“A librarian of the web” is an interesting proposition that might, when fully implemented, help people find documents they are looking for in a more effective way. The first two chapters start by introducing library services, the key roles played by librarians, and the motivation for the new approach of transferring the librarians’ operations to a peer-to-peer (P2P) system.

Is it possible to replace the librarians or, better, equip them (along with everybody else) with new information retrieval tools designed for the ever-increasing volume of documents on the web? One answer is given in this research book.

Chapter 3 reviews web search engines and their working algorithms. Although continuously improved, these engines are still not meeting user expectations. For example, the amount of search results returned is generally too large and lacks structure. Relevant content may not be searched at all. For these and other reasons, the author states that search engines cannot compete with a human’s classification process that is largely based on semantic dependencies. The author proposes the concept of a distributed web search, and the chapter ends with a set of requirements to be met by the new librarian of a web system that will be distributed, scalable, and energy efficient.

In a natural way, the following chapter reviews existing strategies for natural language processing (NLP). While extracting keywords from text is a straightforward operation, returning the semantic relatedness among words and text is much more complex. The chapter discusses topics such as bag-of-concepts and clustering techniques, although a complexity analysis is missing.

The presentation continues with a set of new local tools and services developed and in use for web search. There are interesting comments and suggestions that anybody interested in work in this area can use. In particular, the Android infrared (IR) app seems to be attractive to busy people on the move.

The second part of the book starts with chapter 6, where a concept borrowed from physics, the centroid, represents text documents and queries and properly determines semantic similarities. The centroid is the basis for creating a treelike document structure that can be efficiently searched.

The buildup is made possible by two clustering algorithms. The system was tested to evaluate the quality of the search, and the preliminary results were comparable to those of Google. One benefit pointed out by the author is the distributed nature of the web search strategy he is proposing and the natural implementation as a P2P system.

Overall, due to short and well-contained chapters and graphics, the book is very clear. It comprehensively presents a topic of high interest and debates in a stimulating manner a new approach to web search. Summing up, this book is worth reading, and I can recommend it to all technical people interested in web engineering. It has significant value for graduates and researchers.