/ranked-search

Search system for the Reuters 21578 Corpus.

Primary LanguageC#MIT LicenseMIT

Ranked Search

Build status

Search system for the Reuters 21578 Corpus

It uses an inverted index to store the documents and each document represented as a bag of words model where each word is reduced to its stem before inferring the probability distribution.
The relevance score of a document is computed by summing over the query terms' tf-idf weights.

Clone

git clone --recursive https://github.com/deniskyashif/ranked-search.git

Dependencies

  • .NET 8 LTS