Improving the performance of aspect based sentiment analysis using fine-tuned Bert Base Uncased model

Geetha, M.; Renuka, D. Karthika

doi:10.1016/j.ijin.2021.06.005

Cited by 55 publications

(22 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…We applied BERT Tokenizer based on WordPiece (Muller et al, 2019 ) for the title, and abstracts from the LitCovid corpus. We used a pre-trained model bert-base-uncased (Geetha & Renuka, 2021 ) and the pre-training was performed on a large corpus of English data (BookCorpus and English Wikipedia) in a self-supervised fashion (Geetha & Renuka, 2021 ; Turc et al, 2019 ).…”

Section: Methodsmentioning

confidence: 99%

Multi-class classification of COVID-19 documents using machine learning algorithms

Rabby

Berka

2022

J Intell Inf Syst

View full text Add to dashboard Cite

In most biomedical research paper corpus, document classification is a crucial task. Even due to the global epidemic, it is a crucial task for researchers across a variety of fields to figure out the relevant scientific research papers accurately and quickly from a flood of biomedical research papers. It can also assist learners or researchers in assigning a research paper to an appropriate category and also help to find the relevant research paper within a very short time. A biomedical document classifier needs to be designed differently to go beyond a “general” text classifier because it’s not dependent only on the text itself (i.e. on titles and abstracts) but can also utilize other information like entities extracted using some medical taxonomies or bibliometric data. The main objective of this research was to find out the type of information or features and representation method creates influence the biomedical document classification task. For this reason, we run several experiments on conventional text classification methods with different kinds of features extracted from the titles, abstracts, and bibliometric data. These procedures include data cleaning, feature engineering, and multi-class classification. Eleven different variants of input data tables were created and analyzed using ten machine learning algorithms. We also evaluate the data efficiency and interpretability of these models as essential features of any biomedical research paper classification system for handling specifically the COVID-19 related health crisis. Our major findings are that TF-IDF representations outperform the entity extraction methods and the abstract itself provides sufficient information for correct classification. Out of the used machine learning algorithms, the best performance over various forms of document representation was achieved by Random Forest and Neural Network (BERT). Our results lead to a concrete guideline for practitioners on biomedical document classification.

show abstract

Section: Methodsmentioning

confidence: 99%

Multi-class classification of COVID-19 documents using machine learning algorithms

Rabby

Berka

2022

J Intell Inf Syst

View full text Add to dashboard Cite

show abstract

“…Sedangkan bayes berasal dari prinsip teorema bayes [20]. Keuntungan menggunakan naïve bayes untuk mengklasifikasikan salah satunya karena untuk mengklasifikasikan menggunakan jumlah data pelatihan yang kecil [21].…”

Section: Pendahuluanunclassified

Analisis Sentimen terhadap Opini Feminisme Menggunakan Metode Naive Bayes

Wahyuni¹

2022

INFEB

View full text Add to dashboard Cite

The perspective of the development of feminism centered on women around the world who wants to be free from pressure, oppression and inequality from men, continues to this day. Various public opinions about feminism are now contained in various social media. Long debates about criticism and support for feminism in equalizing women's position both in terms of intellect, and the role of women in making decisions. This research was conducted with the aim of looking at public sentiment based on opinions circulating on social media. Hashtags or hash tags related to feminism from social media are the main data that will be used to analyze public opinion sentiment about feminism and 600 data are obtained about feminism. The data obtained were separated into positive, negative and neutral opinions for analysis using Naïve Bayes (NB). The results of using the Naïve Bayes method obtained a recall value of 84%, precision 94% and Fi-Score of 86% with an accuracy of 88%. Through this research, the results of classification using the nave Bayes method in analyzing sentiment against feminist opinions have good performance.

show abstract

“…They used clinical BERT that had already been trained for embedding using LSTM, and the results were subpar. Additionally, a potent Deep Learning Model called BERT Base Uncased is offered in the study effort [20] to clarify the problem of sentiment analysis. In the experimental assessment, the BERT model outperformed the other machine learning techniques with a good prediction and high accuracy.…”

Section: Related Workmentioning

confidence: 99%

Opinions for Receiving COVID-19 Vaccines Based on Sentiment Analysis

2022

pnr

View full text Add to dashboard Cite

Due to vaccination hesitancy/refusal, a call to action is urgently needed for the COVID-19 vaccine deployment. Find out what the public thinks about the COVID-19 vaccine. Sentiment analysis is a technique for getting text sentiment scores. Therefore, we proposed architecture to analyze the textual data collection of people's opinions on COVID-19 vaccines using two of the best sentiment analysis techniques, the Bidirectional Encoder Representations from Transformers (BERT) technique and the Valence Aware Dictionary for sEntiment Reasoning (VADER) technique of Natural Language Processing (NLP). A questionnaire survey of corona vaccines recipients who recommend COVID-19 collected the data. Finally, recommendations for the corona vaccine were investigated, and various studies were done to determine its efficacy. Accuracy, precision, recall, and f1-score are standard evaluation criteria. The data shows the proposed model's excellent sentiment analysis performance, indicating that most vaccine users prefer to recommend others to get the vaccines.

show abstract

Improving the performance of aspect based sentiment analysis using fine-tuned Bert Base Uncased model

Cited by 55 publications

References 8 publications

Multi-class classification of COVID-19 documents using machine learning algorithms

Multi-class classification of COVID-19 documents using machine learning algorithms

Analisis Sentimen terhadap Opini Feminisme Menggunakan Metode Naive Bayes

Opinions for Receiving COVID-19 Vaccines Based on Sentiment Analysis

Contact Info

Product

Resources

About