Information Retrieval - CS371R

University of Texas at Austin

About: Various aspects of a retrieval system, implemented for class credits

1. Vector Space Retrieval (vsr)

Perform retrieval based on the cosine similarities of documents
Invoke with: java ir.vsr.InvertedIndex -html [path to document corpus]

2. Relevance Feedback

Reassess retrievals with relevance feedback and pseudofeedback
Invoke with: java ir.vsr.InvertedIndex -html -pseudofeedback 8 [path to document corpus]

3. PageRank

Implemented PageRank to spider a corpus of documents and perform link analysis.
Invoke with: java ir.vsr.PageRankInvertedIndex -html -weight 0 crawledPages

4. Text Categorization

Implemented Rocchio and KNN algorithms to categorize a set of sample documents.
Invoke with: java ir.classifiers.TestKNN [-K K] java ir.classifiers.TestRocchio [-neg]