Dependency-based query expansion alteration candidate scoring
View Patent ↗An alteration candidate for a query can be scored. The scoring may include computing one or more query-dependent feature scores and/or one or more intra-candidate dependent feature scores. The computation of the query-dependent feature score(s) can be based on dependencies to multiple query terms from each of one or more alteration terms (i.e., for each of the one or more alteration terms, there can be dependencies to multiple query terms that form at least a portion of the basis for the query-dependent feature score(s)). The computation of the intra-candidate dependent feature score(s) can be based on dependencies between different terms in the alteration candidate. A candidate score can be computed using the query dependent feature score(s) and/or the intra-candidate dependent feature score(s). Additionally, the candidate score can be used in determining whether to select the candidate to expand the query. If selected, the candidate can be used to expand the query.
1. A computer-implemented method, comprising:
scoring an alteration candidate for a query, the alteration candidate comprising multiple alteration terms, the query comprising multiple query terms, and the scoring comprising:
computing one or more query-dependent feature scores that are based on dependencies to multiple query terms from each of one or more of the alteration terms, the one or more query-dependent feature scores comprising one or more bigram scores that are based on dependencies between a pair of the alteration terms and multiple terms in the query; and
computing a candidate score for the candidate using the one or more query-dependent feature scores; and
determining whether to select the candidate to expand the query, the determination using the candidate score.
2. The method of claim 1 , wherein at least one of the one or more query-dependent feature scores is based on a dependency between at least one of the alteration terms and the query overall.
3. The method of claim 1 , wherein the query-dependent feature scores comprise one or more term dependency scores based on dependencies between an alteration term and a plurality of query terms.
4. The method of claim 1 , wherein at least a portion of the one or more query-dependent feature scores are based on one or more translation models.
5. The method of claim 1 , wherein scoring the alteration candidate further comprises computing one or more intra-candidate-dependent feature scores that are based on dependencies between different terms in the alteration candidate, and wherein the computation of the candidate score uses the one or more intra-candidate-dependent feature scores.
6. The method of claim 5 , wherein the one or more intra-candidate dependent feature scores are based on dependencies between all the terms in the alteration candidate.
7. The method of claim 5 , wherein the intra-candidate-dependent feature scores comprise one or more adjacent bigram scores and one or more skip bigram scores.
8. A computer system comprising:
at least one processor; and
a memory comprising instructions stored thereon that when executed by the at least one processor cause the at least one processor to perform acts comprising:
scoring an alteration candidate for a query, the alteration candidate comprising multiple alteration terms, the query comprising multiple query terms, and the scoring comprising:
computing one or more intra-candidate dependent feature scores that are based on dependencies between different terms in the alteration candidate;
computing one or more query-dependent feature scores that are based on dependencies to multiple query terms from each of one or more of the alteration terms; and
using the one or more intra-candidate dependent feature scores and the one or more query-dependent feature scores to compute a candidate score for the candidate; and
determining whether to select the candidate to expand the query, the determination using the candidate score.
9. The computer system of claim 8 , wherein the intra-candidate dependent feature scores comprise at least one skip bigram score.
10. The computer system of claim 8 , wherein the one or more intra-candidate dependent feature scores are based on dependencies between all the terms in the alteration candidate.
11. The computer system of claim 8 , wherein the one or more query-dependant feature scores comprises a plurality of feature scores based on a plurality of translation models.
12. The computer system of claim 8 , wherein the one or more query-dependant feature scores is based on one or more dependencies between at least one of the alteration terms and the query overall.
13. One or more computer-readable storage media having computer-executable instructions embodied thereon that, when executed by at least one processor, cause the at least one processor to perform acts comprising:
scoring a plurality of alteration candidates for a query, the query comprising multiple query terms and each of the alteration candidates comprising multiple alteration terms, the scoring comprising, for each of the alteration candidates, performing the following:
computing one or more intra-candidate dependent feature scores based on dependencies between each term of the candidate and each other term of the candidate, the computation of the one or more intra-candidate dependent feature scores being based on one or more word count models;
computing one or more query-dependent feature scores based on dependencies between each alteration term and the query overall, the computation of the one or more query-dependent feature scores being based on one or more translation models; and
computing a candidate score for the candidate using the feature scores;
determining for each of the candidates whether to select the candidate to expand the query, the determination for each candidate using a corresponding candidate score; and
for each candidate selected to expand the query, using the selected candidate to expand the query.
14. The one or more computer-readable storage media of claim 13 , wherein the one or more query-dependent feature scores comprise:
one or more term dependency scores representing dependencies between each of one or more terms of the candidate and one or more terms of the query; and
one or more bigram translation scores representing dependencies between one or more pairs of terms of the candidate and one or more terms of the query.
15. The one or more computer-readable storage media of claim 13 , wherein the intra-candidate dependent feature scores comprise:
one or more bigram scores representing dependencies between one or more pairs of terms of the candidate; and
one or more unigram feature scores based on based on the alteration terms.
16. The one or more computer-readable storage media of claim 13 , wherein the candidate score computation for each candidate further comprises computing a length feature score based on a length of the candidate.
17. The one or more computer-readable storage media of claim 13 , wherein the candidate score computation for each candidate further comprises computing a translation model score from the query to itself.
18. The one or more computer-readable storage media of claim 13 , wherein the candidate score computation for each candidate comprises computing one or more unigram feature scores based on the alteration terms.