Assessing the Impact of Contextual Information in Hate Speech Detection

Pérez, José Miguel Giner; Luque, Franco M.; Zayat, Demian; Kondratzky, Martín; Moro, Agustín; Serrati, Pablo; Zajac, Joaquín; Miguel, Paula; Debandi, Natalia; Gravano, Agustı́n; Cotik, Viviana

doi:10.1109/access.2023.3258973

Cited by 25 publications

(4 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, the Polyglot library can increase the accuracy rate by up to 1.19% compared to models that do not use NER. These results support the studies [22] and [19].…”

Section: Discussionsupporting

confidence: 92%

Comparison of SVM and Naïve Bayes Algorithms with InNER enriched to Predict Hate Speech

HADİ AL GHOZALİ,

PİRMAN,

INDRA

2023

ECJSE

View full text Add to dashboard Cite

Hate speech is one of the negative sides of social media abuse. Hate speech can be classified into insults, defamation, unpleasant acts, provoking, inciting, and spreading fake news (hoax). The purpose of this study is to compare the SVM and Naïve Bayes methods with feature extraction in the form of Indonesian NER (InNER) for detecting hate speech. To obtain the best model, this study applies five steps: a) data collection; b) data preprocessing; c) feature engineering; d) model development; and e) evaluating and comparing models. In this study, we have collected 7100 tweets as an initial dataset. After manual annotation, this study produced 1681 tweets: 548 insult tweets, 288 blasphemy tweets, 272 provocative tweets, and 573 neutral tweets. This study use two Python libraries that accommodate NER in Indonesian, namely the NLTK library and the Polyglot library. Based on the results of the evaluation of the proposed model, model 5, which develops the SVM algorithm with the NLTK library, is the best model proposed. This model shows an accuracy score of 92.88% with a precision of 0.93, a recall of 0.93, and an F-1 score of 0.92.

show abstract

“…Meanwhile, the Polyglot library can increase the accuracy rate by up to 1.19% compared to models that do not use NER. These results support the studies [22] and [19].…”

Section: Discussionsupporting

confidence: 92%

Comparison of SVM and Naïve Bayes Algorithms with InNER enriched to Predict Hate Speech

HADİ AL GHOZALİ,

PİRMAN,

INDRA

2023

ECJSE

View full text Add to dashboard Cite

show abstract

“…The journal accounts used for the labeling process are still global, while the variations in financial transactions are numerous and unique for each type of business. For example, financial transactions in trading and manufacturing businesses are much more complicated and complex than in service businesses [30].…”

Section: Resultsmentioning

confidence: 99%

Comparison of Classification Algorithm and Language Model in Accounting Financial Transaction Record: A Natural Language Processing Approach

Makayasa,

Siregar,

Sugiantoro

et al. 2024

Int. J. Adv. Sci. Eng. Inf. Technol.

View full text Add to dashboard Cite

The problem of financial recording not following the principles of accounting science has the potential to cause unnecessary problems. However, micro, small, and medium enterprises with their distinctive characteristics, though not all, still face many obstacles in writing financial reports. Even though there is already much financial software available, our study aims to investigate opportunities for implementing automation of accounting financial transaction records using the NLP approach, to interpret financial transactions based on text written on the transaction form into accounting journals (debits and credits). Experiments were carried out by comparing the performance of three classification algorithms, namely SVM, K-Nearest Neighbor, and Random Forest, with traditional (TF-IDF and BOW) and contextual (Word2Vec) Language Models. There are 200 financial transaction datasets consisting of ten classes. The data is divided into two parts, namely, the balance dataset and the imbalance dataset. The pair SVM and Word2Vec in the balanced dataset gave the highest accuracy (92.5%), precision (92.5%), recall/sensitivity (93.33%), and F1 score (92%). However, compared with the results of related semantic research (the average performance reaches 95%), the results obtained in this study are still lower. One point that may have a significant effect is the amount of data in the corpus, which is still lacking. Researchers suggest increasing the number of datasets and using a combination of other language models such as Glove, Bert etc. This study can also be used as a model for more complex financial transaction cases in future research.

show abstract

“…In addition to the tasks described in the main paper, we also provide tools for other related tasks or domains, such as contextualized hate speech detection (Pérez et al, 2023a) and targeted sentiment analysis (Pérez et al, 2023b), both models specialized in Rioplatense Spanish.…”

Section: Appendix a Other Considered Tasksmentioning

confidence: 99%

pysentimiento: A Python Toolkit for Opinion Mining and Social NLP tasks

Perez,

Rajngewerc,

Giudici

et al. 2023

Preprint

View full text Add to dashboard Cite

In recent years, the extraction of opinions and information from user-generated text has attracted a lot of interest, largely due to the unprecedented volume of content in Social Media. However, social researchers face some issues in adopting cutting-edge tools for these tasks, as they are usually behind commercial APIs, unavailable for other languages than English, or very complex to use for non-experts. To address these issues, we present pysentimiento, a comprehensive multilingual Python toolkit designed for opinion mining and other Social NLP tasks. This open-source library brings state-of-the-art models for Spanish, English, Italian, and Portuguese in an easy-to-use Python library, allowing researchers to leverage these techniques. We present a comprehensive assessment of performance for several pre-trained language models across a variety of tasks, languages, and datasets, including an evaluation of fairness in the results.

show abstract

Assessing the Impact of Contextual Information in Hate Speech Detection

Cited by 25 publications

References 47 publications

Comparison of SVM and Naïve Bayes Algorithms with InNER enriched to Predict Hate Speech

Comparison of SVM and Naïve Bayes Algorithms with InNER enriched to Predict Hate Speech

Comparison of Classification Algorithm and Language Model in Accounting Financial Transaction Record: A Natural Language Processing Approach

pysentimiento: A Python Toolkit for Opinion Mining and Social NLP tasks

Contact Info

Product

Resources

About