An application search method and server

Abstract

The invention discloses an application search method and server, comprising: constructing an application label database, the application label database includes a label system of each application; building a search word label database, and the search word label database includes a plurality of search word label systems; Build a tag knowledge base; receive the current search word uploaded by the client, and obtain the tag system of the current search word according to the search word tag database; based on the tag knowledge base, calculate the semantic correlation between the tag system of the current search word and the tag system of each application When the semantic correlation between the label system of the current search word and the label system of an application meets the preset condition, the relevant information of the application is returned to the client for display. This solution matches the tag system of the application with the tag system of the search term based on the tag knowledge base, that is, maps the user intent and the application into the same semantic space, solves the semantic matching problem between the search term and the application, and effectively realizes the application search function.

Description

Technical field[0001]The invention relates to the field of Internet technology, in particular to an application search method and server.Background technique[0002]Application search engine is a mobile terminal software application search engine service that provides mobile app search and download, such as 360 mobile assistant, Tencent app treasure, GooglePlay, Appstore, etc. With the development of the mobile Internet, the number of mobile app applications has increased rapidly. Take the 360 ​​mobile assistant as an example, there are millions of app applications. When users don’t know the exact name of the app they need, they only use what they want. Construct a query description. The application search engine analyzes the user's query intention based on the search query, maps the user's intention to the app application, and accurately finds the app application. This is the application scenario of the functional search technology.[0003]When we analyzed the user query session, we fo...

AI Technical Summary

Problems solved by technology

[0004] The traditional functional search technology solution is based on manual labeling and keyword matching retrieval technology. Manual labeling and classification of app types and manual sorting of query intent types have high accuracy, but low recall and coverage rates, which can only cover popular apps and applications. Its type, the cost of manual labeling to obtain classification and labeling is very high, the workload is heavy, time-consuming and labor-intensive

[0005] Traditional keyword matching-based retrieval algorithms, when associated tags are used to express related semantics, cannot recall related apps due to word mismatch. The general solution is to mine synonyms and use synonyms to supplement semantic matching. However, in the vertical field of apps, synonyms are Very sparse. We have tried to mine synonyms but the results are very poor. One of the characteristics of the vertical search field is that the query requirements are narrow and the ambiguity is small

For example, when a user searches for a query "find a restaurant", it is semantically related to app tags such as "group buying, gourmet food, takeaway, discount, o2o, foodie", but there is no synonym coverage. There is a semantic gap between the user query and the app, which cannot satisfy the user. Search requirements, poor user search experience

Images