Microblog sorting model building and microblog diversity retrieval methods

Abstract

The invention discloses a microblog sorting model building method. The method comprises the steps of S1) constructing a training data set, wherein the training data set comprises a series of query words, each query word corresponds to a plurality of microblogs, and an arrangement sequence of the microblogs is obtained in a manual annotation manner to serve as a training standard answer; S2) extracting attributes of the microblogs corresponding to each query word in the training data set; S3) extracting a relevance feature and a similarity feature of each blog article; and S4) building and training a sorting model. Based on the model, the invention furthermore provides a microblog diversity retrieval method. By use of the method, diversified retrieval results are returned when a user retrieves related information in the microblogs, so that information redundancy is reduced, the accuracy and coverage of a retrieval result of a retrieval system can be effectively improved, and the user experience is enhanced.

Description

technical field [0001] The invention relates to the technical field of information retrieval, in particular to the field of social media retrieval, and in particular to the establishment of a microblog sorting model and a microblog diversity retrieval method. Background technique [0002] Microblog retrieval belongs to the field of information retrieval and is an important means to extract effective information from massive microblog data. In microblog retrieval, users generally express their query intentions by entering shorter query words (average 1.64 words), which often lead to ambiguity or uncertainty in user query intentions. For example, when a user enters the query word "apple", the retrieval system needs to determine whether the user's query intention needs information related to the Apple company or information related to fruit such as apples. [0003] However, the current microblog retrieval system cannot accurately understand the user's query intention. In this ...

AI Technical Summary

Problems solved by technology

[0004] As a kind of social media, Weibo has short text and irregular grammar, which makes the effect of traditional natural language processing methods seriously decline in Weibo retrieval.

Images