To respond to the criticism, you are right that there is room for much more on the subject. To some degree I was constrained by the format–these lectures are supposed to be short. Regardless, I do look forward to diving into more detail either in a follow-up edition or some other format. My role at Endeca does come with an NDA, but there is much I am able (and eager) to discuss freely.

Daniel Tunkelang’s book is a concise introduction to the area of faceted search, its foundations in information science, current research, and practical concerns. The book starts with a whirlwind tour of information science from Aristotle’s classification to Google’s PageRank algorithm. The first part ends with a very short section on faceted search. At this point, faceted search seems so basic and obvious that you wonder why it has only been widely accepted in the last decade. The answer comes in the next chapter, which explains why faceted search is computationally demanding. The last part of the book mentions a number of practical issues in connection with organizing the data, efficient processing and the design of an effective user interface.

The book is rather short (79 pages, including glossary, author’s biography, references, 23 screenshots and 7 blank pages), but it does cover the main points and is very clearly written. It is a pleasant and easy read, it took me less than 3 hours to read from cover to cover (your mileage may vary). It is surprising that Tunkelang, with his extensive experience in this area, does not have more to say about the subject. Maybe his position as Chief Scientist of Endeca does not allow him to go into more detail and give away more of the methodology how to actually build a successful faceted search solution.

The book is great as a first introduction to the field, but I would be looking forward to an expanded second edition.