Thursday, May 5, 2016

Fertilisation: IBM CSIG talk - 05 May - Scott MacLeod - “A Universal Translator as a Cognitive System, beginning as a Guidebook with Test” World University and School, Slides, Recording, WIKIDATA - "database of etymological relationships based on Wiktionary and an associated visualization" and WUaS's Universal Translator collaboration potentially, a) sent only to ... Hi Ester (and Denny), b) sent to the whole Wikidata email list ...

May 5, 2016

Dear Dianne, Jim and friends,

Thank you so much for the opportunity to present today in the IBM Cognitive Systems' Institute Group Speakers' Series, as well! And thanks for posting the recording “A Universal Translator as a Cognitive System, beginning as a Guidebook with Test” from today - - and the slides - - both accessible currently here - And thanks too for this CSIG LinkedIn Group information - - all for extending this Cognitive Systems' Institute Group conversation.  

Thank you!
05 MayScott MacLeod“A Universal Translator as a Cognitive System, beginning as a Guidebook with Test”World University and SchoolSlidesRecording

May 4, 2016

a) sent only to ... 

Hi Ester (and Denny), 

What a great and exciting "database of etymological relationships based on Wiktionary and an associated visualization" project. Thanks for reaching out to the Wikidata community. Our projects have so many potential articulations re Wiktionary especially. 

In what ways could WUaS contribute to helping you to build the Wikitionary dbnary database with RDFs also anticipating a universal translator in this project? WUaS hasn't hired or funded any projects whatsoever yet, but we plan to do both. (And WUaS is currently communicating with the Italian Cultural Center in SF and its related ISS school re hiring the very first director of Italian language World University and School - - where the ICC and ISS would fund this half-time WUaS position). 

In addition to working with all 7,943 languages (re my IBM CSIG Universal Translator talk tomorrow morning with slides here, WUaS wants our wiki-informed Universal Translator to work with so-called "dead" languages too, so the etymological history approach to the dbnary database in your project has many parallels and much value. CC Wikitionary's word/lexical-centricity, as well as potentially parts of words-centricity, as well as potentially phoneme-centricity (for voice) will be central to the planned WUaS Universal Translator, too. Wikimedia's Creative Commons' licensing is also central to Creative Commons' licensed WUaS. And WUaS's Universal Translator in its lexicality focus will eventually further WUaS and other multi-lingual online brain-language research - and via the Brain_and_Cognitive_Sciences wiki subject page - planned in many languages. 

WUaS would like also its universal translator to eventually be in a realistic virtual earth (i.e. film-realistic, interactive, 3D, wiki-build-able, and in all 7,943 languages) such as in Google Street View/Maps/Earth with its time slider with OpenSim/SecondLife, in which your etymological trees eventually could be seen, for example, in classrooms therein (e.g. in rooms at the virtual University of Chicago in this planned realistic virtual earth with time slider), but also in a variety of other ways. 

Could WUaS eventually support your project faculty-wise WUaS University (e..g. Italian language WUaS) by WUaS University (e.g. Serbian language WUaS) in combination with developing a universal translator? Could WUaS apply for/co-author a parallel project which would further support your interesting project and help develop the Universal Translator as well? Could WUaS/I become a co-author? WUaS is planning to facilitate each countries' main language creating their own major research university with free CC accrediting degrees (e.g. on How might you, Ester, see us collaborating further in generative ways? 

(Apologies for not speaking Italian).

Thank you,

From your proposal:
"A java extraction tool based on the Dbnary framework will be developed (see below) and a (RDF) database of etymological relationships will be populated using the English Wiktionary (note that the English Wiktionary contains etymologies of foreign words as well). Then the visualization extension etytree will be developed: a demo of etytree is available here.

Finally, an editing tool will be developed to guide editors in the process of modifying etymologies and therefore nodes and links of the tree."

On Wed, May 4, 2016 at 12:12 PM, Info WorldUniversity <> wrote:
Thanks, Ester,
More soon about this.
Best, Scott
On May 4, 2016 11:43 AM, "Ester Pantaleo" <> wrote:
Thanks Scott, your project is very valuable and interesting. Let me
know how you would like to collaborate.

May 4, 2016

b) sent to the whole Wikidata email list ...

Hi Ester, Denny and Wikidatans, 

CC World University and School is interested in collaborating, if possible, and is planning to develop both in CC Wikidata's ~300 languages, but eventually in all 7,943+ languages, and potentially with Wiktionary, and with developing a Universal Translator (and including an historical component - e.g time slider in Street View - which would work also with etymological history).

I'm glad to be giving an IBM Cognitive Systems' talk ( tomorrow, Thursday, May, at 7:30am PT/ 10:30am ET, as founder and president of World University and School, on "A Universal Translator as a Cognitive System, beginning as a Guidebook with Test" - to which you're invited. The WUaS Universal Translator will help translate WUaS's wiki pages with CC MIT OCW in 7 languages from English into all countries' official languages, and much, much more in all 7,943 languages. Here's a WUaS MediaWiki example from English to German - - having used the Google Translate machine translator. 

(WUaS is like Wikipedia with best STEM CC OpenCourseWare and planning to accredit on CC MIT OCW in 7 languages and CC Yale OYC for online free CC university degrees in all countries' main languages, as well as facilitate wiki schools for open teaching and learning in all 7,943 languages - and become a CC growth story for languages and the web). Best, Scott


May 3, 2016

Denny Vrandečić vrandecic@... via 

10:38 AM (22 hours ago)
to Discussion
I think that is a great idea. But Wikidata is not suited yet a lexical knowledge base, and I think that would be a necessary precondition for your project.

There is a project plan to make Wikidata suitable to be a lexical knowledge base:

And I sure hope that we will hear soon how this will move forward :)

[Wikidata] Wikidata and Wiktionary: an IEG proposal to build a database of etymological relationships based on Wiktionary and an associated visualization

May 3, 2016


I am writing to get some feedback on an IGE grant proposal I submitted to Wikimedia that might be of interest to the Wikidata community as it aims at building a database from Wiktionary data. 

More specifically the aim of the project is to develop an interactive visualization for etymological relationships using dbnary's extraction-framework (for Wiktionary) 

The data behind the visualization will consist of an RDF database of Wiktionary data (definition, part of speech, synonyms, etc) built using dbnary and a database of etymological relationships built using a custom code (to be integrated into dbnary) that translates Wiktionary textual etymology into a graph database of etymological relationships. 

A demo of my interactive visualization etytree is available here:

The visualization will present - in one graph - the etymology of all words deriving from the same ancestor. Users can expand/collapse the tree to visualize what they are interested in. The textual part attached to the graph can be easily translated in any language and the app would become a multilingual resource. 

I am writing to the Wikidata community because I would like to know if the Wikidata community thinks Wikidata could host this data. This project could help integrate dbnary into a Wikimedia environment and create a database from Wiktionary. In particular, the database of etymological relationships will be available for the community and can be used as a resource to study the history of languages, how pronunciation evolved through time, and eventually how semantics evolved through time.

The link to the grant proposal is 
Feedback is very welcome on the grant proposal page or on the talk page of the grant

Looking forward to read your comments.
Thanks a lot!

Ester Pantaleo



No comments: