Interactive movie searching method

Abstract

The invention discloses an interactive movie searching method, which is characterized by comprising the following steps of: extracting full-text index data of a movie from a movie base according to information in a movie knowledge base; generating a movie fragment episode base by movie fragments in the movie base according to the information in the movie database and information in a movie characteristic expert base; and analyzing a user query condition, decomposing the query condition into combination query of a plurality of conditions, integrating the query result of the full-text index data with the information in the movie fragment episode base, carrying out optimizing and sequencing operations and returning the generated final query result to the user. The method further comprises the steps of automatically scoring a video, automatically updating the movie characteristic expert base and recognizing movie names. The interactive movie searching method can improve the accuracy of the movie search result.

Description

technical field [0001] The invention relates to the technical field of search engines, in particular to an interactive video search method. Background technique [0002] With the popularity and development of video network applications, many video websites have emerged, and many video search websites have also emerged. The realization of the search function of these video search websites is still based on the traditional search engine technology, that is, retrieval based on text information. [0003] Although the retrieval technology of text information is very mature, it cannot accurately identify the information of film and television (movies, TV). For example, when the user enters "Kung Fu" query, "Kung Fu King" and "Kung Fu Panda" will also be queried. , the accuracy of the query is greatly reduced. [0004] Due to video size limitations and other reasons, online video often divides the original video set into multiple small segments. When querying traditional search e...

AI Technical Summary

Problems solved by technology

[0003] Although the retrieval technology of text information is very mature, it cannot accurately identify the information of film and television (movies, TV). For example, when the user enters "Kung Fu" query, "Kung Fu King" and "Kung Fu Panda" will also be queried. , the accuracy of the query is greatly reduced

[0004] Due to video size limitations and other reasons, online video often divides the original video set into multiple small segments. When querying traditional search engine technology, the returned results are also in small segments. When users use Very inconvenient

Since the naming methods of small fragments end with Chinese characters, letters, and numbers, there will also be dashes, brackets, etc., in various forms, and it is difficult to list them exhaustively. Traditional search engines cannot organize multiple "broken" small fragments into Functions for Fragments in Sets

[0005] The user's usage of video reflects the degree of attention to the video, encoding quality, network quality, content quality, etc. Traditional search engines do not use these data to improve the quality of search results

Images