Content-based distributed feature extraction method and device, equipment and medium

Abstract

The invention discloses a content-based distributed feature extraction method and device, equipment and a medium. The method comprises the steps of obtaining multiple different dimensions of content information contained in to-be-processed data, wherein each piece of content information corresponds to one feature dimension; preprocessing each of multiple pieces of content information to obtain a plurality of original feature vectors in one-to-one correspondence with each piece of content information; and calling a distributed feature extraction model to perform feature extraction on the plurality of original feature vectors to obtain distributed features corresponding to the to-be-processed data, wherein the distributed features are results of characterizing the to-be-processed data according to a plurality of feature dimensions defined by a standard feature dimension template, and the standard feature dimension template defines the range of content extraction of the to-be-processed data of different sources and different types. According to the embodiment of the invention, the multiple pieces of content information are mapped to a same vector space by utilizing the distributed feature extraction model, so that an accurate recommendation effect is obtained.

Description

technical field [0001] The present application generally relates to the technical field of big data, and in particular relates to content-based distributed feature extraction methods, devices, equipment and media. Background technique [0002] With the development of electronic devices, more and more people choose to read news information on electronic devices. A personalized news recommendation system based on artificial intelligence usually uses machine learning algorithms, especially neural networks, to extract features from news content. For example, representation extraction based on collaborative information, or representation extraction based on content information, or representation extraction based on a combination of the former two. The former relies on the user's interaction information (such as click, favorite, etc.), which relies on the user's interactive operation. The latter relies on the information of the news content (including title, author, text, body, ...

AI Technical Summary

Problems solved by technology

[0003] However, now the forms of news information are diversified and the content information is very rich.

Using the existing representation extraction methods, it is impossible to perform unified feature extraction for different sources and different types of information data, which leads to the inability of the recommendation system to accurately understand different sources and types of information data

Images