Open AccessResearch on Topic Mining Algorithm Based on Deep Learning Extension
Author(s) -
Fuqiang Yang,
Xuan Zhao,
Maohong Zhang
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1345/4/042034
Subject(s) - event (particle physics) , computer science , extension (predicate logic) , similarity (geometry) , set (abstract data type) , algorithm , data mining , artificial intelligence , feature selection , theme (computing) , feature (linguistics) , pattern recognition (psychology) , machine learning , image (mathematics) , linguistics , philosophy , physics , quantum mechanics , programming language , operating system
This paper proposes a new algorithm that can consider both keywords and time of occurrence. Firstly, after the data is preprocessed, the LDA model of the theme event set is established, and the set of subject words is generated as the description mark of the event. The semantic and temporal similarity between the event keywords are calculated by the DTW algorithm to obtain the corresponding similarity matrix. Finally, the collaborative training method is used to iteratively generate the final feature vector and complete the event selection. The simulation results show that the proposed algorithm has higher accuracy and higher efficiency than previous algorithms.