TextNetTopics Pro, a topic model-based text classification for short text by integration of semantic and document-topic distribution information

Abstract

With the exponential growth in the daily publication of scientific articles, automatic classification and categorization can assist in assigning articles to a predefined category. Article titles are concise descriptions of the articles' content with valuable information that can be useful in document classification and categorization. However, shortness, data sparseness, limited word occurrences, and the inadequate contextual information of scientific document titles hinder the direct application of conventional text mining and machine learning algorithms on these short texts, making their classification a challenging task. This study firstly explores the performance of our earlier study, TextNetTopics on the short text. Secondly, here we propose an advanced version called TextNetTopics Pro, which is a novel short-text classification framework that utilizes a promising combination of lexical features organized in topics of words and topic distribution extracted by a topic model to alleviate the data-sparseness problem when classifying short texts. We evaluate our proposed approach using nine state-of-the-art short-text topic models on two publicly available datasets of scientific article titles as short-text documents. The first dataset is related to the Biomedical field, and the other one is related to Computer Science publications. Additionally, we comparatively evaluate the predictive performance of the models generated with and without using the abstracts. Finally, we demonstrate the robustness and effectiveness of the proposed approach in handling the imbalanced data, particularly in the classification of Drug-Induced Liver Injury articles as part of the CAMDA challenge. Taking advantage of the semantic information detected by topic models proved to be a reliable way to improve the overall performance of ML classifiers.

Keywords: feature selection; short text; sparse data; text classification; topic modeling; topic projection; topic selection.

FIGURE 1

Workflow of TextNetTopics Pro.

FIGURE 2

TextNetTopics performance over accumulated top-ranked topics for the CAMDA dataset using various short-text topic models in the T component. Symbols along the line represent the number of accumulated topics.

FIGURE 3

TextNetTopics performance over 140 features/terms for the CAMDA dataset using various short-text topic models in the T component.

FIGURE 4

TextNetTopics performance over accumulated top-ranked topics for the arXiv dataset using various short-text topic models in the T component. Symbols along the line represent the number of accumulated topics.

FIGURE 5

TextNetTopics performance over 140 features/terms for the arXiv dataset using various short-text topic models in the T component.

FIGURE 6

Classification performance of CAMDA dataset when utilizing topical words (TW) extracted by TextNetTopics, topic distribution features (TD) generated by Topic Models, and our proposed approach, combining words of top-ranked topics extracted by TextNetTopics with topic distribution features (TW + TD). The light-colored columns represent the highest achieved values.

FIGURE 7

Classification performance of our proposed approach over the CAMDA dataset, compared with taking all preprocessed terms with the semantic features.

FIGURE 8

Classification performance on the arXiv dataset when utilizing topical words (TW) extracted by TextNetTopics, topic distribution features (TD) generated by Topic Models, and our proposed approach, combining words of top-ranked topics extracted by TextNetTopics with topic distribution features (TW + TD). The light-colored columns represent the highest achieved values.

FIGURE 9

Classification performance on the arXiv dataset when utilizing our proposed approach versus taking all preprocessed terms with the semantic features.

FIGURE 10

Performance of TextNetTopics over accumulated top-ranked topics using various short-text topic models in the T component on regular-sized text, i.e., titles + abstract (CAMDA dataset). Symbols along the line represent the number of accumulated topics.

FIGURE 11

Performance of TextNetTopics Pro over accumulated topic distributions with top-ranked topics using various short-text topic models in the T component on regular-sized text, i.e., titles + abstract (CAMDA dataset). Symbols along the line represent the number of accumulated topics.