面向金融领域的情感词典自动扩展方法的设计与实现
首发时间:2020-05-15
摘要:近几年微博得到了迅猛发展,微博使用人数不断攀升。庞大的用户量以及数据量的背后,蕴藏着巨大的商业、社会等多方面价值。对微博金融领域文本信息的情感研究分析,就是挖据微博潜在的商业价值。面向金融领域的情感词典是研究金融领域文本的重要基础,这方面的研究能应用于金融领域市场分析,用户群研究等金融领域的多个方面。本文以SO-PMI 算法作为基础,提出将标签传递算法(LPA)与PMI相结合,设计并且实现LPA-PMI算法,利用 Micro-F1 作为评判工具,分析研究算法在面向金融领域的情感词典扩展的有效性。实验对比表明,算法构造的金融词典相比于基础情感词典更适用于金融领域的情感分析;LPA-PMI算法相比于SO-PMI算法对于金融领域情感词典扩展的效果更好。验证了本论文提出的LPA-PMI算法的有效性。
关键词: 计算机技术 微博 情感词典 SO-PMI LPA-PMI
For information in English, please click here
Design and Implementation of Automatical Extension Method for
Abstract:In recent years, microblog has developed rapidly, and the number of microblog users is increasing. Behind the huge amount of users and data, there are huge commercial and social values. The emotional research and analysis of text information in the field of micro blog finance is to tap the potential business value of micro blog. Emotion dictionary for financial field is an important foundation to study financial field text. This research can be applied to many aspects of financial field, such as market analysis, user group research and so on. Based on so-pmi algorithm, this paper proposes the combination of LPA and PMI, designs and implements lpa-pmi algorithm, and uses micro-F1 as the evaluation tool to analyze the effectiveness of the algorithm in the expansion of emotion dictionary facing the financial field. The experimental results show that the financial dictionary constructed by the algorithm is more suitable for the emotional analysis in the financial field than the basic emotional dictionary, and the lpa-pmi algorithm is better than the so-pmi algorithm for the expansion of the emotional dictionary in the financial field. The validity of the lpa-pmi algorithm proposed in this paper is verified.
Keywords: Computer Technology Micro-Blog Sentiment Lexicon SO-PMI;LPA-PMI
基金:
引用
No.****
动态公开评议
共计0人参与
勘误表
面向金融领域的情感词典自动扩展方法的设计与实现
评论
全部评论0/1000