Relevance estimation using a search satisfaction metric

摘要

The subject disclosure is directed towards using a satisfaction model's prediction as to whether a user was satisfied or dissatisfied in satisfying a search goal to help estimate the relevance of a URL/document that was returned and clicked by the user. The clickthrough data for a search goal session is processed by either a utility model or a despair model based on whether the satisfaction model indicated that the search goal session ended with the user satisfied or dissatisfied, respectively. The utility model distributes a utility value to each clicked URL, while the despair model distributes a despair value to each clicked URL. The utility value and despair value of each query-URL pair may be used as corresponding feature data for learning a search ranker.

主权项

1. In a computing environment, a method performed at least in part on at least one processor, comprising:
processing search goal session data obtained from a search log, the search goal session comprising a plurality of queries, the search goal session data including information that indicates any uniform resource locator (URL) that was clicked after being returned in search results corresponding to at least one query; inputting satisfaction results for a query-URL pair, the satisfaction results comprising a value indicative of a user satisfied or dissatisfied with satisfying an information need corresponding to the search goal session; and if the satisfaction results indicate that a query within the plurality of queries ended with the user dissatisfied, associating some amount of despair with each clicked query-URL pair for the search goal session, including selecting one or more URLs as effectively clicked and associating some amount of despair with each effectively clicked URL.