Older software and demos

Here is a list of software and demos that I have written, contributed to, released and/or possibly maintain. (Since most of the people involved in these have moved, not all the links may be current.)

NLP Software from Illinois

  • Edison: Edison is a Java library for representing different NLP annotations (views) over text in the form of graphs over constituents. It provides easy-to-use accessors for different types of views and facilitates feature extraction. Here is a presentation I gave which introduces the API.

  • Cogcomp core utilities: This is a collections of utility classes that I found myself writing again and again for several NLP and Machine Learning related stuff. The Java standard library should really have this functionality (the Pair class, for example.) This is not strictly an NLP library, but a collection of Java classes that are useful for NLP applications.

  • Explicit Semantic Analysis: Quang Do and I wrote our own version of Explicit Semantic Analysis.

  • Curator: The curator manages several NLP components and provides uniform access to them. It also caches annotations. The Curator and Edison are closely linked to each other and are described in this LREC 2012 paper.

  • Semantic Role Labeling: The Java version of the Illinois Semantic Role Labeler. The SRL system is based on the work of Punyakanok, Roth and Yih, (2008) and is a complete rewrite of their system, with additional bells and whistles. You can get it by downloading the Curator.

Online Demos at Illinois

  • Semantic Role Labeling: A demo of the SRL system. This also includes the preposition relations that are described in my thesis.

  • Dataless Classification: A demo of our AAAI 2008 paper. It takes text and two arbitrary textual labels and decides which label better describes the text. The demo uses our home-grown implementation of Explicit Semantic Analysis.

Machine learning code

Code that I either use or have somehow helped during development or both.