The newest version of the Database Recommender automatically analyzes a user’s result list to provide recommendations to databases that are relevant to those results. The previous version providing recommendations for databases at your institution was based only on a static list of keyword triggers. This new technology, in addition to the keyword triggers, will base recommendations also on analysis of the search result set. This additional intelligence will allow databases to be recommended for conceptually related search queries. Note that the Database Recommender is able to recommend both databases that are indexed natively by the Summon™ service as well as those that are not.
For example, if you conduct a search for 802.11, the IEEE standard for wireless networking, recommendations may include IEEE Electronic Library Online and Compendex as shown below, depending on your individual library’s subscriptions. The algorithm powering the Database Recommender is being continuously monitored, evaluated and tuned on an ongoing basis.