Release Notes

These release notes summarize new enhanced features and fixed bugs for the most recent releases of Text Categorization Tools ^Java.

Version 2008

The 2008 release is the 2nd version of the Java Text Categorization Tools. This version includes 4 main script programs. They are:

This release includes completing 45 software change requests (SCRs). They are described as follows.

I. Main Feature enhancements

Release Package
- Distributed with JRE, 1.6.0_5
- Distributed with HSqlDb 1.8.0.7 (HyperSonic SQL DB)
- Provide sample Java codes for using Text Categorization tools APIs
- Provide configurable option to auto locate the top directory of TC
- Provide version information on ${TC_DIR}/data/versions.txt
TC Data
- New design of WORD_ST table in DB
- Released with latest tables from MEDLINE.2008 and Metathesaurus.2007AC
- Released with latest tables for JDI/STI/STRI
- Used new designed algorithm to generate stDocuments for STI and STRI
- Latest update on stopWords list
- Latest update on default value of Max. cutoff for normalized count
Java APIs
- New APIs for calculate similarity of JD vectors
- New APIs for calculate similarity of ST vectors
Tool Options
- New features of choosing score display factor to display score in JDI/STI/STRI
- New features of choosing output fields filter in JDI/STI/STRI
- New features of displaying detail scores for all legal words in JDI/STI/STRI
- New features of displaying Semantic Types in TUIs, ST names, and ST abbreviations in STI/STRI
- New features of providing Acronym filter option in JDI/STI/STRI
- New features of indexing MH/SH in STRI
- New features of preserving PMID in MLT
- New features of sorting results in PMID in MLT
- New features of taking unique identifier in input for JDI/STI/STRI

II. Bug fixes: