Real-Time Search has recently emerged as a new and exciting trend in search. It places far greater emphasis on freshness and on minimizing the lag between acquisition of content and its appearance in the index. Real-Time Search completes this process on the order of (fractions of) minutes, possibly even seconds.
A very significant portion of search engine queries (40%) are made by users looking for the latest and freshest information. In addition, users expect these freshest results to be of the highest quality. Real-Time Search greatly improves the user experience for those queries seeking the freshest, highest-quality results. As previously explained, traditional index architectures have been optimized in almost the exact opposite way, with large lag times for the appearance of fresh links in results in order to amortize the very high cost of construction of the entire
index.
Current real-time search architectures, for instance, Twitter Search, place emphasis on freshness and inclusion of all results matching a given keyword. Such an approach sounds appealing, even though the resulting stream of matching results is very large, changes (too) quickly, and may not be as interesting.
Two of the key aspects of the search problem are coverage of the Web content and the lag between discovery of content and its availability in the results. Contemporary general search engines such as Google, Bing and Yahoo have been focused on coverage, trying to maximize the scope of the Web coverage by increasing the size of the index. On the other hand, the newly emerging players in the field of real-time search have been focused on reducing the lag in displaying the results to users.
The 2×2 below depicts the Real-Time Search problem in a typical consulting way. Do you see an opportunity for a new company to emerge and merge what’s good about Google and Twitter?
Tags: Google, real-time search, search, Twitter