ECE328 Information Retrieval

The contents of the course are the following:

• Boolean model, dictionary and postings lists, tolerant retrieval, index construction, index compression, scoring and term weighting,
• vector space retrieval, recall and precision, relevance feedback and query expansion, probabilistic information retrieval
• latent semantic indexing, sparse matrix storage, compressed row storage, compressed column storage, low-rank approximations,
• Web search basics, Web crawling and indexes
  summation formula for PageRank, problems with the iterative process, Markov chain theory, spectrum of the Google matrix,
• parameters in the PageRank model, hyperlink matrix, teleportation matrix
  sensitivity of PageRank, the PageRank problem as a linear system, proof of the PageRank sparse linear system
• large-scale implementation of PageRank, back button modeling, adaptive power method, extrapolation, aggregation, updating the PageRank vector
• HITS method for ranking Webpages, HITS implementation, HITS convergence, HITS’s relationship to bibliometrics, query-independent HITS, HITS sensitivity
• SALSA
• Content and link spam

The course comprises a detailed description of the area of modern information retrieval in the World Wide Web, presenting the topics of content-based ranking and link analysis ranking.
The goal of the course is to offer the knowledge of structures and methods required to develop and execute information retrieval operations in modern networked environments.

The students majoring in the course will:

• Be able to understand the difference between data retrieval and information retrieval.
• Familiarize themselves with the architecture of an information retrieval system, i.e., a search engine.
• Understand the properties of binary, vector and probabilistic information retrieval model.
• Understand the most popular indexing methods of information retrieval systems.
• Acquire the ability to evaluate information retrieval systems.
• Familiarize themselves with the relevance feedback and query expansion techniques.
• Understand the particularities of information retrieval in the World Wide Web.
• Familiarize themselves with Web crawling.
• Understand the concept of link analysis ranking using spectral centralities.