International Conference Machine Learning, Big Data, Cloud and Parallel Computing最新文献

筛选

英文中文

An Enhanced Method for Topic Modeling using Concept-Latent 一种基于概念潜的主题建模改进方法

International Conference Machine Learning, Big Data, Cloud and Parallel Computing Pub Date : 2019-02-01 DOI: 10.1109/COMITCON.2019.8862179

A. Christy, G. Meeragandhi

{"title":"An Enhanced Method for Topic Modeling using Concept-Latent","authors":"A. Christy, G. Meeragandhi","doi":"10.1109/COMITCON.2019.8862179","DOIUrl":"https://doi.org/10.1109/COMITCON.2019.8862179","url":null,"abstract":"In this digital era, Topic Modeling plays a vital role in clustering of documents. In topic modeling, each document is represented as a collection of words. Labeling appropriate word distributions to the defined topic is a challenging task. Topic modeling using unsupervised algorithm does not align topics with those as it exists in the real world metaphor. This problem of aligning topic with real world objects can be carried out by two different methods-either to use a supervised learning algorithm that assigns individual topic with a label acquired from existing data or to use a predefined set of topics and to match them with the distribution of words. The above said methods have some drawbacks as the former method may match a label which does not convey a semantic meaning with the word distributions whereas the latter lacks the knowledge required to detect the topic from the closed set of topics. In this paper an algorithm called Concept-Latent Dirichlet Allocation (Concept-LDA) is proposed which incorporates reinforcement learning for topic modeling which inturn improves the accuracy of resulting topics and of the topic labeling. Experiments were conducted using LDA and Concept-LDA for topic modeling and it is shown that Concept-LDA showing better accuracy than LDA.","PeriodicalId":272832,"journal":{"name":"International Conference Machine Learning, Big Data, Cloud and Parallel Computing","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125792210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 3