A scientific literature clustering method and system based on unsupervised keyword extraction
By constructing an unsupervised keyword extraction model, combining the characteristics of Chinese and English, and automatically adjusting the keyword length and performing differentiated embedding methods, the problem of low efficiency in keyword recognition and clustering in existing literature management systems is solved, achieving efficient clustering and accurate organization of Chinese and English scientific literature.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANDONG UNIV
- Filing Date
- 2023-09-18
- Publication Date
- 2026-07-03
AI Technical Summary
Existing document management systems suffer from inefficiency and inaccuracy in keyword recognition and document clustering, especially in Chinese scientific literature. Existing unsupervised keyword extraction algorithms ignore background knowledge and domain characteristics, resulting in poor clustering performance.
We construct a clustering model based on unsupervised keyword extraction. By comprehensively considering the occurrence of words in document abstracts and titles, the semantic similarity between words and documents, and the characteristics of domain keywords, we use different embedding methods to cluster Chinese and English documents, extract keywords rich in domain concepts, and perform effective clustering.
It improves the accuracy and efficiency of literature clustering, can automatically adjust the keyword length according to the characteristics of the field, realizes effective clustering of Chinese and English scientific literature, and enhances the ability to organize literature and grasp research trends.
Smart Images
Figure CN117453912B_ABST