Information retrieval system
First Claim
1. An information retrieval system comprising:
- an input arranged to receive user query terms;
a processor arranged to generate a plurality of lists of documents using the query terms and by accessing a document index;
a prediction engine arranged to predict, for each of the lists of documents, implicit relevance information and explicit relevance information;
a utility engine arranged to calculate a utility score for each of the lists of documents using the predicted relevance information and a stored utility function;
the utility engine also being arranged to select and store one of the lists of documents on the basis of the utility scores.
2 Assignments
0 Petitions
Accused Products
Abstract
An information retrieval system is described for retrieving a list of documents such as web pages or other items from a document index in response to a user query. In an embodiment a prediction engine is used to predict both explicit relevance information such as judgment labels and implicit relevance information such as click data. In an embodiment the predicted relevance information is applied to a stored utility function that describes user satisfaction with a search session. This produces utility scores for proposed lists of documents. Using the utility scores one of the lists of documents is selected. In this way different sources of relevance information are combined into a single information retrieval system in a principled and effective manner which gives improved performance.
66 Citations
20 Claims
-
1. An information retrieval system comprising:
-
an input arranged to receive user query terms; a processor arranged to generate a plurality of lists of documents using the query terms and by accessing a document index; a prediction engine arranged to predict, for each of the lists of documents, implicit relevance information and explicit relevance information; a utility engine arranged to calculate a utility score for each of the lists of documents using the predicted relevance information and a stored utility function; the utility engine also being arranged to select and store one of the lists of documents on the basis of the utility scores. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13)
-
-
14. A computer-implemented method of retrieving a ranked list of documents from an index of documents comprising
receiving at least one query term; -
generating a plurality of lists of documents from the index using the query term; predicting, for each of the lists of documents, implicit relevance information and explicit relevance information; calculating a utility score for each of the lists of documents using the predicted relevance information and a stored utility function; selecting and storing one of the lists of documents on the basis of the utility scores. - View Dependent Claims (15, 16, 17)
-
-
18. One or more device-readable media with device executable instructions for performing steps comprising:
-
receiving at least one query term being part of a request to retrieve a ranked list of documents from a document index; generating a plurality of ranked lists of documents from the index using the query term; using a machine learning system comprising a stored probabilistic model to predict, for each of the ranked lists of documents, implicit relevance information and explicit relevance information; calculating a utility score for each of the ranked lists of documents using both the predicted implicit relevance information and the predicted explicit relevance information as well as a stored utility function; selecting and storing one of the ranked lists of documents on the basis of the utility scores. - View Dependent Claims (19, 20)
-
Specification