Identification of intent and non-intent query portions
Methods and systems for improved categorization of queries are disclosed. In one aspect, first queries having query results limited to a particular category are identified. Second queries including a first query are also identified. For identified second queries, a pattern is generated based on the second query and the included first query, and a rank of the first query's particular category in results for the second query and a percentage of the second query's results having a category equivalent to the included first query's particular category are determined. The ranks and percentages are aggregated for matching patterns, and second patterns with aggregated ranks and percentages meeting a criterion are determined. Results for a third query are limited to a category equivalent to a particular category for a first query included in the third query, and then transmitted over a computer network to a client device.
1. A method of query processing at an application server using query categorization, the method comprising:
receiving, at the application server, a search query comprising at least a first text portion and a second text portion;
determining the search query matches a pattern from a pattern database, the pattern having a fixed text portion and an uncategorized variable placeholder portion;
in response to determining the search query matches the pattern, using the pattern to identify the first text portion in the search query as a sub-query based on the second text portion in the search query corresponding to the fixed text portion of the pattern and the first text portion in the search query corresponding to the uncategorized variable placeholder portion of the pattern;
based on identifying the first text portion in the search query as the sub-query, categorizing the search query by assigning, to the search query, a category of the sub-query for performing a search with the search query using the category of the sub-query; and
generating search results for the search query using the category of the sub-query.
2. The method of claim 1 , wherein the pattern is identified and stored in the pattern database by:
for each of a plurality of historical queries that are not categorized:
identifying a portion of the historical query as corresponding to a categorized query from a plurality of categorized queries;
generating the pattern by replacing the portion of the historical query corresponding to the categorized query with the uncategorized variable placeholder portion;
comparing a category of search results for the historical query with a category of the categorized query; and
selecting the pattern for storage in the pattern database based on comparison of the category of search results for each historical query with the category of the categorized query within each historical query.
3. The method of claim 2 , wherein comparing the category of search results for a first historical query with the category of the categorized query within the first historical query comprises:
determining a rank of the category of the categorized query in the search results for the first historical query.
4. The method of claim 2 , wherein selecting the pattern for storage in the pattern database based on comparison of the category of search results for each historical query with the category of the categorized query within each historical query comprises:
determining a percentage of the historical search queries in which the category of search results for the historical search queries matches the category of the categorized queries within the historical search queries.
5. The method of claim 2 , wherein the categorized queries are identified based on a database mapping each categorized query to a corresponding category.
6. The method of claim 2 , wherein the placeholder comprises a predetermined string.
7. The method of claim 1 , wherein generating the search results for the search query comprises:
determining an attribute value based on another portion of the search query; and
identifying the search results using the attribute value and the category of the sub-query.
8. The method of claim 7 , wherein determining the attribute value comprises:
determining an ancillary token based on the search query and the sub-query included in the search query;
determining one or more attributes associated with the category of the sub-query;
determining attribute values for the one or more attributes;
determining a token value based on the ancillary token; and
assigning the token value to the attribute value in response to a comparison of the token value to attribute values for the one or more attributes.
9. One or more machine-readable media storing computer-useable instruction that, when used by one or more processor an application server, cause the one or more processors to perform operations, the operations comprising:
receiving, at the application server, a search query comprising at least a first text portion and a second text portion;
determining the search query matches a pattern from a pattern database, the pattern having a fixed text portion and an uncategorized variable placeholder portion;
in response to determining the search query matches the pattern, using the pattern to identify the first text portion in the search query as a sub-query based on the second text portion in the search query corresponding to the fixed text portion of the pattern and the first text portion in the search query corresponding to the uncategorized variable placeholder portion of the pattern;
based on identifying the first text portion in the search query as the sub-query, categorizing the search query by assigning, to the search query, a category of the sub-query for performing a search with the search query using the category of the sub-query; and
generating search results for the search query using the category of the sub-query.
10. The one or more machine-readable media of claim 9 , wherein the pattern is identified and stored in the pattern database by:
for each of a plurality of historical queries that are not categorized:
identifying a portion of the historical query as corresponding to a categorized query from a plurality of categorized queries;
generating the pattern by replacing the portion of the historical query corresponding to the categorized query with the uncategorized variable placeholder portion;
comparing a category of search results for the historical query with a category of the categorized query; and
selecting the pattern for storage in the pattern database based on comparison of the category of search results for each historical query with the category of the categorized query within each historical query.
11. The one or more machine-readable media of claim 10 , wherein comparing the category of search results for a first historical query with the category of the categorized query within the first historical query comprises:
determining a rank of the category of the categorized query in the search results for the first historical query.
12. The one or more machine-readable media of claim 10 , wherein selecting the pattern for storage in the pattern database based on comparison of the category of search results for each historical query with the category of the categorized query within each historical query comprises:
determining a percentage of the historical search queries in which the category of search results for the historical search queries matches the category of the categorized queries within the historical search queries.
13. The one or more machine-readable media of claim 10 , wherein the categorized queries are identified based on a database mapping each categorized query to a corresponding category.
14. The one or more machine-readable media of claim 10 , wherein the placeholder comprises a predetermined string.
15. The one or more machine-readable media of claim 9 , wherein generating the search results for the search query comprises:
determining an attribute value based on another portion of the search query; and
identifying the search results using the attribute value and the category of the sub-query.
16. The one or more machine-readable media of claim 15 , wherein determining the attribute value comprises:
determining an ancillary token based on the search query and the sub-query included in the search query;
determining one or more attributes associated with the category of the sub-query;
determining attribute values for the one or more attributes;
determining a token value based on the ancillary token; and
assigning the token value to the attribute value in response to a comparison of the token value to attribute values for the one or more attributes.
17. An application server comprising:
one or more processors; and
one or more computer readable storage media storing computer-useable instructions that, when used by the one or more processors, cause the one or more processors to:
receive, at the application server, a search query comprising at least a first text portion and a second text portion;
determine the search query matches a pattern from a pattern database, the pattern having a fixed text portion and an uncategorized variable placeholder portion;
in response to determining the search query matches the pattern, use the pattern to identify the first text portion in the search query as a sub-query based on the second text portion in the search query corresponding to the fixed text portion of the pattern and the first text portion in the search query corresponding to the uncategorized variable placeholder portion of the pattern;
based on identifying the first text portion in the search query as the sub-query, categorize the search query by assigning, to the search query, a category of the sub-query for performing a search with the search query using the category of the sub-query; and
generate search results for the search query using the category of the sub-query.
18. The application server of claim 17 , wherein the pattern is identified and stored in the pattern database by:
for each of a plurality of historical queries that are not categorized:
identifying a portion of the historical query as corresponding to a categorized query from a plurality of categorized queries;
generating the pattern by replacing the portion of the historical query corresponding to the categorized query with the uncategorized variable placeholder portion;
comparing a category of search results for the historical query with a category of the categorized query; and
selecting the pattern for storage in the pattern database based on comparison of the category of search results for each historical query with the category of the categorized query within each historical query.
19. The application server of claim 17 , wherein the search results for the search query are generated by:
determining an attribute value based on another portion of the search query; and
identifying the search results using the attribute value and the category of the sub-query.
20. The application server of claim 19 , wherein determining the attribute value comprises:
determining an ancillary token based on the search query and the sub-query included in the search query;
determining one or more attributes associated with the category of the sub-query;
determining attribute values for the one or more attributes;
determining a token value based on the ancillary token; and
assigning the token value to the attribute value in response to a comparison of the token value to attribute values for the one or more attributes.