《面向金融领域的文本情感分析技术研究》由会员分享,可在线阅读,更多相关《面向金融领域的文本情感分析技术研究(83页珍藏版)》请在金锄头文库上搜索。
1、面向金融领域的文本情感分析技术研究硕士学位论文 面向金融领域的文本情感分析技术研究 THE STUDY ON SENTIMENT ANALYSIS FOR FINANCIAL TEXT 王亚伟 哈尔滨工业大学 2011 年 12 月 国内图书分类号 TP3913 学校代码 10213 国际图书分类号 6213 密级公开 硕士学位论文 面向金融领域的文本情感分析技术研究 硕士研究生王亚伟 导 师王晓龙教授 申 请 学 位工学硕士 学 科计算机科学与技术 所 在 单 位深圳研究生院 答 辩 日 期 2011 年 12 月 授予学位单位哈尔滨工业大学 Classified Index TP3913
2、com 6213 Thesis for the Master Degree in Engineering THE STUDY ON SENTIMENT ANALYSIS FOR FINANCIAL TEXT Candidate Yawei Wang Supervisor ProfXiaolong Wang Academic Degree Applied for Master of Engineering Speciality Computer ScienceTechnology Affiliation Shenzhen Graduate School Date of Defence Decem
3、ber 2011 Degree-Conferring-Institution Harbin Institute of Technology 哈尔滨工业大学工学硕士学位论文 摘 要 随着互联网技术和金融产业的飞速发展网络上金融信息日益丰富同时 人们对金融信息的需求也日益增多金融文本的倾向性可能对相关个股或者公 司股票价格走势产生重要影响为此对金融文本进行进一步的情感分析有 助于投资者进行投资决策人工判断新闻评论等金融文本的倾向性的工作量 巨大迫切需要一种面向金融领域的自动文本情感分析技术和工具金融文本 的情感分析技术主要受制于高性能的文本情感分类算法和高质量的金融文本情 感标注语料库 为此本文首
4、先对金融文本的倾向性判定方法进行了研究结合文本情感 分类特点本文对感知器算法进行改进提出了基于优化感知器的文本倾向性 分析方法通过观察和分析金融文本的特点本文提取了情感词评价词等特 征来表示金融文本传统的感知器方法的优点是可以通过错误分类的样本来调 节各个特征的权重但却存在确定学习率的问题而传统的基于情感词典统计 的倾向性分析方法能够采用固定特征权值进行统计所以本文结合感知器和词 典统计方法提出了优化感知器的文本倾向性分析方法实验结果表明优化的 感知器方法能够有效地判断金融文本的倾向性在同一金融语料库上的倾向性 分析效果优于其它方法 其次由于人工标记语料具有主观偏向性的缺点从而影响语料库的一致
5、 性并且导致基于监督学习分析方法的综合性能下降借鉴 PageRank 算法的思 想本文提出了一种新的 DocRank 算法算法采用文本图的结构来表示文档间 相互关系优选出具有代表性的情感样本作为训练语料集合通过多组对比实 验表明 DocRank 算法能够对语料库进行优化选择提高语料库的一致性 最后本文将上述算法应用到了专业的金融信息检索系统中从而帮助用 户更好的理解金融文本提升用户体验同时也验证了算法的有效性 关键词情感分析感知器文本图 - I - 哈尔滨工业大学工学硕士学位论文 Abstract With the rapid development of Internet and finan
6、cial industry both provisions and requirements of financial information people needed increase substantially The tendency of financial texts may have a major impact on the related stocks or companies therefore the further analysis of financial texts could help investors make investment decisions It
7、is a hard work for people to judge the tendencies of news review and so on This situation causes an urgently demand which needs an automated tool with analysis technology to determine news orientation However the technology is subject to the sentimental analysis algorithms with high performance and
8、emotional corpus with high quality For this problem this paper first does research the methods which analyze the inclination of financial news According to the characteristics of the sentimental classification this paper improves the general perceptron algorithm and proposes an optimized perceptron
9、method to distinguish the inclination of the news By observing and analyzing the characteristics of financial texts this paper extracts the words set which includes the sentiment words appraise words and so on to express the financial texts Although the general perceptron method can adjust the weigh
10、ts of each feature by using the misclassified samples it is hard to determine the rate of learning Taking into account the advantage of dictionary-based statistical method which can static the target objects using the fixed weight values of the statistical characteristics this paper achieves optimiz
11、ed perceptron method by combining the advantages of the perceptron method and dictionary-based statistical method The experiments show that the optimal perceptron method can effectively distinguish the inclination of the financial news and be superior to other methods on the effect of doing with the
12、 same financial corpus And then corpus which is marked artificial has the defect of subjective bias that affects the corpus consistency to some extent thereby decline overall performance of the supervised learning method This paper proposes a new DocRank algorithm which chooses the most representati
13、ve sample s as the training data set This algorithm which is based on text-graph structure and draws on the idea of PageRank algorithm is an optimal training method The experiments which - II - 哈尔滨工业大学工学硕士学位论文 are the base of the corpus after training by the DocRank algorithm also prove the effect o
14、f the perceptron method and the purified performance of this rank algorithm Finally above the algorithms proposed are applied to a professional financial information retrieval system This can help users understand the financial text s better and improve their experiences This also verifies the effectiveness of these algorithms Keywords Sentimental Anal