In the area of web search, there is a disconnect between some of the research in the field and what is actually done within a large-scale commercial search engine. Everything from measuring the wrong things to the objectives of the research, have the potential to produce work that although academically interesting, has little commercial value. This paper attempts to reduce this gap by summarizing my experience as both a researcher and an engineer at commercial search engines.

This paper focuses on the theoretical components of a commercial search engine and then describes several of the real-world issues and challenges faced by commercial search engines. Next, it explores briefly the issue of “relevance” and some challenges specific to studying and improving relevance in a commercial search engine.