Kurt’s personal pet project involves conceptual clustering and decision theory. He is constructing mechanisms to extract deep semantic meaning from text, indexing data at the concept level, performing unsupervised classification of semantic content, and providing access to end users through a web-enabled interface. So far this personal research includes the complimentary use of a robust text parser (Link Parser), semantic grounding (extended from WordNet) and conceptual classification (Labrinth a’la Douglass Fisher, et al) with statistical/neural support (extended from DISCERN).

He placed a number of things under MetaNeta, which was a concept developed back in Oakton.

MetaNeta Services

Mar 1994

Project description

MetaNeta is a development project, providing a collection of semantic utility services that are used by related projects.

* Lexi provides an ultra fast response to word searches and lexical identifier lookup in HTML (for humans) or JSON (to support internal reasoning).
– – XML, ProtoBufs and other formats may be added on request.

* Semidx (under construction) is an unsupervised learning system that indexes the semantic content, updating Lexi along the way. Semidx is used by FembOS.

* Parse (under construction) is a natural language parser that uses Semidx and Lexi for semantic grounding and word sense disambiguation. The output is fed to Semidx.

* Repo is a database and publishing system that provides a knowledge repository populated with all data constructs to run an enterprise. Repo is used by CompanyQuick.
See project MetaNeta Services (The site is down as someone else would have to take up the research now).

Welcome to Lexi

Lexi (a.k.a. lexicon_wn) is a node.js application to serve wordnet semantics and morph-syntatic relationships.

The code forces an ‘isa’ relationship with the hypernym “hierarchy,” with selective inheritance, providing local access to deep semantics without having to recurse.

API calls default (anything not .html or .text) to Content-Type: ‘application/json’.
By adding a ‘.text’ suffix, the same JSON is returned with Content-Type: text/plain.
By adding a ‘.html’ suffix, the data is rendered as HTML with clickable hyperlinks.

/surface/

Returns a list of words that have a particular surface (is spelled like) e.g., ‘happiness’

/surface/happiness – returns a list of ‘happiness’ as application/json
/surface/happiness.text – returns a list of ‘happiness’ as text/plain
/surface/happiness.html – renders a list of clickable ‘happiness’ as text/html

/word/

Returns a deep semantic of a particular word as part of a synset.

/word/113987423/1 – returns semantics of happiness (synonym of felicity) as application/json
/word/113987423/1.text – returns semantics of happiness (synonym of felicity) as text/plain
/word/113987423/1.html – returns semantics of happiness (synonym of felicity) as text.html

/synset/

Provides a deep semantic of a synset, and references to incorporated words.

/synset/113987423 – returns semantics of the synset containing happiness and felicity as application/json
/synset/113987423.text – returns semantics of the synset containing happiness and felicity as text/plain
/synset/113987423.html – returns semantics of the synset containing happiness and felicity as text.html

WordNet

Princeton University “About WordNet.” [WordNet](https://wordnet.princeton.edu). Princeton University. 2010.

See also: [morphosemantic-links.xls](http://wordnetcode.princeton.edu/standoff-files/morphosemantic-links.xls)

************

3 D printing

Lexie

The She Bot from Woody Allen Movie

PLaying bass and some spanish guitar

Russian

Indian food and getting weird ingredients *maybe other page