Query Model for Image Search based on User Clicks and NN Features Dmitry Krivokon
Yandex School of Data Analysis Conference Machine Learning: Prospects and Applications We consider a problem of improving the quality of a querybased image search engine by using user click data. The primary purpose of an image search engine (SE) is to assist in finding images that are relevant to the text query entered by a user (such SEs should not be confused with content base image retrieval 10). The resulting images ideally should be sorted by their relevanc
|
|