Open AccessDEN-DIS: “May Get Life in Future” - Hybridized Data Stream Clustering Framework in Market Research Arena
Author(s) -
F.Jasmine Natchial*,
Elango Parasuraman
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.f3593.079920
Subject(s) - cluster analysis , computer science , outlier , data mining , anomaly detection , data stream mining , data stream , exploratory data analysis , data stream clustering , cure data clustering algorithm , artificial intelligence , correlation clustering , telecommunications
Data streams pose several computational challenges due to their large volume of massive data arriving at a very fast rate. Data streams are gaining the attention of today’s research community for their utility in almost all fields. In turn, organizing the data into groups enables the researchers to derive with many useful and valuable information and conclusions based on the categories that were discovered. Clustering makes this organization or grouping easier and plays an important role in exploratory data analysis. This paper focuses on the amalgamation of two very important algorithms namely Density Based clustering used to group the data and the dissimilarity matrix algorithm used to find the outlier among the data. Before feeding the data, the algorithm filters out the sparse data and a continuous monitoring system provides the frequent outlier and inlier checks on the live stream data using buffer timer. This approach provides an optimistic solution in recognizing the outlier data which may later get reverted as inlier based on certain criteria. The concept of DenDis approach will pave a new innovation world of considering every data which “May Get Life in Future”.