Analysing the contents of social media platforms such as YouTube, Facebook and Twitter gained interest due to the vast number of users. One of the important tasks is homophobia/transphobia detection. This paper illustrates the system submitted by our team for the homophobia/transphobia detection in social media comments shared task. A machine learning-based model has been designed and various classification algorithms have been implemented for automatic detection of homophobia in YouTube comments. TF-IDF has been used with a range of bigram models for vectorization of comments. Support Vector Machines have been used to develop the proposed model and our submission reported 0.91, 0.92, 0.88 weighted F1-scores for English, Tamil and Tamil-English datasets respectively.

No abstract

Misinformation about COVID-19 overwhelmed our lives due to the tremendous usage of social media, especially Twitter. Spreading misinformation caused fear and panic among people affecting the national economic security of many countries. Vaccination is the crucial key to limiting the pandemic spread of COVID-19. Therefore, researchers start to detect and fight against the spread of misinformation taking it as a new challenge. This paper illustrates a model for misinformation detection in Arabic tweets using Natural Language Processing (NLP) techniques. A machine learning-based system has been developed regarding COVID-19 vaccination tweets. Term Frequency-Inverse Document Frequency (TF-IDF) has been used as vector space model for feature extraction. Support Vector Machines classification algorithm has been used for implementation the proposed system. Evaluation of the system, using different metrics, has been implemented on Arcov-19Vac, a dataset of Arabic tweets related to COVID-19 vaccination. The results reported by the illustrated model show that the performance of our model is promising.

This paper describes the systems submitted to iSarcasm shared task. The aim of iSarcasm is to identify the sarcastic contents in Arabic and English text. Our team participated in iSarcasm for the Arabic language. A multi-Layer machine learning based model has been submitted for Arabic sarcasm detection. In this model, a vector space TF-IDF has been used as for feature representation. The submitted system is simple and does not need any external resources. The test results show encouraging results.

No abstract