Text Classification Techniques Research Articles

A systematic review of social network sentiment analysis with comparative study of ensemble-based techniques

Sentiment Analysis (SA) of text reviews is an emerging concern in Natural Language Processing (NLP). It is a broadly active method for analyzing and extracting opinions from text using individual or ensemble learning techniques. This field has unquestionable potential in the digital world and social media platforms. Therefore, we present a systematic survey that organizes and describes the current scenario of the SA and provides a structured overview of proposed approaches from traditional to advance. This work also discusses the SA-related challenges, feature engineering techniques, benchmark datasets, popular publication platforms, and best algorithms to advance the automatic SA. Furthermore, a comparative study has been conducted to assess the performance of bagging and boosting-based ensemble techniques for social network SA. Bagging and Boosting are two major approaches of ensemble learning that contain various ensemble algorithms to classify sentiment polarity. Recent studies recommend that ensemble learning techniques have the potential of applicability for sentiment classification. This analytical study examines the bagging and boosting-based ensemble techniques on four benchmark datasets to provide extensive knowledge regarding ensemble techniques for SA. The efficiency and accuracy of these techniques have been measured in terms of TPR, FPR, Weighted F-Score, Weighted Precision, Weighted Recall, Accuracy, ROC-AUC curve, and Run-Time. Moreover, comparative results reveal that bagging-based ensemble techniques outperformed boosting-based techniques for text classification. This extensive review aims to present benchmark information regarding social network SA that will be helpful for future research in this field.

SEmHuS: a semantically embedded humanitarian space

Humanitarian crises are unpredictable and complex environments, in which access to basic services and infrastructures is not adequately available. Computing in a humanitarian crisis environment is different from any other environment. In humanitarian environments the accessibility to electricity, internet, and qualified human resources is usually limited. Hence, advanced computing technologies in such an environment are hard to deploy and implement. Moreover, time and resources in those environments are also limited and devoted for life-saving activities, which makes computing technologies among the lowest priorities for those who operate there. In humanitarian crises, interests and preferences of decision-makers are driven by their original languages, cultures, education, religions, and political affiliations. Hence, decision-making in such environments is usually hard and slow because it solely depends on human capacity in absence of proper computing techniques. In this research, we are interested in overcoming the above challenges by involving machines in humanitarian response. This work proposes and evaluates a text classification and embedding technique to transform historical humanitarian records from human-oriented into a machine-oriented structure (in a vector space). This technique allows machines to extract humanitarian knowledge and use it to answer questions and classify documents. Having machines involved in those tasks helps decision-makers in speeding up humanitarian response, reducing its cost, saving lives, and easing human suffering.

Deep learning for the detection of acquired and non-acquired skills in students' algorithmic assessments

This research is part of the Artificial Intelligence Virtual Trainer (AI-VT) project which aims to create a system that can identify the user's skills from a text by means of machine learning. AI-VT is a case-based reasoning learning support system can generate customized exercise lists that are specially adapted to user needs. To attain this outcome, the relevance of the first proposed exercise must be optimized to assist the system in creating personalized user profiles. To solve this problem, this project was designed to include a preliminary testing phase. As a generic tool, AI-VT was designed to be adapted to any field of learning. The most recent application of AI-VT was in the field of computer science specifically in the context of the fundamentals of algorithmic learning. AI-VT can and will also be useful in other disciplines. Developed in Python with the Keras API and the Tensorflow framework, this artificial intelligence-based tool encompasses a supervised learning environment, multi-label text classification techniques and deep neural networks. This paper presents and compares the performance levels of the different models tested on two different data sets in the context of computer programming and algorithms.

Text classification technique for discovering country-based publications from international COVID-19 publications.

The significant increase in the number of COVID-19 publications, on the one hand, and the strategic importance of this subject area for research and treatment systems in the health field, on the other hand, reveals the need for text-mining research more than ever. The main objective of the present paper is to discover country-based publications from international COVID-19 publications with text classification techniques. The present paper is applied research that has been performed using text-mining techniques such as clustering and text classification. The statistical population is all COVID-19 publications from PubMed Central® (PMC), extracted from November 2019 to June 2021. Latent Dirichlet allocation (LDA) was used for clustering, and support vector machine (SVM), scikit-learn library, and Python programming language were used for text classification. Text classification was applied to discover the consistency of Iranian and international topics. The findings showed that seven topics were extracted using the LDA algorithm for international and Iranian publications on COVID-19. Moreover, the COVID-19 publications show the largest share in the subject area of "Social and Technology in COVID-19" at the international (April 2021) and national (February 2021) levels with 50.61% and 39.44%, respectively. The highest rate of publications at international and national levels was in April 2021 and February 2021, respectively. One of the most important results of this study was discovering a common trend and consistency of Iranian and international publications on COVID-19. Accordingly, in the topic category "Covid-19 Proteins: Vaccine and Antibody Response," Iranian publications have a common publishing and research trend with international ones.

Active Learning Based on Transfer Learning Techniques for Text Classification

Text preprocessing is a common task in machine learning applications that involves hand-labeling sets. Although automatic and semi-automatic annotation of text data is a growing field, researchers need to develop models that use resources as efficiently as possible for a learning task. The goal of this work was to learn faster with fewer resources. In this paper, the combination of active and transfer learning was examined with the purpose of developing an effective text categorization method. These two forms of learning have proven their efficiency and capacity to train correct models with substantially less training data. We considered three types of criteria for selecting training points: random selection, uncertainty sampling criterion and active transfer selection. Experimental evaluation was performed on five data sets from different domains. The findings of the experiments suggest that by combining active and transfer learning, the algorithm performs better with fewer labels than random selection of training points.

BIOMEDICAL TEXT DOCUMENT CLASSIFICATION

Information extraction, retrieval, and text categorization are only a few of the significant research fields covered by "bio medical text classification." This study examines many text categorization techniques utilised in practise, as well as their strengths and weaknesses, in order to improve knowledge of various information extraction opportunities in the field of data mining. We compiled a dataset with a focus on three categories: "Thyroid Cancer," "Lung Cancer," and "Colon Cancer." This paper presents an empirical study of a classifier. The investigation was carried out using biomedical literature benchmarks. Many metaheuristic algorithms are investigated, including genetic algorithms, particle swarm optimisation, firefly, cuckoo, and bat algorithms. In addition, the proposed multiple classifier system outperforms ensemble learning, ensemble pruning, and traditional classification methods. Based on the data, we forecast if it is Thyroid Cancer, Lung Cancer, or Colon Cancer using basic EDA, text preprocessing, and several models such as Logistic Regression, Decision Tree Classification, and Random Forest Classification.

The digital and sustainable transition of the agri-food sector

According to recent trends, food production must double by 2050 to meet the world's growing population's expected demand. To achieve this goal, agri-food companies have begun implementing different digital technologies to increase food production while utilising fewer resources, thus reducing production processes' environmental impact. This study aims to review Industry 4.0 and agri-food sustainability research published in the last decade. Text classification and data extraction machine learning techniques have been used to support the literature review process. Notably, text classification was used to support the screening phase of titles and abstracts, while data extraction was used to support the content analysis phase by identifying the main topics on which documents are focused. The descriptive analysis shows a summary of the leading scientific journals in the research field, as well as the most influential countries and the research topic evolution over time. The results of the study allowed us to identify ten main research clusters, providing in-depth discussions and perspectives on critical areas for future research avenues. Finally, this study provides significant implications for the agri-food industry, suggesting firms redesign their business models according to a logic that prioritises long-term, shared value creation over short-term efficiency, and profitability. Incorporating digital technologies may help control farming activities' impact on soil and air quality, minimising the use of natural resources, pollutants, and CO2 emissions, thus providing long-term economic, environmental, and social advantages.

Review on sentiment analysis for text classification techniques from 2010 to 2021

Review on sentiment analysis for text classification techniques from 2010 to 2021

Machine driven classification of open-ended responses (MDCOR): An analytic framework and no-code, free software application to classify longitudinal and cross-sectional text responses in survey and social media research

Open-ended questions in survey research allow participants to respond freely using their own words. Because such questions offer the possibility of learning how or why respondents may have achieved a goal or behaved in certain ways, these responses can address some of the inherent limitations of quantitative research, which typically does not allow researchers to understand processes or reasons. But such knowledge-based benefits come at the cost of having to label text data into categories or codes to ease their comparison and reach meaningful understandings. Manually classifying open-ended responses is not only time consuming—potentially taking weeks or even months, depending on sample size—but also risks introducing human errors or inconsistencies that can reduce the contribution of these responses in strengthening our understandings. In this study, we discuss the unresolved issue of how to close open-ended responses as rigorously and efficiently as possible relying on machine learning and text classification techniques, without losing context nor the original voices of our research participants, and while leveraging the nuances that human reasoning brings to the qualitative and mixed methods analytic tables. To this end, we offer a rigorous, user-friendly, no-code, and cost-free software application that implements our mixed equal-status design analytic framework: machine driven classification of open-ended responses (MDCOR). To test the performance of MDCOR, we analyzed tens of thousands of open-ended responses from two different surveys—one publicly available and one federally-protected. In all instances, MDCOR consistently offered time-efficient and reliable results and even tested whether non-response was associated with respondents’ attributes. Among its multiple outputs, MDCOR allows researchers to access the fully classified responses that can then be used in traditional quantitative modeling. Since MDCOR runs locally, its versatility to handling cross-sectional and longitudinal responses, enables the analysis of a variety of data, from federally protected/restricted sources to the classification of social media posts. By removing manual classification burdens and computer programming expertise, MDCOR opens the possibility of efficiently and rigorously reaping the knowledge-based benefits of open-ended responses in survey research without losing or altering our participants’ voices. We offer access to the public data analyzed (https://cutt.ly/YNmBOAL or González Canché, 2022c) and the software (Mac version here https://cutt.ly/xv6nnuN, Windows version https://cutt.ly/Lv6nTLG, Code Ocean version at González Canché (2022f)) so that researchers can interact first-hand with MDCOR and start using this tool and analytic framework in their own studies (see also González Canché, 2022a, 2022b, 2022d, 2022e, for related no-code data science applications to analyze qualitative data dynamically).

BHGAttN: A Feature-Enhanced Hierarchical Graph Attention Network for Sentiment Analysis.

Recently, with the rise of deep learning, text classification techniques have developed rapidly. However, the existing work usually takes the entire text as the modeling object and pays less attention to the hierarchical structure within the text, ignoring the internal connection between the upper and lower sentences. To address these issues, this paper proposes a Bert-based hierarchical graph attention network model (BHGAttN) based on a large-scale pretrained model and graph attention network to model the hierarchical relationship of texts. During modeling, the semantic features are enhanced by the output of the intermediate layer of BERT, and the multilevel hierarchical graph network corresponding to each layer of BERT is constructed by using the dependencies between the whole sentence and the subsentence. This model pays attention to the layer-by-layer semantic information and the hierarchical relationship within the text. The experimental results show that the BHGAttN model exhibits significant competitive advantages compared with the current state-of-the-art baseline models.

Development of a fake news detection tool for Vietnamese based on deep learning techniques

With the development of the Internet, social networks and different communication channels, people can get information quickly and easily. However, in addition to real and useful news, we also receive false and unreal information. The problem of fake news has become a difficult and unresolved issue. For languages with few users, such as Vietnamese, the research on fake news detection is still very limited and has not received much attention. In this paper, we present research results on building a tool to support fake news detection for Vietnamese. Our idea is to apply text classification techniques to fake news detection. We have built a database of 4 groups of 2 topics about politics (fake news and real news) and about Covid-19 (fake news and real news). Then use deep learning techniques CNN (Convolutional Neural Network) and RNN (Recurrent Neural Network) to create the corresponding models. When there is new news that needs to be verified, we just need to apply the classification to see which of the four groups they label into to decide whether it is fake news or not. The tool was able to detect fake news quickly and easily with a correct rate of about 85 %. This result will be improved when getting a larger training data set and adjusting the parameters for the machine learning model. These results make an important contribution to the research on detecting fake news for Vietnamese and can be applied to other languages. In the future, besides using classification techniques (based on content analysis), we can combine many other methods such as checking the source, verifying the author's information, checking the distribution process to improve the quality of fake news detection.

Text Classification for Records Management

Automatic classification of electronic records is necessary to address the brewing crisis in the recordkeeping discipline, caused by escalating data volumes and digital rights legislation. Current solutions usually employ expert systems that classify records based on their metadata, but this approach is becoming unfeasible due to the increased variety of records and a growing lack of metadata. Text classification is a promising alternative now that the records themselves are machine readable. In this study, the performance of traditional text classification techniques was compared to newer natural language processing technologies in a series of experiments using authentic records data. While the latest Transformer language models showed superior classification skill, traditional methods still perform well. These results were discussed by a focus group of record managers, who believe that text classification can help them manage risk and meet compliance obligations. This is a first step toward aspirations of being able to synthesize narrative from a corpus of records.

Natural Language Processing for the Study of Institutional Change

Text Classification has been a well-studied task in Natural Language Processing. In this study deep learning methods are applied to train classifiers for the purpose of analyzing a data set of institutional change propaganda technique samples. Specifically this study explores how collective emotions underpin individuals’ participation in disruption of the hydropower institutionalized energy system. By exploring the context of local initiatives which seek to undermine the hydropower energy system, this study taps into micro-foundations of institutional disruption. The preliminary qualitative study reveals existing patterns in how emotions are tied to various steps of instigating and promoting institutional change. Text classification techniques are further applied to confirm and complement the preliminary findings. In particular Long short-term memory neural networks as well as transformer-based models are applied at two levels of classification: sentence level and token level. The trained classification models help investigate how generalizable the preliminary findings are. By combining qualitative and quantitative this study extends prior research by unpacking emotional dynamics implicated in institutional disruption by explaining how actors mobilize collective emotions as a basis to build support for institutional change.

Research on Text Classification based on Deep Learning

Text classification is a fundamental task in natural language processing, and machine learning and deep learning have seen more research and great progress in this task. In this paper, we introduce the current state of research and analyse the application of deep learning networks to text classification, based on the concept of text classification, feature extraction, text representation and other related techniques and text classification methods.

A Novel Text Classification Technique Using Improved Particle Swarm Optimization: A Case Study of Arabic Language

We propose a novel text classification model, which aims to improve the performance of Arabic text classification using machine learning techniques. One of the effective solutions in Arabic text classification is to find the suitable feature selection method with an optimal number of features alongside the classifier. Although several text classification methods have been proposed for the Arabic language using different techniques, such as feature selection methods, an ensemble of classifiers, and discriminative features, choosing the optimal method becomes an NP-hard problem considering the huge search space. Therefore, we propose a method, called Optimal Configuration Determination for Arabic text Classification (OCATC), which utilized the Particle Swarm Optimization (PSO) algorithm to find the optimal solution (configuration) from this space. The proposed OCATC method extracts and converts the features from the textual documents into a numerical vector using the Term Frequency-Inverse Document Frequency (TF–IDF) approach. Finally, the PSO selects the best architecture from a set of classifiers to feature selection methods with an optimal number of features. Extensive experiments were carried out to evaluate the performance of the OCATC method using six datasets, including five publicly available datasets and our proposed dataset. The results obtained demonstrate the superiority of OCATC over individual classifiers and other state-of-the-art methods.

MACHINE LEARNING FOR TEXT CLASSIFICATION IN BUILDING MANAGEMENT SYSTEMS

In building management systems (BMS), a medium building may have between 200 and 1000 sensor points. Their labels need to be translated into a naming standard so they can be automatically recognised by the BMS platform. The current industrial practices often manually translate these points into labels (this is known as the tagging process), which takes around 8 hours for every 100 points. We introduce an AI-based multi-stage text classification that translates BMS points into formatted BMS labels. After comparing five different techniques for text classification (logistic regression, random forests, XGBoost, multinomial Naive Bayes and linear support vector classification), we demonstrate that XGBoost is the top performer with 90.29% of true positives, and use the prediction confidence to filter out false positives. This approach can be applied in sensors networks in various applications, where manual free-text data pre-processing remains cumbersome.

A Method for Classifying Information in Education Policy Texts Based on an Improved Attention Mechanism Model

Text classification techniques in natural language processing can automatically classify text data in a more efficient way, saving human resources and costs. Therefore, text classification techniques can be applied to the automatic classification of education policy data to quickly locate and accurately find education policy data, thus realising the information management and visual analysis of education policy data. This paper proposes a text classification algorithm based on the attention mechanism of headline and body text according to the characteristics of education policy and introduces this algorithm in detail through a model diagram. The experimental comparison with existing classification algorithms verifies the superiority of the algorithm in the classification method of education policy text classification information.

Attention Based Encoder Architecture for Automatic Text Classification: A Case Study on Text-news Classification

Text classification is a classical problem of natural language processing where given texts are classified into relevant classes. Text classification techniques are extensively used in finding news categories, search engine optimization, automated textual response and many more. With the exponential growth of digital data, it is inevitable that an automatic classification of documents is needed for efficient information retrieval and document archival. Over the years, text classification problems are approached by using classical machine learning techniques like naive bayes(NB), support vector machine(SVM), logistic regression(LR), random forest(RF) and others. In this article, an Attention based deep learning (DL) model is proposed and applied to the news corpus and then the results are compared with the results of classical machine learning models. A comparative analysis of classification accuracy of the DL model with prominent models like LR, RF and NB are presented in this work.

Automatic Classification of National Health Service Feedback

Text datasets come in an abundance of shapes, sizes and styles. However, determining what factors limit classification accuracy remains a difficult task which is still the subject of intensive research. Using a challenging UK National Health Service (NHS) dataset, which contains many characteristics known to increase the complexity of classification, we propose an innovative classification pipeline. This pipeline switches between different text pre-processing, scoring and classification techniques during execution. Using this flexible pipeline, a high level of accuracy has been achieved in the classification of a range of datasets, attaining a micro-averaged F1 score of 93.30% on the Reuters-21578 “ApteMod” corpus. An evaluation of this flexible pipeline was carried out using a variety of complex datasets compared against an unsupervised clustering approach. The paper describes how classification accuracy is impacted by an unbalanced category distribution, the rare use of generic terms and the subjective nature of manual human classification.

A Survey on Text Classification Algorithms: From Text to Predictions

In recent years, the exponential growth of digital documents has been met by rapid progress in text classification techniques. Newly proposed machine learning algorithms leverage the latest advancements in deep learning methods, allowing for the automatic extraction of expressive features. The swift development of these methods has led to a plethora of strategies to encode natural language into machine-interpretable data. The latest language modelling algorithms are used in conjunction with ad hoc preprocessing procedures, of which the description is often omitted in favour of a more detailed explanation of the classification step. This paper offers a concise review of recent text classification models, with emphasis on the flow of data, from raw text to output labels. We highlight the differences between earlier methods and more recent, deep learning-based methods in both their functioning and in how they transform input data. To give a better perspective on the text classification landscape, we provide an overview of datasets for the English language, as well as supplying instructions for the synthesis of two new multilabel datasets, which we found to be particularly scarce in this setting. Finally, we provide an outline of new experimental results and discuss the open research challenges posed by deep learning-based language models.