Two-level text similarity calculation method based on subjective and objective semantics
A technology of semantic similarity and text similarity, applied in computing, special data processing applications, instruments, etc., can solve the problems that the accuracy or rationality needs to be improved, and the text expression dimension is high, and achieve the improvement of accuracy and rationality, saving Storage space, the effect of concise expression of words
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment Construction
[0051] For the convenience of description, we assume the following application example: Randomly extract several different categories of texts from Sogou news texts, and calculate the similarity between these texts.
[0052] The specific embodiment of the present invention is:
[0053] (1) Download a corpus containing tens of thousands of Sogou news from the Internet, build a text corpus, segment words and extract keywords, and build a text index;
[0054] (2) Divide each text to be calculated for similarity into two parts: topic information and text content information;
[0055] (3) Process the topic information of the text as a sentence, divide the topic sentence into words, and filter out adverbs, prepositions, pronouns, conjunctions, and words that are substrings of other words to obtain sentence-word vectors, which combine subjective and objective semantics of words The similarity calculation method is applied to the word vector similarity calculation to obtain the text ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com