Samet's book is the definitive reference, but I have trouble understanding his explanations. They do a good job of summarizing what's distinctive about each technique relative to the others, but a poor job of explaining what the techniques are. This is a pitfall of technical writing, especially by experts: it's easy to produce something that's intelligible only if you already know the material, which defeats the purpose (assuming your purpose is explanatory and not only compository). Still, the book is indispensable as a reference.
When I was building OBsearch.net I used this book extensively. I grant you that some explanations are not so detailed. On the other hand, the explanations of the most relevant techniques are quite good. Check for instance the section that explains LSH or the explanation about Navarro's SAT.
