期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：2
页码：2322
DOI：10.15680/IJIRCCE.2017.0502044
出版社：S&S Publications
摘要：Due to the increasing interest of people in online purchasing there has been an inclination towards thedigital marketing. So to make more and more purchasing possible from people the advertisement should be related tothe people’s interest. If the advertisement is not in interest with the people, it will not attract them. This paper attemptsto study the person’s interest and show them the relevant data using Sentiment Analysis with the help of Density BasedClustering Algorithms (DBSCAN). Density Based Clustering Algorithms need the Epsilon(Eps) value and Minimumpoints value(MinPts) to create the clusters. The proposed method accepts the domain knowledge about the data set asan input and calculation of Eps and MinPts is automated which helps to make the data certain to some extent. We firstcreate the grids for the dataset related to domain then it derives the default Eps and MinPts which are input forDBSCALE[Density-Based Clustering Algorithm for Larger Datasets] Algorithm. Experimental results indicate that theproposed method gives the most relevant data which matches people’s interest.
关键词：DBSCAN; DBSCALE; Density – Based Clustering; Epsilon; Domain Knowledge; Grid.
