Anzeige
Erschienen in:
2019 | OriginalPaper | Buchkapitel
Chinese News Keyword Extraction Algorithm Based on TextRank and Topic Model
verfasst von : Ao Xiong, Qing Guo
Erschienen in: Artificial Intelligence for Communications and Networks
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
TextRank tends to choose frequent words as keywords of a document. In fact, some infrequent words can also be keywords. In order to improve this situation, a Chinese news keyword extraction algorithm LDA-TextRank based on TextRank and LDA topic model is proposed. The algorithm is a single document, unsupervised algorithm. It defines the diffusivity of two candidate words, constructs a new weight formula, and improves the weight of the edges in the text graph. At the same time, it combines with the LDA topic model, and the damping factor in TextRank is adjusted by calculating the word’s topic relevance of the document. The experiment was carried out on the Chinese corpus. The results show that compared with TextRank, LDA-TextRank has an improvement in Precision, Recall and F1-measure.