Personalized search results
The disclosure relates to personalizing search results based on the device features of a user device. An example method for personalizing search results includes receiving an encoded device profile indicating the device features of the device. The device features indicate one or more native applications installed on the device. The device features may further indicate native applications recently executed by the device, native applications that have been deleted from the user device, and/or web applications recently accessed by the user device. The method further includes identifying the device features of the device based on the encoded device profile, receiving a search query from the device, and performing an application search based on the search query to identify a consideration set of records. The method further includes adjusting the consideration set based on the device features and generating search results based on the adjusted consideration set.
1. A method comprising:
receiving, by a processing device, an encoded device profile indicating device features of a user device, wherein the device features indicate one or more native applications installed on the user device;
identifying, by the processing device, the device features of the user device based on the encoded device profile;
receiving, by the processing device, a search query from the user device;
performing, by the processing device, an application search based on the search query to identify a consideration set of records, wherein each record in the consideration set of records corresponds to a respective native application; and
for each record in the consideration set of records:
generating, by the processing device, a result score based on at least one of an application scoring feature, a query scoring feature, or an application-query scoring feature, wherein:
the application scoring feature includes data corresponding to a popularity of each record the consideration set of records,
the query scoring feature includes at least one of a number of words in the search query, a popularity of the search query, or an expected frequency of words in the search query, and
the application-query scoring feature includes parameters that indicate a degree of matching between words of the search query and words of each record in the consideration set of records,
determining, by the processing device, a boosting factor based on the device features of the user device, wherein the boosting factor is a first value in response to the respective native application not being installed on the user device, and wherein the boosting factor is a second value in response to the respective native application being installed on the user device, and
adjusting, by the processing device, the result score based on the boosting factor,
generating, by the processing device, search results based on a subset of the consideration set of records, wherein the subset is selected based on the result scores of the consideration set of records, and
transmitting, by the processing device, the search results to the user device.