Rewriting Keyword Information Using Search Engine Results

摘要

A computer-implemented technique is described herein for modifying original keyword information to increase the probability that it will match the queries input by users. The technique operates by using a search engine to provide supplemental information that is relevant to the original keyword information. The technique then mines the supplemental information to extract frequently-occurring n-grams. Next, the technique removes n-grams that are considered to represent noise, and then uses a deep-structured machine-learned model to assign score values to the remaining n-grams. Finally, the technique supplements and/or replaces the original keyword information with the highest-scoring n-grams.

主权项

1. A method for modifying keyword information, implemented by at least one hardware processor of one or more computing devices, comprising:
identifying a target item having original keyword information that warrants modification; submitting the original keyword information to a computer-implemented search engine; receiving supplemental information from the search engine that has been determined, by the search engine, to be related to the keyword information; producing a collection of n-grams based on tokens which appear in the supplemental information; selecting n-grams in the collection of n-grams that satisfy a frequency threshold test, to provide a subset of frequently-occurring n-grams; filtering out n-grams from the subset of frequently-occurring n-grams that are determined to represent noise, to provide a subset of noise-removed candidate n-grams; using a scoring model to assign a score value to each candidate n-gram, the score value reflecting a similarity between the candidate n-gram and the original keyword information, to overall provide score information associated with a subset of scored n-grams; selecting one or more scored n-grams based on the score information, to provide selected keyword information; replacing and/or supplementing the original keyword information with the selected keyword information, to provide new keyword information; and storing the new keyword information in association with the target item.