Sashank Kurra
← projects

Search / NLP · CS 410 / information retrieval

Information Retrieval

Search, indexing and ranking experiments

Python · Pyserini · Indexing · Ranking · Text retrieval

Built information-retrieval experiments with Pyserini to understand how search systems turn a document collection into ranked results. The work covered indexing, querying, ranking, and evaluating whether the returned documents were actually relevant.

01Document corpus
02Index
03Query processing
04Retrieval model
05Ranked results
06Evaluation

Built retrieval pipelines with Pyserini.

Tested how indexing and ranking choices affect search results.

Kept document preparation, retrieval, and result evaluation as separate stages.

A search system can return results correctly while still ranking poor matches highly, so relevance evaluation matters beyond simply making the query run.

Built hands-on experience with the core pieces of a text search and ranking pipeline.

← All projectsHome →