Use of statistical language modeling for generating exploratory search results

摘要

Methods and arrangements for generating exploratory search results. A first search query is accepted in a browsing session, and first search result set is derived, comprising at least one search result from the first search query. A second search query is predicted, and a second search result set is derived, comprising at least one search result from the second search query. A final search result set is yielded via combining the first search result set and the second search result set.

主权项

1. A computer program product comprising:
a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising: computer readable program code configured to accept a first search query in a browsing session; computer readable program code configured to derive a first search result set comprising at least one search result from the first search query; computer readable program code configured to predict a second search query; computer readable program code configured to derive a second search result set comprising at least one search result from the second search query; computer readable program code configured to yield a final search result set via combining the first search result set and the second search result set; wherein said computer readable program code is configured to derive a first search result set via ascertaining a quantity k of at least one search result first occurring responsive to the first accepted search query; wherein said computer readable program code is configured to determine k based on a comparison of a first derivative of a cumulative confidence score of retrieved ranked documents to a predetermined threshold; the first derivative being determined as a function of the cumulative confidence score compared to a number of retrieved ranked documents; and wherein the retrieved ranked documents comprise documents retrieved from the first search result set and ranked on a basis of confidence scores from compared N-grams from within the documents comprising at least one of: compared bigrams, and compared trigrams.