Article
Mathematics
Francisco J. Ribadas-Pena, Shuyuan Cao, Victor M. Darriba Bilbao
Summary: In this paper, a multi-label lazy learning approach is proposed for automatic semantic indexing in large document collections with complex label vocabularies and high inter-label correlation. The method is evaluated on a portion of the MEDLINE biomedical document collection, using different document representation approaches and label autoencoder configurations.
Article
Mathematical & Computational Biology
Abdullah Alqahtani, Habib Ullah Khan, Shtwai Alsubai, Mohemmed Sha, Ahmad Almadhor, Tayyab Iqbal, Sidra Abbas
Summary: This paper utilizes machine learning and deep learning techniques to classify textual data, achieving high accuracy with LSTM model outperforming others.
FRONTIERS IN COMPUTATIONAL NEUROSCIENCE
(2022)
Article
Oncology
Khajamoinuddin Syed, William C. Sleeman, Michael Hagan, Jatinder Palta, Rishabh Kapoor, Preetam Ghosh
Summary: Standardization of radiotherapy structure names is crucial for personalized treatment plans, and machine learning models that integrate different data types can improve the accuracy of classification. Combining various views of data helps in building better models for structure name standardization, enabling big data analytics in radiation oncology.
Article
Chemistry, Multidisciplinary
Omer Koksal, Bedir Tekinerdogan
Summary: This paper presents an automated bug classification approach applied in an industrial case study. The approach utilizes machine learning, text mining, and natural language processing techniques to support the classification of software bugs. The results show that bug classification can be automated and even outperform manual classification.
APPLIED SCIENCES-BASEL
(2022)
Article
Mathematical & Computational Biology
Longjia Jia, Bangzuo Zhang
Summary: There are two main factors involved in document classification: document representation method and classification algorithm. This study focuses on the document representation method and introduces a strategy called Document Representation based on Global Policy (DRGP), which shows superior results compared to other text representation strategies.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2022)
Article
Computer Science, Information Systems
Andrea Esuli
Summary: This article presents the Interactive Classification System (ICS), a web-based application that utilizes machine learning to provide classification suggestions to users. The key feature of ICS is giving users total freedom to modify classification schemes and labels. The article discusses the challenges that this requirement poses to traditional machine learning research and introduces an unobtrusive machine learning model that addresses these challenges.
Review
Computer Science, Information Systems
Li Kong, Chuanyi Li, Jidong Ge, Vincent Ng, Bin Luo
Summary: In recent years, the number of online product reviews has been increasing rapidly, making it difficult for customers to read through all the reviews. To address this issue, researchers propose a new model for predicting review helpfulness, combining Convolutional Neural Network (CNN) and TransE. The experimental results demonstrate that this approach outperforms the state of the art.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2022)
Article
Computer Science, Interdisciplinary Applications
Ekaterina Ilgisonis, Mikhail A. Pyatnitskiy, Svetlana N. Tarbeeva, Artem A. Aldushin, Elena A. Ponomarenko
Summary: The paper presents a scheme for comparative analysis of Pubmed publications by comparing the frequencies of occurrence of keywords-MeSH terms. The analysis aims to identify MeSH terms that characterize specific research areas and trends in the number of published works. The proposed approach was tested on medical publications and articles in the field of personalized medicine. The analysis revealed increasing research interest in various topics and demonstrated shifts in scientific priorities over the past 10 years. The findings can be used to predict the relevance and significance of scientific work direction in the upcoming years. The proposed analysis can be expanded to larger sets of publications in the future and adjusted by introducing filters or selecting keywords.
Article
Computer Science, Software Engineering
Walid Cherif, Abdellah Madani, Mohamed Kissi
Summary: The article discusses the application of machine learning techniques in text data classification, introducing a new classification approach and highlighting its advantages. Experimental results demonstrate that the method performs well in automatic text categorization.
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE
(2022)
Article
Computer Science, Information Systems
Ali Saleh Alammary
Summary: Classifying assessment questions based on Bloom's taxonomy is important for designing effective assessments. This study addressed the lack of research on classifying Arabic questions and proposed a new feature extraction method tailored for Arabic questions. The evaluation results showed that the proposed method outperformed traditional methods, indicating its potential for application in other languages.
Article
Computer Science, Artificial Intelligence
Mayur Wankhade, Annavarapu Chandra Sekhara Rao, Chaitanya Kulkarni
Summary: This article provides an overview of sentiment analysis, discussing its applications, advantages, and challenges. Sentiment analysis involves extracting subjective information from text and evaluating and categorizing people's opinions.
ARTIFICIAL INTELLIGENCE REVIEW
(2022)
Article
Psychology, Mathematical
Ryan C. Yeung, Myra A. Fernandes
Summary: Detecting and removing invalid texts is a crucial step in text data analysis. However, existing methods lack effectiveness and practicality. In this study, a supervised machine learning approach was proposed and implemented to accurately detect invalid texts, outperforming existing data quality indicators. This approach enables researchers to improve data quality in text-based studies.
BEHAVIOR RESEARCH METHODS
(2022)
Article
Computer Science, Information Systems
Saima Sadiq, Turki Aljrees, Saleem Ullah
Summary: Recent advancements in natural language production and language modelling have empowered deep neural models to generate realistic and influential content, allowing adversaries to manipulate public opinion on social media. To address this problem, current research focuses on identifying machine-generated text on social networks like Twitter. This study proposes a simple deep learning model using word embeddings to classify tweets as human-generated or bot-generated, achieving a superior accuracy of 93% when compared to other baseline methods.
Article
Multidisciplinary Sciences
Jake Lever, Russ B. Altman
Summary: CoronaCentral resource utilizes machine learning to process research literature related to SARS-CoV-2, providing researchers with categorized analysis of content, pace, and focus. The resource, updated daily, covers topics such as therapeutics, disease forecasting, and long COVID.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2021)
Review
Computer Science, Information Systems
Brian Ondov, Kush Attal, Dina Demner-Fushman
Summary: Plain language in medicine has long been advocated to improve patient understanding and engagement. As Natural Language Processing advances, methods for automatic simplification of biomedical text have become increasingly sophisticated, with procedural and neural methods being the main approaches.
JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION
(2022)