated data is kept up to date with current standards.&lt;/p&gt;
&lt;h2&gt;
	Keeping up with the Joneses&lt;/h2&gt;
&lt;p&gt;Things change and so does the paperwork. With citations as terms it becomes possible to link them to document management systems and to query them directly within the ontology or the knowledge graph. This allows for the monitoring of which documents are currently referenced by any ontology and whether these will need to be updated upon the publication of a new edition of a reference document.&lt;/p&gt;
&lt;p&gt;From an application developer or end user perspective, consider that this allows for greater transparency of process and communication. Design decisions can be linked backed to the narrative of the original standards with minimal clerical effort since it is integrated to the ontology graph.&lt;/p&gt;
&lt;p&gt;We often talk about AI explainability as if this was an algorithmic design problem, but this isn&#039;t completely true. There is no point explaining the workings of an algorithm if the underlying rules, facts and definitions that communicate with the outside world aren’t available. No amount of abstract logical proofs have value if you are unable to explain if the coded output means bi-annual interest or monthly interest.&lt;/p&gt;
&lt;h2&gt;
	Document, Document, Document&lt;/h2&gt;
&lt;p&gt;Writing documentation is a thankless task: everyone demands it in their preferred style, few will actually eventually read it and more often than not, the select few individuals that do write it have limited resources to do it properly. An ontology-based bibliography lets you automate some of the documentation generation, referencing and cross-referencing in whatever citation format is required.&lt;/p&gt;
&lt;p&gt;Ontologies are primarily objects of communication and interoperability. Some schools of thought, &quot;The Ontology Is The Documentation&quot;, fail to consider that no matter its logical coherence, an ontology is part of a bigger world. The ability to reference outside documents and terms ensures that the ontology functions within an operational contexts and adds to the value of the ontology and saves valuable time for its users.&lt;/p&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 English
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/67&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;rdf&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/64&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;owl&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/68&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;owl2&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/69&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;bibo&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/21&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;ontologies&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/70&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;documentation&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Mon, 22 Jul 2024 01:20:21 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">56 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/56#comments</comments>
</item>
<item>
 <title>Ontologies, what are they good for really?</title>
 <link>https://www.dbdump.org/node/55</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;div&gt;
	Ontologies have a chequered past in the business arena, although many significant benefits have been realised over the past two decades in biological and medical research. Wrapped up in the AI hype and occasionally lumped in with knowledge management projects or taxonomy projects they are often shelved curiosities whose original purpose is lost by their champions. What is an ontology? Generally speaking, it is a set of concepts, their properties and the relationships that link the concepts together.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	The power of Ontologies lies in the management of complex, and often overlooked, facts about an Enterprise. Which of your brands are specific to a market and a language? What is the trade name of this widget in some foreign language? What is its common name? Are shipping manifests linked to a single invoice or can they be composed from multiple invoices? Is this within a single division or across the entire enterprise? All of this data is somewhere within the company be it in employees’ minds, in software configurations or explicitly listed within a standards document. &lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	The benefit of an ontology is that it becomes a single point of reference for all of these items in a format that is machine readable and writable, that permits multilingual documentation and labelling and that can differentiate between what an object is and what it is called. An important point is that while an ontology has to be consistent, it can represent multiple truths and perspectives at the same time. Not all parts of the organisation work the same way; this is perfectly normal and there is great value in documenting how they actually work.&lt;/div&gt;
&lt;h2&gt;
	It’s not just Taxonomies&lt;/h2&gt;
&lt;div&gt;
	People have a natural tendency to categorise and organise information. Taxonomies are one way to create simple hierarchical &quot;trees&quot; to organise concepts and values.  One issue with this method is that different departments in the business have different taxonomy layering structures for the same concepts. This results in many taxonomies being used for some concepts (Products being a typical example) and trying to report or find product characteristics against these different taxonomies can become complex.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	But ontologies go beyond this by adding the capacity to relate the concepts, instantiate the concepts and add labels and descriptions in multiple languages, thus decoupling the identifiers from the representation data. From an enterprise perspective, many applications and systems fail to differentiate between these and the result is that small changes trigger a complete reconstruction of the application’s database. &lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	The separation of the identifier from the content and multilingualism means that data can be updated automatically within downstream software without having to navigate across what the “official translation” for a concept is: Who would you even ask to find out what the localised name of your corporation is in German or whether the description for a particular widget is the same in the German and Austrian markets are the same?&lt;/div&gt;
&lt;h2&gt;
	Master data management. Done right&lt;/h2&gt;
&lt;div&gt;
	Master data management often promotes the idea of capturing information as a single source of truth. This makes sense at first glance: everyone should be on the same page, using the same processes and nomenclatures across the enterprise. However, most businesses are more nuanced, with information viewed and used in multiple business perspectives.  This becomes even more of a challenge with the common use of 3rd party applications that can embed different nuanced meanings depending on the degree of fit for the purpose of the business. Often this implied context, and how to handle the exception, is undocumented and lives on an ad hoc basis within the minds of company personnel. This creates an additional unreported business continuity risk.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	A simple example is a list of office locations: some locations are plants with a shipping yard, multiple buildings, a mailing address for the front office and street addresses for all of the above. The fact that a parcel makes it to the right location will often require either local knowledge from the courier company or a few phone calls and false starts as the right destination is found. &lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	The assignment of an identifier to each sub-location helps to differentiate them all from an ambiguous label “The Pittsburgh Plant”, but fails to record the relationship between them. This creates some concerns for downstream users of the data who have to track the identifier within their own system and attempt to identify the “right one” amongst the list while aggravating end users who just want to send something to “Bob at the Pittsburgh Plant”.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	It is in these cases that ontological solutions shine because of their ability to handle complex data and handle ambiguity. An ontology can easily represent that the office space in building ‘D’ of the Pittsburgh Plant in Westmoreland has its mail routed through the main plant office at 567 Alpha Drive. Finding the “right” entry is left to the system integrators who have full knowledge of their context and who can query the right entry.&lt;/div&gt;
&lt;h2&gt;
	Interoperability&lt;/h2&gt;
&lt;div&gt;
	Lastly, ontologies can be designed to promote Interoperability. Enterprise information systems are made up of multiple pieces of software that look at data from different perspectives and with a specific focus. This often results in having various pieces of the same thing in different locations while making it prohibitively complex to resolve the meaning of the information. &lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	Ontologies and their associated knowledge graphs allow for information that is redundant across different applications to be reconciled and loaded from a central point that can be used to link the data. While one application may only need to be aware of products name and description, another may need compliance information and a third may only be aware of product weight and dimensions for shipping purposes. Keeping all of this information in sync is difficult since applications have different ideas about object identity and identifiers; something that an ontology can easily keep track of.&lt;/div&gt;
&lt;h2&gt;
	Wrapping it all up&lt;/h2&gt;
&lt;div&gt;
	Ontologies and their associated knowledge graphs provide a useful mechanism for handling complex information and translating across different viewpoints of the same instances. Beyond their application to master data management, interoperability and harmonising multiple concurrent taxonomies ontologies are an effective tool to record complex enterprise information. They have been used with success by large companies such as the ‘Fargo’ and ‘Erica’ virtual assistants at Wells Fargo and Bank of America and Google Knowledge Graphs to drive effective business tools. In future posts we will be exploring what tooling, management structure and processes are required to ensure effective ontology use within the company.  &lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	This article was written as part of the &lt;a href=&quot;https://bizonontology.org/&quot;&gt;Bizon Business Ontology&lt;/a&gt; project.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/64&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;owl&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/21&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;ontologies&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/65&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;data integration&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/66&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;interoperability&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Thu, 01 Feb 2024 15:46:50 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">55 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/55#comments</comments>
</item>
<item>
 <title>Books Worth Reading: Most Secret War</title>
 <link>https://www.dbdump.org/node/50</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;(This is the beginning of a series of blog posts on books that aren&#039;t as well known, have an odd topic niche and represent the awesomeness of human ingenuity.)&lt;/p&gt;
&lt;p&gt;&lt;img alt=&quot;Most Secret Book Cover&quot; src=&quot;/sites/one.dbdump.org/images/most_small.png&quot; style=&quot;width: 83px; height: 132px; float: left; border-width: 1px; border-style: solid; margin: 2px;&quot; /&gt;Most Secret War&lt;a href=&quot;#ref1&quot; title=&quot;Reference 1&quot;&gt;[1]&lt;/a&gt; is a book by &lt;a href=&quot;https://en.wikipedia.org/wiki/Reginald_Victor_Jones&quot;&gt;Reginald Jones&lt;/a&gt; about scientific intelligence in the Second World War. One of the inspirations (along with &lt;a href=&quot;https://en.wikipedia.org/wiki/Robert_Watson-Watt&quot;&gt;Robert Watson-Watt&lt;/a&gt;) for Ian Fleming&#039;s James Bond Character Q, Reginald role was to look over incoming intelligence reports for actionable scientific tidbits. This led to serious electronic warfare actions such as the &lt;a href=&quot;https://en.wikipedia.org/wiki/Battle_of_the_Beams&quot;&gt;Battle of the Beams&lt;/a&gt; where the British attempted to jam the German bomber radio navigation signals, the intercept of the German radar station reports during the &lt;a href=&quot;https://en.wikipedia.org/wiki/Battle_of_Britain&quot;&gt;Battle of Britain&lt;/a&gt; and the, at times comical, corruption of German night fighter radio channels.&lt;/p&gt;
&lt;!--break--&gt;&lt;p&gt;He also participated in the planning of the &lt;a href=&quot;https://en.wikipedia.org/wiki/Operation_Biting&quot;&gt;Bruvenal Raid&lt;/a&gt;: faced with a new German radar system (&lt;a href=&quot;https://en.wikipedia.org/wiki/Freya_radar&quot;&gt;Freya&lt;/a&gt;), the British decided that the best way to learn how to defeat it was to mount a raid and steal one. One of his claims to fame was being accused (along with a number of people involved with the project) of chemical warfare: in one of the initial deployment of &lt;a href=&quot;https://en.wikipedia.org/wiki/Chaff_(countermeasure)&quot;&gt;Window&lt;/a&gt; (Radar jamming chaff), the foil strips had been painted black with a toxic paint. The strips landed in a pasture and subsequently poisoned the cows that ate them with the forage. That led some of the German propaganda outlets to charge the British with chemical warfare. In many ways, this book documents both the insanity and the ingenuity of human beings in war time.&lt;/p&gt;
&lt;p&gt;In his later years Reginald Jones kept on the outskirts of government service. His attitudes to Intelligence was that people should have access to as much of the intelligence as they required which didn&#039;t always align with the decreased urgency of post-war Britain and the increase paranoia of the Cold War. He wrote a series of books, including Reflections on Intelligence&lt;a href=&quot;#ref2&quot; title=&quot;Reference 2&quot;&gt;[2]&lt;/a&gt; and lectured often, including to the Royal Canadian Airforce. He passed away in 1997 and in buried in Aberdeenshire.&lt;/p&gt;
&lt;hr /&gt;
&lt;h3&gt;References&lt;/h3&gt;
&lt;div class=&quot;references&quot;&gt;
&lt;ol&gt;
&lt;li id=&quot;reference1&quot;&gt;&lt;a name=&quot;ref1&quot; id=&quot;ref1&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=18&quot; rel=&quot;nofollow&quot;&gt;R. Victor Jones&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;/node/47&quot;&gt;&lt;span class=&quot;biblio-title&quot; style=&quot;font-style: italic;&quot;&gt;Most Secret War&lt;/span&gt;&lt;/a&gt;. Hamish Hamilton Ltd, 1978, p. 556.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&amp;amp;rft.title=Most+Secret+War&amp;amp;rft.date=1978&amp;amp;rft.tpages=556&amp;amp;rft.aulast=Jones&amp;amp;rft.aufirst=Reginald&amp;amp;rft.pub=Hamish+Hamilton+Ltd&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;li id=&quot;reference2&quot;&gt;&lt;a name=&quot;ref2&quot; id=&quot;ref2&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=18&quot; rel=&quot;nofollow&quot;&gt;R. Victor Jones&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;/node/49&quot;&gt;&lt;span class=&quot;biblio-title&quot; style=&quot;font-style: italic;&quot;&gt;Reflections on Intelligence&lt;/span&gt;&lt;/a&gt;. William Heinemann Ltd, 1989, p. 352.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&amp;amp;rft.title=Reflections+on+Intelligence&amp;amp;rft.issn=0434377244&amp;amp;rft.date=1989&amp;amp;rft.tpages=352&amp;amp;rft.aulast=Jones&amp;amp;rft.aufirst=Reginald&amp;amp;rft.pub=William+Heinemann+Ltd&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;/ol&gt;
&lt;/div&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 English
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/60&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Books worth reading&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/3&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;History&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/61&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Books&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;ul class=&quot;links inline&quot;&gt;&lt;li class=&quot;translation_fr first last&quot;&gt;&lt;a href=&quot;/node/51?language=fr&quot; title=&quot;Des livres qui ont de la valeur: Most Secret War&quot; class=&quot;translation-link&quot; xml:lang=&quot;fr&quot;&gt;Français&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;</description>
 <pubDate>Mon, 26 Nov 2018 14:39:32 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">50 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/50#comments</comments>
</item>
<item>
 <title>AI and the Law (Part III) - Algorithmic Bias Is Good For You</title>
 <link>https://www.dbdump.org/node/34</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;&lt;em&gt;&lt;span&gt;(Note: This is the third part of a series of posts that were based on several conversations &lt;/span&gt;with lawyers and executives about &lt;a href=&quot;http://one.dbdump.org/taxonomy/term/10&quot;&gt;AI&lt;/a&gt;, the nature of technology and its application to business problems. The second part is &lt;a href=&quot;http://www.dbdump.org/WhatisAI/PartII&quot;&gt;here&lt;/a&gt;. Warning: Cultural references are sprinkled left, right and centre. You’ve been warned.)&lt;/em&gt;&lt;/p&gt;
&lt;p&gt;This post is about bias in Artificial Intelligence, what it is, what it isn&#039;t and why it&#039;s a feature and not a bug. Scary looking headlines appear in the press about AI taking over people&#039;s job, shouts of &lt;a href=&quot;http://business.financialpost.com/technology/personal-tech/silicon-valley-is-ethically-lost-google-grapples-with-reaction-to-its-new-horrifying-and-uncanny-ai-tech&quot;&gt;silicon valley being ethically lost&lt;/a&gt;, machines taking over the world or generally making a mess of things. Concerns about algorithmic unfairness are not unfounded, but the same remedies that we’ve previously used to resolve unfairness apply.&lt;/p&gt;
&lt;h2&gt;
	Bias Explained&lt;/h2&gt;
&lt;p&gt; &lt;/p&gt;
&lt;p&gt;&lt;img alt=&quot;&quot; src=&quot;/sites/www.dbdump.org/images/AIFirstExample.png&quot; style=&quot;width: 315px; height: 160px; float: left;&quot; /&gt;Bias is defined as &quot;prejudice in favour of or against one thing, person, or group compared with another, usually in a way considered to be unfair&quot;, yet in the AI and mathematical context bias turns out to be something desirable. Figure 1 is the previous classification example from an earlier blog where we have a simple classifier that separates the orange balls from the green balls. The classifier isn&#039;t perfect: it does a reasonable job in most cases but misclassifies one of the oranges balls and there is no way to improve the classifier instance using this linear regression technique because no straight line will separate the green balls from the orange balls.&lt;/p&gt;
&lt;p&gt;However, let&#039;s say that the orange balls represent some outcome that is really negative. Like a cancer going undetected or a critical failure in an engine not being reported, the impact of making that wrong classification is disproportionate with someone having to go through an unnecessary biopsy or the machine being shutdown for unwarranted preventive maintenance. Since the cost of misclassifying that orange ball is so high, we would rather have additional misclassified green balls rather than miss one of the orange ones. That process of deliberately preferring one class over another one is called &lt;em&gt;bias&lt;/em&gt;.&lt;/p&gt;
&lt;!--break--&gt;&lt;p&gt;&lt;img alt=&quot;&quot; src=&quot;/sites/www.dbdump.org/images/biasApplied.png&quot; style=&quot;width: 167px; height: 160px; float: right;&quot; /&gt; Figure two is a graphical representation of this process. Mathematically, if the linear regression is represented by &lt;em&gt;y&lt;/em&gt; = &lt;em&gt;mx&lt;/em&gt; + &lt;em&gt;b&lt;/em&gt;, we increase the value of b so that all orange balls are below the classification threshold. In practice, classification is never perfect and even the application of an extreme bias may still result in an unsatisfactory condition. &lt;strong&gt;However it provides a risk mitigating mechanism that can be quantified based on the expected costs/benefits of each outcome and the classifier that is available.&lt;/strong&gt;&lt;/p&gt;
&lt;p&gt;Let&#039;s say you are using the classifier to route incoming forms to one of two offices for processing, office green and office orange. Office orange has a single clerk whose processing of the form is clerical, office green involves a review committee of five people who must review the form in depth. Assume that the time needed to send a misrouted form back to the other office is the same in each case. The cost of a form misrouted to the office green is thus five times more than office orange since it is taking up the time of five people concurrently. To mitigate the cost of misrouted forms, the classifier will be biased to favour the green office slightly over the orange office, calculated based on the estimated number of green and orange forms and the relative costs of misrouted orange and green forms. This optimized classified allows us to mitigate the risks of the situation using AI and &lt;strong&gt;not the other way around.&lt;/strong&gt;&lt;/p&gt;
&lt;h2&gt;
	&quot;Plus ça change, plus c&#039;est la même chose&quot;&lt;/h2&gt;
&lt;p&gt; &lt;/p&gt;
&lt;p&gt;So why are people so concerned about bias? It&#039;s the possibility of having one&#039;s autonomy restricted by a machine without recourse. Office Space&#039;s &lt;a href=&quot;https://en.wikipedia.org/wiki/PC_LOAD_LETTER&quot;&gt;PC LOAD LETTER&lt;/a&gt; or &quot;&lt;a href=&quot;https://www.youtube.com/watch?v=AJQ3TM-p2QI&quot;&gt;Computer Say No&lt;/a&gt;&quot; jokes aside, the risks of AI are no different than that of other technologies. The perception is that algorithms are somehow biased against ethical principals have been fuelled by everyday examples such as &lt;a href=&quot;https://www.theguardian.com/technology/2018/jan/12/google-racism-ban-gorilla-black-people&quot;&gt;Google auto-tagging black people as gorillas&lt;/a&gt; or Google Street View&#039;s privacy filter failing to blur out someone’s face. But in the final analysis, the AI is executing what humans programmed it to do. AIs are not autonomous, they act as agents of a person or organizations. Speculation about the liability issues of software borne decisions was covered in a well written article by John Kington &lt;a href=&quot;#ref1&quot; title=&quot;Reference 1&quot;&gt;[1]&lt;/a&gt; in the context of current laws: essentially it comes down to whether the product was defective or whether it was being operated properly.&lt;/p&gt;
&lt;p&gt;&lt;strong&gt;Unfairness and biases against certain social groups are nothing new and we’ve developed regulatory frameworks to deal with them.&lt;/strong&gt; But these are human rights that need a human being to make them happen (interestingly, AI works both ways: it can be trained for look for algorithmic unfairness using the same machine readable data). &lt;a href=&quot;https://en.wikipedia.org/wiki/Redlining&quot;&gt;Redlining&lt;/a&gt;, the process of ignoring or promoting services to certain communities based on their physical location is still an ongoing problem today and as long as there is an economic motivation for doing so, it will be necessary to have regulatory enforcement effort on it. Whether the redlining is the result of an unethically trained AI or an unethically planned massed mailing is irrelevant: a person made it happen.&lt;/p&gt;
&lt;p&gt;The possibility of an AI within an organization becoming sentient and achieving their own agency through (il)legal and financial maneuvering is something that will likely remain with &lt;a href=&quot;http://altered-carbon.wikia.com/wiki/Poe&quot;&gt;Altered Carbon characters&lt;/a&gt; for years to come. Efforts to somehow embed ethics into machines date back to 1942, remember &lt;a href=&quot;https://en.wikipedia.org/wiki/Three_Laws_of_Robotics&quot;&gt;the law of Robotics by Isaac Asimov&lt;/a&gt;? Three-quarters of a century later, robots are still dumb as a fence-post, do repetitive labour and whenever a human gets injured, it’s because someone wired the emergency stop button wrong or didn’t lockout the controller. Contrast this with another science fiction scenario: &lt;a href=&quot;https://en.wikipedia.org/wiki/2001:_A_Space_Odyssey_(film)&quot;&gt;instruct a computer system with contradictory and irreconcilable information, neglecting an Emergency Stop button resulting in the machine murdering its human crew&lt;/a&gt;. Humanity has already achieved this: it&#039;s called an industrial accident. They happen daily and we have an entire ecosystem of organizations who work to prevent them from occurring: the safety and ethical principals are the same whether the machine runs on electricity, hydraulics or pneumatics.&lt;/p&gt;
&lt;p&gt;Toy problems with a strong moralistic implication make for a great book, magazine article or panel discussion topic. They don’t map well to operational planning or real life situations for two reasons: a) emergencies occur because of a number of concurrent failures and b) by their nature they are unplanned. By the time an AI, and we are talking about one that would need to be very sophisticated, has actually formulated the problem, the emergency will likely have resolved itself. Will AI really have to choose between two immoral choices? Not in the near future, primarily because AI isn’t all that smart to begin with and there’s a finite amount of computing power to work with. Ethical decisions require a lot of contextual information, processing and are algorithmically ill-behaved even if you know what the “right decision” is. &lt;strong&gt;The self-driving car isn’t going to choose between the baby or grandma &lt;/strong&gt;&lt;a href=&quot;#ref2&quot; title=&quot;Reference 2&quot;&gt;[2]&lt;/a&gt;&lt;strong&gt;, it’s going to choose the path with the longest braking distance, because it is the simplest, fastest and most reliable metric to acquire and make a safe decision with.&lt;/strong&gt;&lt;/p&gt;
&lt;p&gt;Lastly, whenever people talk about algorithmic bias or AI bias, they usually mean an error. Coming back to the google gorilla story, it was an error: classify hundreds of millions of images every year and you’ll eventually make an embarrassing decision eventually. Same for people: sit a person at a chair and have them classify images of traffic light and tombstones every day and eventually they will make an error, either by inattention or due to terminal boredom.&lt;/p&gt;
&lt;p&gt;The actual reasons that unintentional AI training bias occurs is bad training data and/or the repurposing of an engine whose design goals are misaligned with the current objective. Generating training data is expensive, time consuming and a little back and forth is needed to get the system working properly. An efficient means of creating training data isn’t always available. Graduate students trying to test things scour their research department for training data (old emails, mugshots, sound bytes) and invariably end up drafting their friends and acquaintances which creates a bias all of its own. Good enough for academic testing, those models are sometimes coop’ted by cash starved startups. Similarly, there are limited code bases available for each artificial intelligence methodology. Whatever optimization was built in and long forgotten may have unintended consequences when used for a different application. Careful with that axe Eugene.&lt;/p&gt;
&lt;h2&gt;
	Why do we really need AI?&lt;/h2&gt;
&lt;p&gt; &lt;/p&gt;
&lt;p&gt;It is a fair question. Why should we use AI when we could be using people to do the work and generate employment? Irrespective of the market forces that effect the labour costs, the fact is that&lt;strong&gt; the rate at which we generate information is many times greater than that at which we are capable of processing it&lt;/strong&gt;. Or, as every parent will tell you, the capacity of a 5 year old to ask questions is wildly disproportionate to theirs parents combined capacity to answer them all. Part of the problem is human beings: we haven&#039;t changed much since Julius Caesar crossed the Rubicon but our society has changed dramatically: only a very few of the people reading this will remember a world without the internet where just getting the phone number for someone out of town was a serious information search. Our capacity to personally process information has not kept up with the dramatic increase in data.&lt;/p&gt;
&lt;p&gt;Getting more people to process the information is also not an answer: as more people enters the system, they themselves generate more information and the situation becomes a vicious circle. According to a popular quote from Google&#039;s CEO, every 48 hours humankind creates more data than it did before 1993. There is no catching up with current methods. Our current approaches to data management have resulted in some embarrassing responses &lt;a href=&quot;#ref3&quot; title=&quot;Reference 3&quot;&gt;[3]&lt;/a&gt; to ATIP requests with estimates of over 80 years! It is obviously not acceptable for straightforward information requests not to be answerable within the lifetime of a human being.&lt;/p&gt;
&lt;p&gt;The only real viable solution is cognitive automation through AI. There has been much research on information retrieval and question answering over the years, most it comes from an age where the rarity of data made finding what you were looking for easy. Studies about how human beings handle information has revealed that we aren’t really good at it: attempting to create gold standard classification sets at the American National Insitute of Standards revealed that for non-trivial tasks, people are only accurate 40% of the time &lt;a href=&quot;#ref4&quot; title=&quot;Reference 4&quot;&gt;[4]&lt;/a&gt;&lt;a href=&quot;#ref5&quot; title=&quot;Reference 5&quot;&gt;[5]&lt;/a&gt;. Using reviewing panels, multiple accessors and some statistical tricks, the precision / recall curve can only be risen to 65% at 65%. This means that careful thinking about accuracy should be done when using both AI and people: AI’s are great at making simple decisions over and over again while humans do better at decision that require a lot of context.&lt;/p&gt;
&lt;p&gt;An example that I like to use is the most important business meeting of your life. You’ve locked the office door, unplugged the phone, turned off the cell phone and made everyone swear that they wouldn’t disturb you. When your spouse has a car accident, calls from the emergency room payphone with your child going in for emergency surgery, your colleague will charge into your office to tell you to take the call. The AI will do what you asked and leave you alone. &lt;strong&gt;The value of people is that they have judgement, AIs don’t.&lt;/strong&gt;&lt;/p&gt;
&lt;h2&gt;
	Conclusion&lt;/h2&gt;
&lt;p&gt; &lt;/p&gt;
&lt;p&gt;Artificial Intelligence is here to stay and at the pace that things are going, necessary for our civilization to keep functioning. As with any new tool, reading the safety warnings and the manual is a good idea. Writers, wether academic, press or science fiction, tend to play up the good story, predict exciting situations that can end up in gloom and doom. But the reality is that to work well, AI projects need to be focused on a single problem and well tested. The thoughts of &lt;a href=&quot;https://en.wikipedia.org/wiki/Skynet_(Terminator)&quot;&gt;skynet&lt;/a&gt;, or the &lt;a href=&quot;https://en.wikipedia.org/wiki/WarGames&quot;&gt;War Games&lt;/a&gt; WOPR, taking over NORAD computers is a good movie plot, but it won’t happen anytime soon.&lt;/p&gt;
&lt;hr /&gt;
&lt;h3&gt;References&lt;/h3&gt;
&lt;div class=&quot;references&quot;&gt;
&lt;ol&gt;
&lt;li id=&quot;reference1&quot;&gt;&lt;a name=&quot;ref1&quot; id=&quot;ref1&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=9&quot; rel=&quot;nofollow&quot;&gt;J. K. C.  Kingston&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;https://arxiv.org/ftp/arxiv/papers/1802/1802.07782.pdf&quot;&gt;“&lt;span class=&quot;biblio-title&quot;&gt;Artificial intelligence and legal liability&lt;/span&gt;”&lt;/a&gt;, in &lt;span style=&quot;font-style: italic;&quot;&gt;International Conference on Innovative Techniques and Applications of Artificial Intelligence&lt;/span&gt;, 2016.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&amp;amp;rft.title=Artificial+intelligence+and+legal+liability&amp;amp;rft.date=2016&amp;amp;rft.aulast=Kingston&amp;amp;rft.aufirst=John&amp;amp;rft.pub=Springer&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;li id=&quot;reference2&quot;&gt;&lt;a name=&quot;ref2&quot; id=&quot;ref2&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=10&quot; rel=&quot;nofollow&quot;&gt;K.  Hao&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;https://www.technologyreview.com/s/612341/a-global-ethics-study-aims-to-help-ai-solve-the-self-driving-trolley-problem/&quot;&gt;“&lt;span class=&quot;biblio-title&quot;&gt;Should a self-driving car kill the baby or the grandma? Depends on where you’re from&lt;/span&gt;”&lt;/a&gt;, &lt;span style=&quot;font-style: italic;&quot;&gt;Technology Review&lt;/span&gt;, 2018.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&amp;amp;rft.title=Should+a+self-driving+car+kill+the+baby+or+the+grandma%3F+Depends+on+where+you%E2%80%99re+from&amp;amp;rft.date=2018&amp;amp;rft.aulast=Hao&amp;amp;rft.aufirst=Karen&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;li id=&quot;reference3&quot;&gt;&lt;a name=&quot;ref3&quot; id=&quot;ref3&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=8&quot; rel=&quot;nofollow&quot;&gt;A.  Boutilier&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;/node/41&quot;&gt;“&lt;span class=&quot;biblio-title&quot;&gt;Your document request will be fulfilled by 2098&lt;/span&gt;”&lt;/a&gt;, &lt;span style=&quot;font-style: italic;&quot;&gt;The Toronto Star&lt;/span&gt;, Toronto, Ontario, 2018.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&amp;amp;rft.title=Your+document+request+will+be+fulfilled+by+2098&amp;amp;rft.date=2018&amp;amp;rft.aulast=Boutilier&amp;amp;rft.aufirst=Alex&amp;amp;rft.place=Toronto%2C+Ontario&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;li id=&quot;reference4&quot;&gt;&lt;a name=&quot;ref4&quot; id=&quot;ref4&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=11&quot; rel=&quot;nofollow&quot;&gt;D.  Harman&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.27.7118&amp;amp;rep=rep1&amp;amp;type=pdf&quot;&gt;“&lt;span class=&quot;biblio-title&quot;&gt;What we have learned, and not learned&lt;/span&gt;”&lt;/a&gt;, &lt;span style=&quot;font-style: italic;&quot;&gt;The BCS/IRSG 22nd Annual Colloquium on Information Retrieval Research&lt;/span&gt;, pp. 2-20, 2000.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&amp;amp;rft.atitle=What+we+have+learned%2C+and+not+learned&amp;amp;rft.title=The+BCS%2FIRSG+22nd+Annual+Colloquium+on+Information+Retrieval+Research&amp;amp;rft.date=2000&amp;amp;rft.spage=2&amp;amp;rft.epage=20&amp;amp;rft.aulast=Harman&amp;amp;rft.aufirst=Donna&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;li id=&quot;reference5&quot;&gt;&lt;a name=&quot;ref5&quot; id=&quot;ref5&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=12&quot; rel=&quot;nofollow&quot;&gt;M. R.  Grossman&lt;/a&gt;, &lt;a href=&quot;/biblio?f%5Bauthor%5D=13&quot; rel=&quot;nofollow&quot;&gt;Cormack, G. V.&lt;/a&gt;, &lt;a href=&quot;/biblio?f%5Bauthor%5D=14&quot; rel=&quot;nofollow&quot;&gt;Hedin, B.&lt;/a&gt;, and &lt;a href=&quot;/biblio?f%5Bauthor%5D=15&quot; rel=&quot;nofollow&quot;&gt;Oard, D. W.&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;http://www.ee.umd.edu/ oard/pdf/trecov11.pdf&quot;&gt;“&lt;span class=&quot;biblio-title&quot;&gt;Overview of the TREC 2011 legal track&lt;/span&gt;”&lt;/a&gt;, in &lt;span style=&quot;font-style: italic;&quot;&gt;Proc. 20th Text REtrieval Conference&lt;/span&gt;, 2010.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&amp;amp;rft.title=Overview+of+the+TREC+2011+legal+track&amp;amp;rft.date=2010&amp;amp;rft.aulast=Grossman&amp;amp;rft.aufirst=Maura&amp;amp;rft.au=Cormack%2C+Gordon&amp;amp;rft.au=Hedin%2C+Bruce&amp;amp;rft.au=Oard%2C+Douglas&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;/ol&gt;
&lt;/div&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 English
&lt;/div&gt;
&lt;ul class=&quot;links inline&quot;&gt;&lt;li class=&quot;translation_fr first last&quot;&gt;&lt;a href=&quot;/node/54?language=fr&quot; title=&quot;L&amp;#039;IA et la Loi (Partie III) - La Partialités Algorithmique est Bonne Pour Vous&quot; class=&quot;translation-link&quot; xml:lang=&quot;fr&quot;&gt;Français&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;</description>
 <pubDate>Tue, 20 Nov 2018 14:37:23 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">34 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/34#comments</comments>
</item>
<item>
 <title>What is going wrong with the Semantic Web?</title>
 <link>https://www.dbdump.org/node/37</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;The &lt;a href=&quot;http://us2ts.org/&quot;&gt;US Semantic Technologies Symposium&lt;/a&gt; was held at &lt;a href=&quot;https://www.wright.edu/&quot;&gt;Wright State University&lt;/a&gt; a month ago where there were great discussions with &lt;a href=&quot;https://twitter.com/caknoblock&quot;&gt;Craig Knoblock&lt;/a&gt; about SPARQL servers reliability&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref1_7rtdexj&quot; title=&quot;Short version, my experiences with the Muninn Project, CWRC, CLDI and Myra have been positive. Overall SPARQL servers have had less engineering calendar time than other comparable software: Apache and Mysql have been worked on since 1995, Postgresql since 1986. In contrast, Virtuoso has had SPARQL since 2005, Alleograph 2004 and ARC2 2010. 10+ extra years of development work helps. Furthermore, Mondeca&#039;s SPARQL endpoint monitor show that SPARQL servers do have good uptime. The often misquoted 63% of endpoints being offline applies to every SPARQL server ever seen since 2013. The statistic that should be worrisome is that only 13% of them have ever had a machine readable description!&quot; href=&quot;#footnote1_7rtdexj&quot;&gt;1&lt;/a&gt;, &lt;a href=&quot;https://twitter.com/ekansa?lang=en&quot;&gt;Eric Kansa&lt;/a&gt; about storing archeology data and &lt;a href=&quot;https://opencontext.org/&quot;&gt;Open Context&lt;/a&gt;, &lt;a href=&quot;https://www.w3.org/People/EM/&quot;&gt;Eric Miller&lt;/a&gt; about the workings of the W3, mid West farmers and old bikes, &lt;a href=&quot;https://foodscience.ucdavis.edu/people/matthew-lange&quot;&gt;Matthew Lange&lt;/a&gt; about tracking crops with LOD and a &#039;fruitful&#039; talk with &lt;a href=&quot;https://www.nist.gov/people/evan-k-wallace&quot;&gt;Evan Wallace&lt;/a&gt; about farm data storage standards. &lt;/p&gt;
&lt;p&gt;Thinking through these conversations, I decided to outline what I think are the troubling conclusions for our area, namely that a) &lt;a href=&quot;#part_1&quot;&gt;Semantic Web adoption is lagging&lt;/a&gt;, b) &lt;a href=&quot;#part_2&quot;&gt;we keep rehashing old problems without moving on&lt;/a&gt; and c) &lt;a href=&quot;#part_3&quot;&gt;our ongoing lack of support for our own projects&lt;/a&gt; after which I&#039;ll suggest a &lt;a href=&quot;#conclusion&quot;&gt;few solutions&lt;/a&gt;.&lt;/p&gt;
&lt;!--break--&gt;&lt;h3&gt;
	&lt;a name=&quot;part_1&quot; id=&quot;part_1&quot;&gt;Semantic Web adoption is not where we&#039;d like it to be&lt;/a&gt;&lt;/h3&gt;
&lt;p&gt;Very, very few people care about data management&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref2_czzri57&quot; title=&quot;Data management is the simplest redux of the semantic web and ontology. I&#039;m setting the bar low on purpose...&quot; href=&quot;#footnote2_czzri57&quot;&gt;2&lt;/a&gt;. Even fewer people understand data management. I&#039;d go so far to say that the majority of the IT community spends its time moving strings to the end-user&#039;s screen, focusing primarily on user communications &lt;strong&gt;and getting told &lt;em&gt;what&lt;/em&gt; to communicate&lt;/strong&gt;. Other developers may worry about analysis, networking stacks or storage, but the number of them that care about the data itself are few.&lt;/p&gt;
&lt;p&gt;That leaves the database developer whose entire attention is on bread and butter issues. &lt;strong&gt;Why did we have the hubris to think that people would care about the Semantic Web when most of them have no data management problem to worry about?&lt;/strong&gt; &lt;a href=&quot;https://queue.acm.org/detail.cfm?id=2857276&quot;&gt;31% of webpages are reported to contain schema.org&lt;/a&gt;&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref3_qjxki1e&quot; title=&quot;It would be interesting to see how many of these triples are well formed, sensical and form a data structure that makes syntactically.&quot; href=&quot;#footnote3_qjxki1e&quot;&gt;3&lt;/a&gt;, primarily because Web Developers believe it will help them with SEO issues, not because it helps them with data management or interoperability.&lt;/p&gt;
&lt;h3&gt;
	&lt;a name=&quot;part_2&quot; id=&quot;part_2&quot;&gt;Reinventing the wheel. Again.&lt;/a&gt;&lt;/h3&gt;
&lt;p&gt;I regret I didn&#039;t note the speaker who said &quot;&lt;strong&gt;&lt;em&gt;If your developers care about JSON, I don&#039;t care about your developers&lt;/em&gt;&lt;/strong&gt;&quot;, because it goes to the heart of the matter about poor Semantic Web training and education. At this stage &lt;strong&gt;arguments about serializations are about as relevant as debating whether submarines can swim&lt;/strong&gt;&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref4_cg443oq&quot; title=&quot;With apologies to Edsger Dijkstra.&quot; href=&quot;#footnote4_cg443oq&quot;&gt;4&lt;/a&gt;. The was a lot of talk at the meeting about creating new json standards to handle corner cases without knowledge or regards for previous standards because &quot;&lt;em&gt;it&#039;s not JSON and people want JSON&lt;/em&gt;&quot;. The Semantic Web stack translates the model to whatever serialization is needed, in most cases negotiated without programmer involvement. JSON is really nice for web developers, RDF/XML for XPATH, turtle for authoring, n3 for throughput, et al. David Booth also noted the panoply of standards and vocabularies. A number of them have been beautifully engineered by domain experts (&lt;a href=&quot;http://www.opengeospatial.org/standards/geosparql&quot;&gt;GeoSPARQL&lt;/a&gt;&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref5_snt0njk&quot; title=&quot;The name is somewhat of a misnomer since the standard contains both an ontology to describe both feature and geometry, as well as SPARQL extensions meant to spatially reason over the data. It is based on previous OGC work and is rock solid.&quot; href=&quot;#footnote5_snt0njk&quot;&gt;5&lt;/a&gt;, &lt;a href=&quot;https://www.w3.org/TR/owl-time/&quot;&gt;OWL-TIME&lt;/a&gt;, &lt;a href=&quot;https://www.w3.org/TR/vocab-ssn/&quot;&gt;SOSA&lt;/a&gt; and &lt;a href=&quot;https://www.w3.org/TR/prov-o/&quot;&gt;PROV&lt;/a&gt; come to mind), it&#039;s an outright waste of everyone&#039;s time not to reuse them.&lt;/p&gt;
&lt;h3&gt;
	&lt;a name=&quot;part_3&quot; id=&quot;part_3&quot;&gt;We expect research groups to act like service providers&lt;/a&gt;&lt;/h3&gt;
&lt;p&gt;The lack of reliable services and exemplars was also noted: the curated &lt;a href=&quot;https://stackoverflow.com/questions/37858222/new-york-times-rdf-data-dump-or-sparql-endpoint&quot;&gt;New York Times RDF dataset is no longer answering&lt;/a&gt;, the BBC has cut back on &lt;a href=&quot;http://iphylo.blogspot.ca/2017/12/blue-planet-ii-bbc-and-semantic-web.html&quot;&gt;outward&lt;/a&gt; &lt;a href=&quot;http://lists.infradead.org/pipermail/get_iplayer/2016-January/008598.html&quot;&gt;looking&lt;/a&gt; Semantic Web services and DBPedia, at the heart of the &lt;a href=&quot;https://medium.com/openlink-software-blog/what-is-dbpedia-and-why-is-it-important-d306b5324f90#.a7naa79ht&quot;&gt;LOD cloud&lt;/a&gt;, is still running on a borrowed virtual machine with the &lt;a href=&quot;http://wiki.dbpedia.org/dbpedia-association&quot;&gt;DBPedia Association&lt;/a&gt; having a hard time raising funds. I would like to echo &lt;a href=&quot;http://www.juansequeda.com/blog/2018/03/05/trip-report-2018-us2ts/&quot;&gt;Juan Sequeda&lt;/a&gt;&#039;s post that we should set aside some grant monies for resources such as &lt;a href=&quot;http://lov.okfn.org/dataset/lov/&quot;&gt;Linked Open Vocabularies&lt;/a&gt;, a  great vocabulary/ontology location tool&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref6_20qt4hb&quot; title=&quot;Developed by Bernard Vatant and Pierre-Yves Vandenbussche .&quot; href=&quot;#footnote6_20qt4hb&quot;&gt;6&lt;/a&gt; Getting operational funding is always a slog but we cannot &lt;strong&gt;advocate for a technology when the exemplars are not maintained and disappear overnight.&lt;/strong&gt;&lt;/p&gt;
&lt;p&gt;In the past we&#039;ve gotten away with a lot by stuffing machines under graduate students&#039;s desks and getting them to write applications between course work and thesis submission. This is not sustainable and we need to make an effort on long term sustainability.&lt;/p&gt;
&lt;h3&gt;
	&lt;a name=&quot;conclusion&quot; id=&quot;conclusion&quot;&gt;What we should be doing&lt;/a&gt;&lt;/h3&gt;
&lt;p&gt;The Semantic Web stack is annoyingly complex, not because of the technology but because of the problems that it is trying to solve. Its critics abound (&lt;a href=&quot;https://www.youtube.com/watch?v=4egkml10rN8&quot;&gt;even Hitler apparently&lt;/a&gt;) but there is no real alternative to deal with data at scale. Organizationally, it sits uncomfortably between two communities:&lt;/p&gt;
&lt;p&gt;The first is the small group of developers that deal with &lt;a href=&quot;https://en.wikipedia.org/wiki/Web_API&quot;&gt;web apis&lt;/a&gt;, mostly independently from each other. Integrations are done on an ad-hoc basis when the one-off business requirement presents itself. These are the people that came up with ideas like &lt;a href=&quot;https://swagger.io/&quot;&gt;Swagger&lt;/a&gt;: simple documentation that&lt;strong&gt; focuses on programmatic operations with little semantics about the transaction itself&lt;/strong&gt;. Want it in &lt;span style=&quot;color:#ff8c00;&quot;&gt;Orange&lt;/span&gt;? Set colour_id to 2. Why 2? Because that&#039;s the value some developer arbitrarily decided on at the time. Why is your self-evident use case not handled? Because no one has needed it before. Development is incremental. If an error occurs, put in a ticket into github, no harm done.&lt;/p&gt;
&lt;p&gt;The second is the &lt;a href=&quot;https://en.wikipedia.org/wiki/Enterprise_resource_planning&quot;&gt;Enterprise Resource Planning&lt;/a&gt; crowd that has has been doing this for a very long time, albeit usually within a single organization and with massive amounts of corporate resources. Because they care deeply that orders of 5,000 sheets of 8.5x11 paper aren&#039;t interpreted as orders of 8,511 sheets of 5000in&lt;sup&gt;2&lt;/sup&gt; paper, &lt;strong&gt;they tend to document everything (a single API document may run 100&#039;s of pages) and have a neurotic attention to change management&lt;/strong&gt;. There have been &lt;a href=&quot;https://en.wikipedia.org/wiki/Phoenix_Pay_System&quot;&gt;spectacular&lt;/a&gt; &lt;a href=&quot;https://www.computerdealernews.com/news/sobeys-fires-sap-over-erp-debacle/22906&quot;&gt;failures&lt;/a&gt; when implementing these mammoth&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref7_9ackphz&quot; title=&quot;Without putting Alessandro Oltramari on the spot, it took Robert Bosch over a decade to get everything running and it is considered a case-worthy installation.&quot; href=&quot;#footnote7_9ackphz&quot;&gt;7&lt;/a&gt; systems, but generally you can order something from across the world and it will show up on your doorstep next week. &lt;/p&gt;
&lt;p&gt;The Semantic Web has a lot to offer to both these communities: a ready made &lt;strong&gt;semantic modelling language&lt;/strong&gt;&lt;a class=&quot;see-footnote&quot; id=&quot;footnoteref8_rx4ansr&quot; title=&quot;Notwithstanding some early OWL missteps mentioned by Deborah McGuinness, the basic ontological framework underneath the semantic web is extremely powerful and a godsend for data integration.&quot; href=&quot;#footnote8_rx4ansr&quot;&gt;8&lt;/a&gt; that is reusable by web apis, &lt;strong&gt;URL-based global identifiers&lt;/strong&gt; and a &lt;strong&gt;unified multilingual documentation framework&lt;/strong&gt; than fits corporate needs. Bridges need to be built with application domain experts and with existing data eco-systems. Logistics systems such as &lt;a href=&quot;https://en.wikipedia.org/wiki/Global_Trade_Item_Number&quot;&gt;Global Trade Item Number&lt;/a&gt; are pushing the limits of what we can do with barcodes and relational databases. We want the &lt;a href=&quot;https://en.wikipedia.org/wiki/Internet_of_things&quot;&gt;Internet of Things&lt;/a&gt;, the &lt;a href=&quot;https://www.ic-foods.org/&quot;&gt;Internet of Food&lt;/a&gt;, a &lt;a href=&quot;https://en.wikipedia.org/wiki/Smart_grid&quot;&gt;smart power and transportation grid&lt;/a&gt; and a bibliographic system that isn&#039;t going to split it&#039;s seams. The only way that we can achieve all of this is to have the data that is being generated supported by content and the Semantic Web.&lt;/p&gt;
&lt;p&gt;  &lt;/p&gt;
&lt;!--break--&gt;&lt;ul class=&quot;footnotes&quot;&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote1_7rtdexj&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref1_7rtdexj&quot;&gt;1.&lt;/a&gt; Short version, my experiences with the &lt;a href=&quot;http://www.muninn-project.org&quot;&gt;Muninn Project&lt;/a&gt;, &lt;a href=&quot;http://www.cwrc.ca&quot;&gt;CWRC&lt;/a&gt;, &lt;a href=&quot;https://connect.library.utoronto.ca/display/U5LD/Canadian+Linked+Data+Initiative+Home&quot;&gt;CLDI&lt;/a&gt; and &lt;a href=&quot;http://www.myraanalytics.ca&quot;&gt;Myra&lt;/a&gt; have been positive. Overall SPARQL servers have had less engineering calendar time than other comparable software: Apache and Mysql have been worked on since 1995, Postgresql since 1986. In contrast, Virtuoso has had SPARQL since 2005, Alleograph 2004 and ARC2 2010. &lt;strong&gt;10+ extra years&lt;/strong&gt; of development work helps. Furthermore, Mondeca&#039;s &lt;a href=&quot;http://sparqles.ai.wu.ac.at&quot;&gt;SPARQL endpoint&lt;/a&gt; monitor show that SPARQL servers do have good uptime. The often misquoted 63% of endpoints being offline applies to every SPARQL server &lt;a href=&quot;https://github.com/pyvandenbussche/sparqles/issues/20&quot;&gt;ever seen since 2013&lt;/a&gt;. The statistic that should be worrisome is that &lt;strong&gt;only 13% of them have &lt;em&gt;ever&lt;/em&gt; had a machine readable description&lt;/strong&gt;!&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote2_czzri57&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref2_czzri57&quot;&gt;2.&lt;/a&gt; Data management is the simplest redux of the semantic web and ontology. I&#039;m setting the bar low on purpose...&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote3_qjxki1e&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref3_qjxki1e&quot;&gt;3.&lt;/a&gt; It would be interesting to see how many of these triples are well formed, sensical and form a data structure that makes syntactically.&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote4_cg443oq&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref4_cg443oq&quot;&gt;4.&lt;/a&gt; With apologies to &lt;a href=&quot;https://en.wikiquote.org/wiki/Edsger_W._Dijkstra&quot;&gt;Edsger Dijkstra&lt;/a&gt;.&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote5_snt0njk&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref5_snt0njk&quot;&gt;5.&lt;/a&gt; The name is somewhat of a misnomer since the standard contains both an ontology to describe both feature and geometry, as well as SPARQL extensions meant to spatially reason over the data. It is based on previous &lt;a href=&quot;http://www.opengeospatial.org/&quot;&gt;OGC&lt;/a&gt; work and is rock solid.&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote6_20qt4hb&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref6_20qt4hb&quot;&gt;6.&lt;/a&gt; Developed by &lt;a href=&quot;https://bvatant.blogspot.ca/&quot;&gt;Bernard Vatant&lt;/a&gt; and &lt;a href=&quot;https://plus.google.com/+PierreYvesVandenbussche&quot;&gt;Pierre-Yves Vandenbussche&lt;/a&gt; .&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote7_9ackphz&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref7_9ackphz&quot;&gt;7.&lt;/a&gt; Without putting Alessandro Oltramari on the spot, it took Robert &lt;a href=&quot;https://books.google.de/books?hl=de&amp;amp;lr=&amp;amp;id=sVS9AQAAQBAJ&amp;amp;oi=fnd&amp;amp;pg=PP1&amp;amp;ots=Zs1hm0f7C2&amp;amp;sig=4BhqgAq5dVGuIIr772fmKwmXo-0#v=onepage&amp;amp;q&amp;amp;f=false&quot;&gt;Bosch over a decade to get everything running&lt;/a&gt; and it is considered a case-worthy installation.&lt;/li&gt;
&lt;li class=&quot;footnote&quot; id=&quot;footnote8_rx4ansr&quot;&gt;&lt;a class=&quot;footnote-label&quot; href=&quot;#footnoteref8_rx4ansr&quot;&gt;8.&lt;/a&gt; Notwithstanding some early OWL missteps mentioned by Deborah McGuinness, the basic ontological framework underneath the semantic web is extremely powerful and a godsend for data integration.&lt;/li&gt;
&lt;/ul&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 English
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/13&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;lod&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/46&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Semantic Web&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/57&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Education&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/58&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Progress&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/21&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;ontologies&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/59&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;US2TS&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;ul class=&quot;links inline&quot;&gt;&lt;li class=&quot;translation_fr first last&quot;&gt;&lt;a href=&quot;/node/38?language=fr&quot; title=&quot;Ce qui va mal avec le Web Sémantique&quot; class=&quot;translation-link&quot; xml:lang=&quot;fr&quot;&gt;Français&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;</description>
 <pubDate>Tue, 10 Apr 2018 17:15:34 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">37 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/37#comments</comments>
</item>
<item>
 <title>Talk: Ontologies, Semantic Web, and Linked Data for Business</title>
 <link>https://www.dbdump.org/node/35</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;&lt;img alt=&quot;&quot; src=&quot;http://one.dbdump.org/sites/www.dbdump.org/images/cqads-l-logo.png&quot; style=&quot;width: 331px; height: 125px; float: right;&quot; /&gt;&lt;/p&gt;
&lt;div&gt;
	&lt;a href=&quot;http://cqads.carleton.ca/content/ontologies-semantic-web-and-linked-data-business&quot;&gt;Ontologies, Semantic Web, and Linked Data for Business &lt;/a&gt;&lt;/div&gt;
&lt;div&gt;
	13 February 2018, from 10am to 2pm.&lt;/div&gt;
&lt;div&gt;
	Carleton University, &lt;a href=&quot;http://cqads.carleton.ca/&quot;&gt;Centre for Quantitative Analysis and Decision Support&lt;/a&gt;&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;p&gt;This is a half-day workshop about the current business uses of the semantic web. It is targeted at executives, project managers and subject matter experts who want to understand what problems the technology can solve. This workshop will concern itself with the basic building blocks of the semantic web and the solutions that each aspect brings to an organization. The objective of the workshop is not to provide in-depth technical training, rather, we wish to present an overview that will enable a varied audience to determine what this technology provides for their organization. Specific aspects will include recent standards such as the &lt;a href=&quot;https://www.edmcouncil.org/financialbusiness&quot;&gt;FIBO&lt;/a&gt; and &lt;a href=&quot;http://schema.org&quot;&gt;schema.org&lt;/a&gt; and the recruitment and training of staff, as well as opportunities for localization to different markets and lowering the cost of regulatory reporting. In order to anchor the discussions the tribulations of a fictional company, &quot;The Triples Coffee Company&quot;, will be used to present business cases within different areas of an enterprise. Example solutions will then be outlined using a semantic web approach for each business case.&lt;/p&gt;
&lt;p&gt;Jointly presented with Robert Warren, Ph.D., Jennifer Schellinck, Ph.D., Patrick Boily, Ph.D.&lt;/p&gt;
&lt;!--break--&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/49&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;FIBO&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/50&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;schema.org&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/13&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;lod&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/51&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Training&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Tue, 28 Nov 2017 15:03:32 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">35 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/35#comments</comments>
</item>
<item>
 <title>AI and the Law (Part II) - How AI Works</title>
 <link>https://www.dbdump.org/WhatisAI/PartII</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;&lt;cite&gt;&lt;em&gt;&lt;span&gt;(Note: This is the second part of a series of posts that were based on several conversations &lt;/span&gt;with lawyers and executives about &lt;a href=&quot;http://one.dbdump.org/taxonomy/term/10&quot;&gt;AI&lt;/a&gt;, the nature of technology and its application to business problems. The first part is &lt;a href=&quot;http://www.dbdump.org/WhatisAI/PartI&quot;&gt;here&lt;/a&gt;.)&lt;/em&gt;&lt;/cite&gt;&lt;/p&gt;
&lt;p&gt;At the heart of it, AI is about asking the following question: &quot;Can I use the computer to make decisions that would normally require a human being?&quot; Of course, the obvious answer is yes; human beings make all sorts of decisions all day long, ranging from the complex to the mundane. &lt;img alt=&quot;&quot; src=&quot;/sites/www.dbdump.org/images/AIFirstExample.png&quot; style=&quot;width: 315px; height: 160px; float: right;&quot; /&gt;Accounting and operations systems have been making decisions for human beings for years, be it from calculating credit scores and interest rates to determining the best time to order feedstock.&lt;/p&gt;
&lt;p&gt;Let&#039;s use a simple example to explain the difference. Take the plot on the right hand side where I have orange dots and green dots. With basic statistical methods (linear regression was invented in the early 1900&#039;s&lt;a href=&quot;#ref1&quot; title=&quot;Reference 1&quot;&gt;[1]&lt;/a&gt;), we can create a simple classifier that will separate the green from the orange by simply drawing a line through the graph. It&#039;s not perfect, some oranges are misclassified as green and vice-versa, but we do very well with a really simple method.&lt;img alt=&quot;&quot; src=&quot;/sites/www.dbdump.org/images/AI1a.png&quot; style=&quot;width: 164px; height: 160px; float: left;&quot; /&gt; We can do better using mathematical techniques that are more sophisticated, see a non-linear method on the left, but fundamentally the problem remains simple: telling the granny smith apples from the oranges.&lt;/p&gt;
&lt;p&gt;The problem is simple to solve because it is well defined. The objective is clear, the definition of success is clear (keep both sets of dots separated) and the way to tell them apart is by their colour. The only thing that remains is applying the recipe that matches the problem, a linear regression in this case, to solve the problem. Loosely speaking, there is no intelligence needed because the problem defines its own process to a solution.&lt;/p&gt;
&lt;p&gt;Similarly, let&#039;s take another toy problem: tic-tac-toe. Every school child, even those that don&#039;t eventually end up working on AI, learns that the game can be tied or won by the first person playing. The second player can always force a tie, but can never win the game. All of us learn this by playing the game repeatedly while young: children over time explore the set of possible game layouts in tic tac toe and eventually learn that there are a finite number of starting X&#039;s and O&#039;s that can lead to victory or loss.&lt;/p&gt;
&lt;p&gt;Very roughly, there are &lt;a href=&quot;https://en.wikipedia.org/wiki/Tic-tac-toe&quot;&gt;over 360,000 possible positions in tic-tac-toe&lt;/a&gt;. One basic machine learning method to learn to play tic-tac-toe is brute-force: try every single move and counter-move until every game is enumerated and then only choose moves that end in winning the game. Obviously, children can&#039;t keep track of 360,000 tic-tac-toe boards at the same time and while effective, the method does not scale well even for computers (Your desktop computer can&#039;t store the &lt;a href=&quot;https://en.wikipedia.org/wiki/Solving_chess&quot;&gt;10^47 possible combinations of chess&lt;/a&gt;). Therefore, children learn to take shortcuts and reduce those 360,000 positions into a set of starting moves that ensures that they win the game every time. That is intelligence: no one taught them the process required to find the solution, they just did. Artificial Intelligence is the science of creating algorithms that can do the same for certain classes of problems.&lt;/p&gt;
&lt;p&gt;We say &quot;classes of problems&quot; because we only have the computing power and the programming know-how to handle limited problems, like recognizing a person, a text or a musical genre. This is different from what people sometimes refer to as &quot;&lt;a href=&quot;https://en.wikipedia.org/wiki/Artificial_general_intelligence&quot;&gt;True AI&lt;/a&gt;&quot;, which is the human-like machine that walks and talks (and invariably tries to take over the world) on television shows. In my opinion, you are unlikely to have the Terminator do your filing for you anytime in the near future.&lt;/p&gt;
&lt;p&gt;However for specific types of problems, AI works very well: classification, clustering, searching, reduction, etc... In turn, this means that the most of the work that goes into implementing an AI engine is actually trying to match very simple mathematical solutions to complex business problem. Going back to our first example of apples and oranges, the problem was delivered on a plate: colour and position. The solution needs more thinking when the problem isn&#039;t so well defined. In some cases, we know that some of the dots are different but not why or which ones (eg: Outlier Detection). In others, the objective may be to &quot;Group the similar dots together&quot; without having any idea of what makes the dots similar (eg: Market Segmentation). &lt;/p&gt;
&lt;h2&gt;
	Machine Learning vs Artificial Intelligence&lt;/h2&gt;
&lt;p&gt;In the vernacular, the terms Machine Learning and Artificial Intelligence are sometimes used interchangeably through they refer to different things. Artificial Intelligence is the catch all phrase for different computational techniques that have an intelligence component to them irrespective of the flexibility or adaptability of the method. Machine Learning refers to methods that are capable of learning themselves from the data without having their decision model encoded by a human being.&lt;/p&gt;
&lt;p&gt;Take a program that can play tic-tac-toe. It clearly has an intelligence component in order to function, but the software will not learn from its interaction with the user (The problem is simple enough that there is no point to). But a program that &lt;a href=&quot;https://googleblog.blogspot.ca/2012/06/using-large-scale-brain-simulations-for.html&quot;&gt;recognizes cats in videos&lt;/a&gt; needs a machine-driven learning component is order for it to learn what a cat looks like.&lt;/p&gt;
&lt;h2&gt;
	Classifying And Finding Things With Artificial Intelligence&lt;/h2&gt;
&lt;p&gt;The figure below represents a very simplified block diagram of the AI process for classification; starting from the left to the right. We have a data set that we want processed, which can be documents, images, songs, video, etc... In practice, not everything within that data set is relevant to the context of what we are trying to do and so we transform each document into a set of features that we think are valuable to solve our problem. A feature might be a specific word in a document, another might be the word&#039;s part-of-speech (verb, noun, adverb, etc) or a typographical aspect such as the word being underlined.&lt;/p&gt;
&lt;p&gt;&lt;img alt=&quot;&quot; src=&quot;/sites/www.dbdump.org/images/AI2.png&quot; style=&quot;width: 554px; height: 160px; float: left;&quot; /&gt;&lt;/p&gt;
&lt;p&gt;Since the feature set &lt;em&gt;explicitly&lt;/em&gt; determines what part of the data set the algorithm will actually look at, feature generation is an extremely important part of the Artificial Intelligence process. It has spawned it&#039;s own field of study, &lt;a href=&quot;https://en.wikipedia.org/wiki/Feature_engineering&quot;&gt;Feature Engineering&lt;/a&gt;, and at times some have insisted that Artificial Intelligence is carefully crafted Feature Engineering. In practice, many engines have enough computational resources that they will simply generate every feature possible from the input data and the algorithm will simply choose the features that are most promising (This is called the &quot;Throwing things at the wall to see what sticks&quot; approach). It&#039;s wasteful, but computing time has become much cheaper than the people time required to create an efficient design.&lt;/p&gt;
&lt;p&gt;The algorithm is the brain of AI, which is ironic in that the algorithm in itself is usually very simple and generic; the same algorithm that flies a flying drone might be the same that keeps your phone camera images from being blurry. However, the devil is in the details and the implementation of the algorithm is usually not portable from the phone to the drone. Examples of algorithms might be &lt;a href=&quot;https://en.wikipedia.org/wiki/K-means_clustering&quot;&gt;k-means&lt;/a&gt;, &lt;a href=&quot;https://en.wikipedia.org/wiki/C4.5_algorithm&quot;&gt;C4.5&lt;/a&gt; or &lt;a href=&quot;https://en.wikipedia.org/wiki/Okapi_BM25&quot;&gt;okapi&lt;/a&gt;, each one trying to perform the task of &lt;a href=&quot;https://en.wikipedia.org/wiki/Cluster_analysis&quot;&gt;clustering&lt;/a&gt;, rule generation or &lt;a href=&quot;https://en.wikipedia.org/wiki/Information_retrieval&quot;&gt;information retrieval&lt;/a&gt;. As part of the process, the algorithm will take in the features and select the most promising. As part of that selection, some external information such as a trained model or parameter might be provided to the algorithm to guide it&#039;s decision making.&lt;/p&gt;
&lt;p&gt;The results are then checked against a benchmark, sometimes called a &lt;a href=&quot;https://en.wikipedia.org/wiki/Gold_standard_(test)&quot;&gt;gold standard&lt;/a&gt;, to ensure that the system is doing what it is supposed to. If the results aren&#039;t exactly what is required the model or the parameters of the algorithm might be changed.&lt;/p&gt;
&lt;p&gt;Overall, the basics of AI aren&#039;t that complex, but its implementation and arrangement needs to be focused on the objectives of the project, otherwise one gets into the loop of &quot;&lt;a href=&quot;https://en.wikipedia.org/wiki/Garbage_in,_garbage_out&quot;&gt;garbage in, garbage out&lt;/a&gt;&quot;. Depending on the case, the tuning of parameters can be frustrating and model generation becomes an art and not a science. There are many different frameworks, libraries and code bases available both freely and commercially to experiment with which I encourage you to do.&lt;/p&gt;
&lt;p&gt;Next: Part III - Algorithmic Bias is good for you.&lt;/p&gt;
&lt;!--break--&gt;&lt;hr /&gt;
&lt;h3&gt;References&lt;/h3&gt;
&lt;div class=&quot;references&quot;&gt;
&lt;ol&gt;
&lt;li id=&quot;reference1&quot;&gt;&lt;a name=&quot;ref1&quot; id=&quot;ref1&quot;&gt;&lt;span class=&quot;biblio-authors&quot;&gt;&lt;a href=&quot;/biblio?f%5Bauthor%5D=1&quot; rel=&quot;nofollow&quot;&gt;F.  Galton&lt;/a&gt;&lt;/span&gt;, &lt;/a&gt;&lt;a href=&quot;/node/33&quot;&gt;&lt;span class=&quot;biblio-title&quot; style=&quot;font-style: italic;&quot;&gt;Natural Inheritance&lt;/span&gt;&lt;/a&gt;. London, UK: Macmillan, 1889.&lt;span class=&quot;Z3988&quot; title=&quot;ctx_ver=Z39.88-2004&amp;amp;rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&amp;amp;rft.title=Natural+Inheritance&amp;amp;rft.btitle=Natural+Inheritance&amp;amp;rft.date=1889&amp;amp;rft.aulast=Galton&amp;amp;rft.aufirst=Francis&amp;amp;rft.pub=Macmillan&amp;amp;rft.place=London%2C+UK&quot;&gt;&lt;/span&gt; &lt;/li&gt;
&lt;/ol&gt;
&lt;/div&gt;
&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/41&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Artificial Intelligence&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/43&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Automation&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/3&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;History&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Fri, 31 Mar 2017 14:01:34 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">31 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/WhatisAI/PartII#comments</comments>
</item>
<item>
 <title>AI and the Law (Part 1)</title>
 <link>https://www.dbdump.org/WhatisAI/PartI</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;p&gt;&lt;em&gt;&lt;span&gt;(Note: This is the first part of a series of posts that were based on several conversations &lt;/span&gt;with lawyers and executives about &lt;a href=&quot;http://one.dbdump.org/taxonomy/term/10&quot;&gt;AI&lt;/a&gt;, the nature of technology and its application to business problems.)&lt;/em&gt;&lt;/p&gt;
&lt;h1&gt;
	What is Artificial Intelligence?&lt;/h1&gt;
&lt;p&gt;Does it really represent an improvement over what we already have? An entirely new class of solutions to ongoing problems? Or the flavour of the week in a market that is overwhelmed with buzzwords?&lt;/p&gt;
&lt;p&gt;Skepticism is endemic to technology culture in industry, government or academia. It&#039;s a byproduct of working in an area who&#039;s foundation is innovation and ideas. When it costs significantly less to say that you have something than to actually get it to work: a &quot;show me&quot; attitude is necessary. Ironically, IT has been so far primarily about what we would call classical Management Information System. The software may be really slick, the hardware may be really fast and we can store a lot of data, but really most of what the industry has focusing on so far is simply replacing physical forms and paperwork with the electronic equivalent: tabulating ledgers for accounting, generating reports and mailing checks and invoices over the Internet instead of in paper form. These are boring, unglamorous tasks but they have been IT&#039;s big success: taking things that were mundane, repetitive and that were cost centres and streamline it using technology.&lt;/p&gt;
&lt;p&gt;And now, we have Artificial Intelligence.&lt;/p&gt;
&lt;p&gt;&lt;a href=&quot;https://commons.wikimedia.org/wiki/File:Turk-engraving4.jpg&quot;&gt;&lt;img alt=&quot;Image of a copper engraving from Karl Gottlieb von Windisch&#039;s 1783 book Briefe über den Schachspieler des Hrn. von Kempelen, nebst drei Kupferstichen die diese berühmte Maschine vorstellen.&quot; src=&quot;/sites/one.dbdump.org/images/Turk-engraving4.jpeg &quot; style=&quot;width: 229px; height: 200px; float: right;&quot; /&gt;&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;The underlying idea that a machine would replace a human being in making decisions isn&#039;t all that new. One of the better know historical exhibits (and fraud) is &lt;a href=&quot;https://en.wikipedia.org/wiki/The_Turk&quot;&gt;The Turk&lt;/a&gt;, a mechanical automaton that would play chess against a person. Of course, playing chess was asking a lot of simple clockwork mechanisms and the builder had constructed a false compartment in which a human player would hide and move the mannequin using a system of pulleys and cams. This may have been the first &lt;a href=&quot;https://en.wikipedia.org/wiki/Vaporware&quot;&gt;vapourware&lt;/a&gt; product ever, but the idea that a machine could perform tasks at a human level had taken root.&lt;/p&gt;
&lt;p&gt;What we&#039;ll call modern Artificial Intelligence appeared in the mid-1950&#039;s when scientist began to look at ways that elements of human cognition could be modelled using mathematics. That in itself wasn&#039;t novel, humankind had moved on from the &lt;a href=&quot;https://en.wikipedia.org/wiki/Abacus&quot;&gt;abacus&lt;/a&gt;. What they were aiming at were higher cognition functions like learning from examples and extrapolating solutions for problems that the machine had never seen before.&lt;/p&gt;
&lt;p&gt;In this series of blog posts, the basics of AI will be reviewed and its application to practical business problems outlined. As with many technologies it has had its false starts and the causes of the &lt;a href=&quot;https://en.wikipedia.org/wiki/AI_winter&quot;&gt;AI Winters&lt;/a&gt; periods will be reviewed which in turn, will give a sense of why it is making such a resurgence.&lt;/p&gt;
&lt;p&gt;Next: Part II - How AI Works.&lt;/p&gt;
&lt;!--break--&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/41&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Artificial Intelligence&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/42&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Machine Learning&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/11&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Law&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Mon, 30 Jan 2017 15:34:23 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">30 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/WhatisAI/PartI#comments</comments>
</item>
<item>
 <title>Presentation at Derby University: Artificial Intelligence and the Law</title>
 <link>https://www.dbdump.org/node/4</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;div&gt;
	&lt;img alt=&quot;&quot; src=&quot;/sites/one.dbdump.org/derby.jpg&quot; style=&quot;width: 222px; height: 90px; float: right;&quot; /&gt;Artificial Intelligence and the Law&lt;/div&gt;
&lt;div&gt;
	&lt;a href=&quot;http://www.derby.ac.uk&quot;&gt;University of Derby&lt;/a&gt;&lt;/div&gt;
&lt;div&gt;
	&lt;a href=&quot;http://www.derby.ac.uk&quot;&gt;Venue: N613, &lt;/a&gt;&lt;a href=&quot;http://www.derby.ac.uk&quot;&gt;University of Derby&lt;/a&gt;, &lt;a href=&quot;https://www.google.com/maps/place/52%C2%B056&#039;18.2%22N+1%C2%B029&#039;47.7%22W/@52.938401,-1.4976369,209m/data=!3m2!1e3!4b1!4m5!3m4!1s0x0:0x0!8m2!3d52.938401!4d-1.496589&quot;&gt;Kedleston Road, Derby&lt;/a&gt;&lt;/div&gt;
&lt;div&gt;
	Wednesday, 30 November 2016 at 5.30 PM&lt;/div&gt;
&lt;p&gt; &lt;/p&gt;
&lt;div&gt;
	This talk is dedicated to the memory of David F. Evans.&lt;/div&gt;
&lt;p&gt; &lt;/p&gt;
&lt;div&gt;
	After years of over-promising and under-delivering, and two so-called “winters”, Artificial Intelligence is aggressively making inroads in the legal and financial markets. Affordable computational power, bandwidth, and storage each played a part in this revival, but adoption is really the result of targeted products that do what machine learning is historically excellent at: domain specific applications. Concerns over lost employment, reduced professional prestige, and the accountability of such approaches (Do algorithms behave ethically? Should they be regulated?) remain a hot topic for the layman and the practitioner alike as the technology finds new niches to occupy.&lt;/div&gt;
&lt;div&gt;
	This talk is about some lessons learned about the application of AI to legal document review and some of the unexpected corner-cases found along the way, including creative system training methods, tackling entrenched cultural beliefs, and the thorny (and multifaceted) issues of information security.&lt;/div&gt;
&lt;div&gt;
	Working on these problems has highlighted the value of human judgement, the limits of computation, and how scaling problems affect both people and machines. More than ever, sound theory, good mathematics, and a well-rounded approach are needed to tackle an increasingly complex and globalized world.&lt;/div&gt;
&lt;!--break--&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/10&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;AI&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/11&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;Law&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/12&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;HCI&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Mon, 21 Nov 2016 15:12:54 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">4 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/4#comments</comments>
</item>
<item>
 <title>Presentation to the Security Group, University of Kent</title>
 <link>https://www.dbdump.org/node/25</link>
 <description>&lt;div class=&quot;field field-name-body field-type-text-with-summary field-label-hidden&quot;&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; property=&quot;content:encoded&quot;&gt;&lt;div&gt;
	&lt;img alt=&quot;&quot; src=&quot;/sites/one.dbdump.org/globalHeaderLogo.png&quot; style=&quot;float: right;&quot; /&gt;Processing legal documents with AI: Notes from the field.&lt;/div&gt;
&lt;div&gt;
	&lt;a href=&quot;https://www.kent.ac.uk/&quot;&gt;University of Kent&lt;/a&gt;, &lt;a href=&quot;https://www.cs.kent.ac.uk/index.html&quot;&gt;School of Computing&lt;/a&gt;, &lt;a href=&quot;https://www.cs.kent.ac.uk/research/groups/security/&quot;&gt;Security Group&lt;/a&gt;&lt;/div&gt;
&lt;div&gt;
	Wednesday, 23 November 2016, 5PM&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	In a world where the marginal cost of copying information is nil, the marginal cost of storing information is nil, and the marginal cost of transmitting information is nil, the opportunities for data disasters are numerous and their consequences are devastating to individuals and corporations. Even well-intentioned gestures such as AOL&#039;s release of anonymized query logs have resulted in privacy violations in ways not immediately obvious to the principals.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	Interestingly, portability and interoperability are now working against us: in the world of single sign on, once you have access to the user you have access to all of their systems, and the first leak tends to be the last leak. This talk will be about the security of AI engines in contract analysis and the unexpected and counterintuitive lessons learned along the way. Security auditors want audit logs; forensics-aware systems people want traceability; privacy advocates want minimal information kept; developers want ease of maintenance and debugging; and users want to be minimally impacted. Now all of these views are necessarily compatible and some interesting conflicts arise in the creation and operations of applications.&lt;/div&gt;
&lt;div&gt;
	 &lt;/div&gt;
&lt;div&gt;
	In the end, the creation of online apps that handle sensitive information, preserve privacy and security, while enabling distributed teams to collaborate requires much more than following best practices. They rest on a culture of security within the organization.&lt;/div&gt;
&lt;!--break--&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;&lt;div class=&quot;form-item form-type-item&quot;&gt;
  &lt;label&gt;Language &lt;/label&gt;
 Undefined
&lt;/div&gt;
&lt;div class=&quot;field field-name-field-tags field-type-taxonomy-term-reference field-label-above&quot;&gt;&lt;div class=&quot;field-label&quot;&gt;Tags:&amp;nbsp;&lt;/div&gt;&lt;div class=&quot;field-items&quot;&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/31&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;security&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/32&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;communications&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item even&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/33&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;penn-testing&lt;/a&gt;&lt;/div&gt;&lt;div class=&quot;field-item odd&quot; rel=&quot;dc:subject&quot;&gt;&lt;a href=&quot;/taxonomy/term/34&quot; typeof=&quot;skos:Concept&quot; property=&quot;rdfs:label skos:prefLabel&quot; datatype=&quot;&quot;&gt;corporate memory&lt;/a&gt;&lt;/div&gt;&lt;/div&gt;&lt;/div&gt;</description>
 <pubDate>Tue, 15 Nov 2016 12:55:32 +0000</pubDate>
 <dc:creator>warren</dc:creator>
 <guid isPermaLink="false">25 at https://www.dbdump.org</guid>
 <comments>https://www.dbdump.org/node/25#comments</comments>
</item>
</channel>
</rss>
