A feature-centric view of information retrieval. (Q639421)

This book is organized in 6 chapters (Introduction, Classical retrieval models, Feature-based ranking, Feature-based query expansion, Query-dependent feature weighting, Model learning), two appendices (Data sets and Evaluation metrics) and a comprehensive bibliography. Chapter 1, as the name suggests, is a short introduction into the topics of this book, serving also as a strong motivation for the reader. Chapter 2 presents some classical retrieval models making a transition to the 3rd chapter which exposes the Markov random field model, the fundaments for the feature-based retrieval models studied in this book. Chapter 4 presents `latent concept expansion', a method used for feature-based query expansion, while Chapter 5 presents weighted sequential dependence model and its adaptive techniques for evaluating the importance of features based on properties. The last chapter presents various techniques for estimating the parameters of feature-based models. The book is recommended for an advanced master's or PhD-level course in information retrieval, being also a valuable reference for the researchers with professional interests in this domain.