NSLPCD: Topic based tweets clustering using Node significance based label propagation community detection algorithm

Social networks like Twitter, Facebook have recently become the most widely used communication platforms for people to propagate information rapidly. Fast diffusion of information creates accuracy and scalability issues towards topic detection. Most of the existing approaches can detect the most pop...

Full description

Saved in:
Bibliographic Details
Published in:Annals of mathematics and artificial intelligence 2021-03, Vol.89 (3-4), p.371-407
Main Authors: Singh, Jagrati, Singh, Anil Kumar
Format: Article
Language:English
Subjects:
Algorithms
Analysis
Artificial Intelligence
Clustering
Communications software
Complex Systems
Computer Science
Diffusion rate
Frequency distribution
Keywords
Labels
Mathematics
Social media
Social networks
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Social networks like Twitter, Facebook have recently become the most widely used communication platforms for people to propagate information rapidly. Fast diffusion of information creates accuracy and scalability issues towards topic detection. Most of the existing approaches can detect the most popular topics on a large scale. However, these approaches are not effective for faster detection. This article proposes a novel topic detection approach – Node Significance based Label Propagation Community Detection (NSLPCD) algorithm, which detects the topic faster without compromising accuracy. The proposed algorithm analyzes the frequency distribution of keywords in the collection of tweets and finds two types of keywords: topic-identifying and topic-describing keywords, which play an important role in topic detection. Based on these defined keywords, the keyword co-occurrence graph is built, and subsequently, the NSLPCD algorithm is applied to get topic clusters in the form of communities. The experimental results using the real data of Twitter, show that the proposed method is effective in quality as well as run-time performance as compared to other existing methods.
ISSN:1012-2443
1573-7470
DOI:10.1007/s10472-020-09709-z