Semantically meaningful group detection within sub-communities of Twitter blogosphere

Abstract

This paper addresses the problem of semantically meaningful group detection within a sub-community of twitter micro-bloggers by utilizing a topic modeling, multi-objective clustering approach. The proposed group detection method is anchored on the Latent Dirichlet Allocation (LDA) topic modeling technique, aiming at identifying clusters of twitter users that are optimal in terms of both spatial and topical compactness. Specifically, the group detection problem is formulated as a multiobjective optimization problem taking into consideration two complementary cluster formation directives. The first objective, related to spatial compactness, is achieved by minimizing the overall deviation from the corresponding cluster centers. The second, related to topical compactness, is achieved by minimizing the portion of probability mass assigned to low probability topics for the corresponding cluster centroids. In our approach, optimization is performed by employing a multi-objective genetic algorithm ,which results in a variety of cluster structures that are significantly more interpretable than cluster assignments obtained with traditional single-objective clustering algorithms.

A novel approach for text categorization by applying hybrid genetic bat algorithm through feature extraction and feature selection methods

“…The results demonstrated the superiority of the ACA as a feature selector among other algorithms. Several studies showed the application of the LDA method as a feature extractor in text categorization using the GA in the feature selection stage (Chen et al, 2017;Panichella et al, 2013;Sotiropoulos et al, 2014;Sotiropoulos et al, 2016). The clustering problem was reformulated by Sotiropoulos et al (2016) as a discrete optimization problem within the n-dimensional standard simplex.…”

“…Results demonstrated that LDA-GA can determine robust LDA configurations. Sotiropoulos et al (2014) proposed a group detection method within a subcommunity of Twitter micro-bloggers by utilizing a topic modeling. They utilized the LDA topic modeling technique to identify clusters of Twitter users considering both spatial and topical compactness.…”

Web Services Classification Based on Wide & Bi-LSTM Model

“…Cao et al [20] proposed a K-Means algorithm based on the similarity of Mashup services to cluster services by combining the description document and the corresponding tags of Mashup services. Sotiropoulos et al [15] presented a service clustering model based on topic probability and domain characteristics. Shi et al [16] provided an enhanced-LDA service clustering method based on word vectors to cluster all the words in the Web service description document, and to take these words clustering information into the training process of the LDA model.…”

“…The purity of each C i and the mean purity of all Web service classification in EWSC are respectively defined in the formulas ( 13) and (14). Similarly, the entropy of each C i and the mean entropy of all Web service categories in EWSC are respectively defined in the formulas (15) and (16).…”