System and methods for predicting correct spellings of terms in multiple-term search queries
View Patent ↗A search engine process predicts the correct spellings of search terms within multiple-term search queries. In one embodiment, when a user submits a multiple-term search query that includes a non-matching term and at least one matching term, a table is accessed to look up a set of terms that are “related” to the matching term or terms. A spelling comparison function is then used to determine whether any of these related terms is sufficiently similar in spelling to the non-matching term to be deemed a candidate correctly-spelled replacement. A candidate replacement term may automatically be substituted for the non-matching term, or may be suggested to the user as a replacement. The invention also includes a process for identifying terms that are related to each other based on the relatively high frequencies with which they co-occur within search queries of users, database records, and/or specific database fields.
1. A method of predicting a correct spelling of a potentially-misspelled search term within a multiple-term search query, the method comprising:
identifying at least one matching search term of the multiple-term search query, said matching search term being different from the potentially-misspelled search term;
identifying at least one additional term that, based on an analysis of search query submissions of users, has occurred in combination with the matching search term relatively frequently; and
comparing a spelling of the additional term to a spelling of the potentially-misspelled search term.
2. The method of claim 1 , wherein identifying at least one additional term comprises taking into consideration a search field in which the potentially-misspelled search term was entered.
3. The method of claim 2 , wherein identifying at least one additional term further comprises taking into consideration a search field in which the matching search term was entered.
4. The method of claim 1 , further comprising automatically replacing the potentially-misspelled term with the additional term if the additional is similar in spelling to the potentially-misspelled term to within a defined threshold.
5. The method of claim 1 , further comprising performing said analysis of search query submissions.
6. The method of claim 5 , wherein performing the analysis comprises according a greater weight to a search query submission that, based on an action performed by a user with respect to a search result item, is deemed to have produced a successful result.
7. The method of claim 6 , wherein said action is a purchase of a search result item.
8. A method of correcting spelling errors in search queries, the method comprising:
receiving a multiple-term search query from a user;
identifying a non-matching search term within the multiple-term term search query;
identifying a matching search term within the multiple-term term search query; and
selecting a replacement term that is a candidate correctly-spelled replacement for the non-matching search term, wherein the candidate correctly-spelled replacement term is selected based at least in part upon an identity of the matching search term.
9. The method as is claim 8 , wherein selecting the candidate correctly-spelled replacement term comprises selecting a term that, based on an analysis of search query submissions of users, has appeared in combination with the matching search term relatively frequently.
10. The method as is claim 8 , wherein selecting the candidate correctly-spelled replacement term comprises selecting a replacement term that, based on an analysis of a database to which the multiple-term search query is directed, has appeared in combination with the matching search term within the database relatively frequently.
11. A method of predicting a correct spelling of a non-matching search term within a multiple-term search query, the method comprising:
identifying a plurality of matching search terms within the multiple-term search query, each matching search term being different from the non-matching search term;
for each matching search term, looking up a corresponding set of related terms, to thereby obtain multiple sets of related terms;
combining the multiple sets of related terms to form a related terms set; and
comparing spellings of individual terms in the related terms set to a spelling of the non-matching term.
12. The method of claim 11 , wherein looking up a corresponding set of related terms comprises identifying related terms that, based on an analysis of search query submissions of users, has occurred in combination with a matching search term relatively frequently.