基于文法的产品舆情分析方法研究

发布时间：2018-06-27 21:50

本文选题：产品舆情 + 舆情主题　；参考：《广西师范大学》2016年硕士论文

【摘要】：当今社会,市场经济的迅猛发展,社会经济在各种复杂的深层次的社会矛盾下日益凸显,由此引发的产品质量问题日益引起民众的关注,产品质量和售后服务问题究成为决定企业发展和用户消费的关键因素。产品网络舆情分析成为政府、行业、企业和消费者关注的重要研究课题。目前,产品舆情分析的主要研究方法都是依靠关键词从社交网络等渠道获取消费者的产品评价。由于缺乏对舆情语义的分析,这种基于统计的方法和关键词的抽取有以下缺陷：(1)传统舆情系统会识别出比较多的噪音数据,舆情识别的精度低,无法达到政府和企业所需要产品舆情的效果。(2)网络舆情的去重、分类、正负面评价等自然语言分析技术不够实用。(3)在舆情发展初期,由于数据稀疏,基于关键词的统计方法难以及时发现舆情。为解决这些问题,有些学者提出了基于领域语义文法的舆情文法分析方法,但是由于产品种类的多样性,这种方法不完全适用于产品舆情分析。本文提出了一种基于产品舆情主题结构和产品舆情文法(称为PPOG)的产品舆情分析方法。通过大量实际的产品舆情例子测试,基于PPOG的分析方法具有良好的应用前景。本文的主要研究工作如下：(1)利用国家标准《统计用产品分类目录》对各类产品建立统一的分类框架,给出了基于该框架的产品舆情主题分类,体现了产品舆情分析的基础。(2)提出了利用产品分类结构和舆情主题采用自顶向下的产品舆情文法设计方法来建立产品舆情文法。(3)通过对舆情语料和种子模式库的观察和总结,提出了一种利用种子模式库对产品舆情文法中的冗余义项词类进行义项清除的方法。(4)提出了产品舆情文法设计过程中,以及文法设计师需要遵循的重要原则和方法。(5)在LL(1)基础上,实现了一个产品舆情分析的原型系统,并且进行了试验,取得了初步的成果。
[Abstract]:In today's society, with the rapid development of the market economy, the social economy is becoming increasingly prominent under various complicated and deep-seated social contradictions, and the product quality problems arising therefrom are attracting more and more people's attention. The problem of product quality and after-sale service becomes the key factor of enterprise development and consumer consumption. The analysis of product network public opinion has become an important research topic of government, industry, enterprises and consumers. At present, the main research methods of product public opinion analysis rely on keywords to obtain product evaluation from social network and other channels. Due to the lack of semantic analysis of public opinion, the statistical method and keyword extraction have the following shortcomings: (1) the traditional public opinion system will recognize more noise data, and the accuracy of public opinion recognition is low. (2) the natural language analysis technology of network public opinion is not practical enough. (3) in the initial stage of the development of public opinion, the data is sparse. Keyword-based statistical methods are difficult to find public opinion in time. In order to solve these problems, some scholars have proposed a method of public opinion grammar analysis based on domain semantic grammar. However, due to the diversity of product types, this method is not fully applicable to product public opinion analysis. In this paper, a product public opinion analysis method based on product public opinion theme structure and product public opinion grammar (PPOG) is proposed. The analysis method based on PPOG has a good application prospect through a large number of practical examples of product public opinion. The main research work of this paper is as follows: (1) A unified classification framework for all kinds of products is established by using the national standard "Product Classification catalogue for Statistics", and the topic classification of product public opinion based on this framework is given. It reflects the basis of product public opinion analysis. (2) A top-down design method of product public opinion grammar is proposed by using product classification structure and topic of public opinion. (3) through the observation and summary of public opinion corpus and seed pattern database, In this paper, a method of removing redundant meanings in product public opinion grammar by using seed pattern library is proposed. (4) in the process of designing product public opinion grammar, And the important principles and methods to be followed by grammar designers. (5) A prototype system of product public opinion analysis is implemented on the basis of LL (1).
【学位授予单位】：广西师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.1

【相似文献】