Introduction to probabilistic models.

Description of boolean retrieval, vector space model, probabilistic retrieval, latent semantic indexing and other IR topics. An introduction to various classical ranking methods is also provided.

It describes key issues in document ranking techniques based on the vector­ space model. Several TF*IDF variants are discussed. The cosine measure, recall and precision are introduced. [PS format]

A survey of probabilistic models in information retrieval. [PDF format]

Formal introduction to latent semantic indexing. [PS format]

A Chapter in a book which introduces probabilistic retrieval.

"Ranking Algorithms" is chapter 14 in the Frakes and Baeza-Yates book. It gives a good discussion of the tradeoffs and choices among different term-weighting strategies.

Evaluation of many combinations of term frequency statistics, document frequency statistics and document length normalization. [PDF format]

This study evaluates the performance of a state-of-the-art keyword-based document ranking algorithm (coming out of TREC) on a popular web search task.