时变论域下区间二型模糊集合的在线评论分析

发布时间：2019-05-19 14:24

【摘要】：在线评论不仅实时的反应了消费者当前的关注点,而且作为已购买者反馈的评论,实时的、准确的传递了对购买过程及购买商品的理性认知;可以给潜在的购买者提供更为有用的感知风险的参考点;这样给消费者带来更为便利的帮助。论文首先通过时变论域下区间二型模糊集合表述评论语言,建立基于评论语言分析所得的店铺商品现状的动态模糊规则及其词计算。利用模糊聚类分析方法,对在线评论语言关键词包含的内容进行分类。简化语言评论词汇的丰富性,同时通过此方法建立了目标对象的因素集。然后,采用了离散时变论域来描述在线店铺商品,用二型模糊集合表述基词分析在线评论,从在线评论文本关键词为语言文字和具体的数字两方面综合考虑,确定商务店铺商品现状各因素集的区间二型模糊集合。并运用模糊综合评判方法及语言动力系统理论对不同店铺的同类商品的现状进行动态评估,同时采用二型模糊集合来表述由在线评论分析所得的店铺同类商品现状结果。最后,以连续的区间二型模糊集合为基础,对在线评论进行分析,建立商务店铺现状的评估和措施改善模型。设定控制对象预期目标,根据其模糊规则及其词计算,确定可参考的干扰或者改善措施,形成闭环控制,得到预期的语言动力学轨迹,完成目标控制任务,并应用到在线店铺现状改善措施的确定的研究中。论文提出了一种区间二型模糊集合的方法,对在线评论进行建模与分析。在当前的环境下,给顾客带来更多便利地、有效地、直接地决策依据,同时给商家制定更合理的改进方案提供参考依据。由于在线评论语言具有模糊性,随着时间的变化人们评价事物的标准也会发生变化。因此,动态模糊规则及其时变论域更适合解决语言评价类问题。同时由于时变论域的可变实时性,使该方法自适应能力增强,推理更为准确。
[Abstract]:Online comments not only reflect the current concerns of consumers in real time, but also convey the rational cognition of the purchase process and goods in real time and accurately as the feedback of the purchasers. It can provide potential buyers with more useful reference points for risk perception; this will bring more convenient help to consumers. In this paper, the dynamic fuzzy rules and word calculation of shop goods based on the analysis of comment language are established by expressing the comment language by interval two fuzzy sets under the time-varying domain. The fuzzy clustering analysis method is used to classify the content contained in the keywords of online comment language. The richness of language comment vocabulary is simplified, and the factor set of target object is established by this method. Then, the discrete time-varying domain is used to describe the online store goods, and the two-type fuzzy set is used to express the base words to analyze the online comments, which is considered comprehensively from two aspects: the keywords of the online comment text are the language and the specific numbers. Determine the interval type 2 fuzzy set of each factor set of the present situation of business store commodities. The fuzzy comprehensive evaluation method and linguistic power system theory are used to dynamically evaluate the present situation of the same kind of goods in different stores, and the two-type fuzzy set is used to express the present situation results of the same kind of goods obtained from the online review and analysis. Finally, based on the continuous interval two-type fuzzy set, the online comments are analyzed, and the evaluation of the present situation of business stores and the improvement model of measures are established. Set the expected goal of the control object, according to its fuzzy rules and its word calculation, determine the interference or improvement measures that can be referred to, form the closed-loop control, obtain the expected language dynamics trajectory, and complete the target control task. And applied to the online store status quo improvement measures to determine the research. In this paper, a method of interval two fuzzy set is proposed to model and analyze online comments. In the current environment, it brings more convenient, effective and direct decision-making basis for customers, and provides a reference for businesses to make more reasonable improvement programs. Because of the fuzziness of online comment language, people's criteria for evaluating things will also change with the change of time. Therefore, dynamic fuzzy rules and their time-varying domains are more suitable for solving language evaluation problems. At the same time, due to the variable real-time performance of time-varying domain, the adaptive ability of the method is enhanced and the reasoning is more accurate.
【学位授予单位】：长沙理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F724.6;TP391.1

【参考文献】