Year: 2024
Type: conference-paper
Abstract: The traditional classification method based on manual labelling has problems such as low efficiency and high cost. To address this problem, this project proposes to combine the three feature extraction techniques of TF-IDF, Word2Vec and GloVe, and use K-means clustering algorithm to cluster and analyse the patent text. The experiment proves that compared with the conventional TF-IDF algorithm, the algorithm has a greater improvement in clustering effect, classification accuracy and retrieval effect, and can better reflect the semantics in patent text. On this basis, this project will propose a new type of patent text classification method based on image segmentation, and compare it with the conventional image segmentation technology, so that it has better classification effect and higher computational efficiency.
Authors: Xiaoxin Gao, Yukai Liu, Yuangeng Zhu
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4409832941