Nice Stemmer Project

This is an extension of the Nice Stemmer Project at University of North Carolina.

Project Members: Adam Wead, Kiduk Yang

Based on the previous research in the TREC project, this study will focus on the effect of different stemming algorithms on retrieval precision with respect to term weights and corpus data types

Data	Stemmer	Term Weight
Robust	Simple	SMART
HARD	Porter	Okapi
Genomics	Krovetz
Web	Combo
24 Result Sets

Plan

Literature Review
- stemmer
- term weights
Research Question
- What is the effect of stopping and stemming with respect to term weights and data
Update Nice Stemmer
Run Experiement
Data Analysis
Conclusion
- Effect: good/bad/none
Paper Submission

Reference:

Hull, D. A. (1996). Stemming algorithms: A case study for detailed evaluation. Journal of the American Society for Information Science, 47, 70-84.