Short text characteristic extension and fitting characteristic library building method and device

Abstract

The invention discloses a short text characteristic extension and fitting characteristic library building method and device. The short text characteristic extension method includes the steps of a, obtaining a short text to be extended, b, deleting an item, without meaning expression ability, in the short text to be extended to obtain a fitting short text, c, inquiring the fitting short text in a fitting characteristic library, returning a characteristic item of the fitting short text as an extension characteristic item if the fitting short text is found in the fitting characteristic library, and executing the step d if the fitting short text is not found in the fitting characteristic library, d, omitting an item with the lowest significance weight in the fitting short text to obtain an omitting short text, e, judging whether a sum of the significance weights of all items in the omitting short text is smaller than a threshold value or not, returning ineffectiveness if the sum of the significance weights of all items in the omitting short text is smaller than the threshold value, and executing the step f if the sum of the significance weights of all items in the omitting short text is larger than the threshold value, and f, inquiring the omitting short text in the fitting characteristic library, returning a characteristic item of the omitting short text as an extension characteristic item if the omitting short text is found, using the omitting short text as the fitting short text if the omitting short text is not found, and executing the step d. By means of the short text characteristic extension and fitting characteristic library building method and device, the transferred meaning risks in the short text characteristic extension process are reduced, and the characteristic extension accuracy is improved.

Description

technical field [0001] The invention relates to computer text processing technology, in particular to a short text feature expansion and fitting text feature database construction method and device. Background technique [0002] With the widespread use of applications such as e-mail, web forums, and microblogs, a large amount of text information data has been generated within the Internet, and these information are usually only fragmentary descriptions or opinion comments, with only a short text content, so called short text. Facing the massive text data generated by the rapid development of the Internet, how to accurately and effectively obtain the required materials and information has become a topic of widespread concern and research in the Internet industry. [0003] Due to the short length of the short text and the weak signal of the concept described, the retrieval results may not be obtained in the short text retrieval, or the obtained retrieval results are not what ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the embodiments of the present invention is to propose a short text feature extension and fitting feature library construction method and device to solve the problems of high escape risk and inaccurate extension results in the process of short text feature extension

Images