Keyword extraction method for machine reading comprehension based on natural language processing

• Published in: Journal of physics. Conference series 2021-06, Vol.1955 (1), p.12072
• Main Authors: Li, Ruiheng, Zhang, Xuan, Li, Chengdong, Zheng, Zhongju, Zhou, Zihang, Geng, Yuyin
• Format: Article
• Language: English
• Subjects: Algorithms; Annotations; Information retrieval; Keywords; Natural language processing; Physics; Preprocessing; Reading comprehension; Websites
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1955/1/012072

The traditional keyword extraction method has the problem that the accuracy of identification and extraction decreases when the amount of data is increasing. This paper introduces natural language processing technology and designs a keyword extraction method for machine reading comprehension. In this process, considering that the information in the text of machine reading comprehension may come from the web page, we can refer to the NLP system of natural language algorithm. Firstly, analyze the content of a single document, then analyze the short string level, and finally organize the preprocessing of natural language text information, so as to achieve the preprocessing of machine reading and comprehension of Chinese text. At the same time, the SKEA keyword extraction algorithm system is designed to merge text words with the same meaning, locate the location information of document resources, and identify the annotated words in the resource base. After the annotation processing is completed, the annotation information is extracted. In addition, a comparative experiment is presented to prove the effectiveness of the proposed method compared with the traditional method.