Network-Based Bag-of-Words Model for Text Classification

Abstract

The rapidly developing internet and other media have produced a tremendous amount of text data, making it a challenging and valuable task to find a more effective way to analyze text data by machine. Text representation is the first step for a machine to understand the text, and the commonly used text representation method is the Bag-of-Words (BoW) model. To form the vector representation of a document, the BoW model separately matches and counts each element in the document, neglecting much correlation information among words. In this paper, we propose a network-based bag-of-words model, which collects high-level structural and semantic meaning of the words. Because the structural and semantic information of a network reflects the relationship between nodes, the proposed model can distinguish the relation of words. We apply the proposed model to text classification and compare the performance of the proposed model with different text representation methods on four document datasets. The results show that the proposed method achieves the best performance with high efficiency. Using the Eccentricity property of the network as features can get the highest accuracy. We also investigate the influence of different network structures in the proposed method. Experimental results reveal that, for text classification, the dynamic network is more suitable than the static network and the hybrid network.

A Survey of Automatic Text Summarization: Progress, Process and Challenges

“…Computer vision, NLP, Bayesian spam filters, document categorization, and information retrieval utilizing machine learning are all areas where the BoW technique is used. [186], [101], [187], and [188] are the papers in which BOW feature extraction approaches are used.…”

Vision-Based Human Detection Techniques: A Descriptive Review

“…The "part filters" compute features two times of the spatial resolution of the "root filter". Here "root filter" [94], [102], object recognition [103], text classification [104], image retrieval [105] • quite simple to comprehend and implement [94] • can categorize the objects [94] • computationally expensive [95] • skips geometric relationships among visual words [96] • less annotation accuracy [97] SPM object recognition [106], image classification [98], [107] • computationally effective [98] • improves the classification accuracy [98] • weight's mechanism is not sophisticated [99] • insufficient discriminative power [100] BOVW scene classification [108], [108], land-use classification [109], object classification [110] • improves the classification accuracy [108] • visual vocabulary's computational cost is high [111] behaves similar to a Dalal-Triggs method. The outcome of this method at a specific point and scale in a picture is the result of the "root filter" on the window plus the sum over parts, of the limit over positions of that component, of the "part filter" record on the coming from sub-window minus the cost of deformation.…”

An Evolutionary-Based Sentiment Analysis Approach for Enhancing Government Decisions during COVID-19 Pandemic: The Case of Jordan

“…. .x n ], where x i indicates the occurrence for the ith term [51,52]. The occurrence of the term can be binary, term frequency, or TF-IDF [51].…”