- From: carl mattocks <carlmattocks@gmail.com>
- Date: Wed, 11 Jul 2018 13:25:10 -0400
- To: Chris Harding <chris@lacibus.net>
- Cc: Paola Di Maio <paoladimaio10@googlemail.com>, public-aikr@w3.org
- Message-ID: <CAHtonukcP0tEoO2LFK1W_PZ542VhJ=njtWU8j316szTpyo9XFw@mail.gmail.com>
Greetings: I am Carl Mattocks, for 30 + years I have taken a pragmatic approach in developing AI & KR capabilities. Thus I easily accepted Paola's suggestion to name the group AIKR. I suggested the ChatBot centric areas because I believe to have any chance of proven success, AIKR must be applied as in, have a national language / discipline specific context / domains / dimensions / vectors etc. I support the idea that (1) the vocabulary should include KR terms as well as AI terms; (2) knowledge is best stewarded as a property of a community-of-practice; (3) AIKR should be published in a compendium (4) which is likely to have some kind of blockchain & AIKR miners. As reference material I start with links to UIMA work .. so we can take advantage of the NLP oriented capabilities. cheers Carl Mattocks http://www.linkedin.com/in/carlmattocks carlmattocks@gmail.com Reference: A number of NLP research institutes based their software development efforts on UIMA specifications to adhere to emerging standards. http://www.lrec-conf.org/proceedings/lrec2008/workshops/W16_Proceedings.pdf UIMA, the Unstructured Information Management Architecture, is an open-platform middle-ware for dealing with unstructured information (text, speech, audio, video data), originally launched by IBM. Apache Software Foundation has a project for developing UIMA-based software https://uima.apache.org/d/uimaj-2.6.0/tutorials_and_users_guides.html. Organization for the Advancement of Structured Information Standards (OASIS) hosted a Technical Committee to standardize the UIMA specification. https://www.oasis-open.org/committees/download.php/27025/UIMA_Specification_Draft_v1.doc A number of NLP research institutes based their software development efforts on UIMA specifications to adhere to emerging standards. http://www.lrec-conf.org/proceedings/lrec2008/workshops/W16_Proceedings.pdf It was a pleasure to clarify On Wed, Jul 11, 2018 at 5:45 AM, Chris Harding <chris@lacibus.net> wrote: > It was the "Knowledge Representation" aspect that attracted me to join the > group. > > AI appears to be about creating computer systems that are like people. > AIKR is partly about the internal design of such computer systems, but it > is also about about communication between them and about communication > between them and people. The communication aspect is to my mind more > important than internal design. > > My ideas on this are not based on research, but are derived from many > years of experience of participating in and managing groups of people > developing standards - at least some of which incorporate knowledge. They > are not fully formed, or well expressed, and I hope that through this group > I will be able to connect them with academic and commercial research and > contribute to a group conensus. > > The list of proposed outcomes at https://www.w3.org/community/aikr/ looks > good, and I suggest starting with the list of resources and the natural > language vocabulary. I propose that the vocabulary should include KR terms > as well as AI terms. (This may have been intended, but is not explicit.) > > > Paola Di Maio wrote: > > Thank you everyone who introduced themselves, and others please do so in > your own time. > > What motivated the formation of this group initially is a need to keep up > with the extremely fast moving pace of AI progress - even when writing a > simple research paper or teaching a basic class like 'Introduction to > Machine Learning' these days it is necessary to keep an eye out for new > approaches and contexts. There is a lot happening. Things are changing by > the minute. > I don't understand the words people use anymore. Everyday I need to start > a new learning curve. What were once fairly clear cut fields of study seem > to have shifted (or is it my consciosness expanding?) There seems to be new > concepts coming up all the time. Cognitive ability is linked to knowledge > representation, both for humans and machines, and I thought maybe we can > try to grab some of this rapidly developing AI knowledge by creating some > KR tabs to help us figure things out. Even KR is moving but at a less fast > pace. > > So I started digesting words and concepts and hope that the cognitive load > is going to become more bearable. Then realised there is much to be done in > this space, and since the web is now part of us in more than one way, our > interface, our constantly updated database, W3C is open to new ideas, why > not make a CG. > > I am so glad to have so many clever and experienced people on the list. > Its easy to turn sun open informal efforts into a rave party, get > distracted with interesting discussions, since we all have our own agendas > and interesting backgrounds and a lot to tell. > > Our challenge is to keep the focus, so I beg everyone please lets put down > explicitly (write in our web pages) some stuff we want doing (goals, > deliverables whatever) and let's stick to it diligently and systematically. > > Please share with the group - on separate threads perhaps - what you think > we should try to achieve, and how? > What motivates you to join this group? What would like to contribute, and > what can the group do for you? > > I started a few pages with initial thoughts, they should editable > https://www.w3.org/community/aikr/welcome/ > > On a housekeeping note, we may benefit from a co-chair if anyone would > like to volunteer > (no particular commitment required) I hope this group will become a WG as > soon as we have enough juice, and produce a lot of nice deliverables, and > after achieving some of the goals I have initially set (maye publish one or > two reports, or a journal special issue or something) then I ll be ready to > pass on the torch to the next chair. > > Since it's the summer, I suggest we may relax and ponder on the tasks > already set out and contribute some thinking and some items (start > submitting resources, edit the TO DO list a bit, etc) and postpone our > first virtual meeting until September. but if someone/anyone is inclined > to call out a virtual meeting before the end of the summer please feel free > to do so. > > I am not big with online meetings, prefer to exchange via emails, but if > virtual meetings become productive I ll make the effort. > > Personally I am now busy reading new research - tons and tons of stuff - > and simply > understanding what it all means. I ll try to pin down a few concept/terms > definition and submit them to our resource depository and will ping the > list when I need some input on what I am doing > For now these pages are still drafts, but when everyone has had the > opportunity to review > them, we can make the firm and start actioning them, set dates etc > > Please do the same! > > Look forward to learn more about what everybody is doing and how can this > community > help! Questions? > > Thanks again for joining > > PDM > > > > > > > Dr Paola Di Maio > Center For Technology Ethics > ISTCS.org > *A bit about me <https://about.me/paoladimaio>* > > > On Tue, Jul 10, 2018 at 5:04 PM, Alberto Nogales Moyano < > alberto.nogales@ceiec.es> wrote: > >> Dear all >> thanks for creating this group. >> My name is Alberto Nogales, I am a computer scientist and have a PhD from >> University of Alcalá (Spain) related with Social Network Analysis of the >> Web of Linked Data in order to improve data retrieval strategies. During >> the last years I have also worked in Technical University in Vienna, >> Technical University in Tallinn and University of Málaga. Actually, I work >> as a researcher and lecturer in Universidad Francisco de Vitoria (also in >> Spain). The head of my lab has a huge background in neural networks, so we >> have started to apply Deep Learning models to some issues like automatic >> classification of Linked Data sources or data retrieval strategies by >> applying learning to rank. More information about my lab here: >> http://www.ceiec.es/en/ >> >> >> Alberto Nogales Moyano >> PostDoctoral researcher and Lecturer >> CEIEC (Francisco de Vitoria University) >> https://www.researchgate.net/profile/Alberto_Nogales >> alberto.nogales@ceiec.es >> >> >> >> On 06 Jul 2018, at 12:34, Paola Di Maio <paola.dimaio@gmail.com> wrote: >> >> seminal article >> <https://onlinelibrary.wiley.com/doi/epdf/10.1207/s15516709cog0101_2> >> >> >> > > -- > Regards > > Chris > ++++ > > Chief Executive, Lacibus <https://lacibus.net> Ltd > chris@lacibus.net > >
Received on Wednesday, 11 July 2018 17:26:15 UTC