当前位置: X-MOL 学术IEEE Lat. Am. Trans. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Instance Genetic Selection for Fuzzy Rule-based Systems Optimization to Opinion Classification
IEEE Latin America Transactions ( IF 1.3 ) Pub Date : 2020-07-01 , DOI: 10.1109/tla.2020.9099762
Tayane Leite Cerqueira 1 , Fabiana Cristina Bertoni 2 , Matheus Giovanni Pires 2
Affiliation  

Opinion Mining aims to identify people feelings about some item of interest based on content available on the Web, without the user having to find and read all the news about it. As opinions are related to feelings that are often described by imprecise terms, Fuzzy Systems appear as an alternative to treat the information subjectivity. An important task in developing Fuzzy Systems is define the rule base usually based on textual databases in case of opinion mining. However, these databases are extensive and the algorithms used to generate the rule base often result in many rules, making it difficult to achieve accuracy with a low computational cost. To deal with this problem, instance selection can be applied to reduce databases in order to save only relevant data. Thus, the aim of the present study is optimize a Fuzzy System, using instance selection to generate a reduced rule base. As the issue mentioned is a multiobjective problem, which seeks to increase accuracy and reduce the number of rules, we have chosen to apply a Multiobjective Genetic Algorithm, since it has been acknowledged as a promising approach in the literature. The results demonstrate that the Multiobjective Genetic Algorithms can be applied in instance selection for opinion classification problems, presenting a reduction in the number of instances and execution time, without significant changes in accuracy.

中文翻译:

基于模糊规则的系统优化到意见分类的实例遗传选择

意见挖掘旨在根据 Web 上可用的内容识别人们对某个感兴趣项目的感受,而无需用户查找和阅读有关它的所有新闻。由于意见与通常用不精确术语描述的感觉有关,因此模糊系统似乎是处理信息主观性的替代方案。开发模糊系统的一个重要任务是在意见挖掘的情况下定义通常基于文本数据库的规则库。然而,这些数据库是广泛的,并且用于生成规则库的算法通常会产生许多规则,因此很难以较低的计算成本实现准确性。为了解决这个问题,可以应用实例选择来减少数据库,以便只保存相关数据。因此,本研究的目的是优化一个模糊系统,使用实例选择来生成简化的规则库。由于提到的问题是一个多目标问题,它旨在提高准确性并减少规则数量,我们选择应用多目标遗传算法,因为它在文献中被认为是一种有前途的方法。结果表明,多目标遗传算法可以应用于意见分类问题的实例选择,减少实例数量和执行时间,而不会显着改变准确性。
更新日期:2020-07-01
down
wechat
bug