This paper presents an alternative event detection model based on the integration between the DistilBERT and a new meta-heuristic technique named the Hunger Games Search (HGS). The DistilBERT aims to extract features from the text dataset, while a binary version of HGS is developed as a feature selection (FS) approach, which aims to remove the irrelevant features from those extracted. To assess the developed model, a set of experiments are conducted using a set of real-world datasets. In addition, we compared the binary HGS with a set of well-known FS algorithms, as well as the state-of-the-art event detection models. The comparison results show that the proposed model is superior to other methods in terms of performance measures.

This paper presents a transformer-based pipeline for extracting and classifying events from social media, particularly Twitter. The proposed approach integrates Named Entity Recognition (NER) for event extraction, abstractive summarization to condense Twitter content, and multi- label classification for categorization. Our method demonstrates strong performance across various datasets, showcasing its robustness and scalability for real-world applications such as disaster management and trend monitoring. Social media platforms enable users to collaborate on ideas and organize events. However, recent incidents related to social media have raised widespread concerns. A thorough investigation was conducted to identify and address any alarming situations. This study primarily focuses on event detection, including disasters, traffic events, sports, real-time events, and others. These detected events can quickly reflect the overall state of society, making them particularly valuable for analyzing occurrences that pose a threat to social security. We observed that one of the most significant challenges for event detection algorithms is ensuring compatibility with various languages, spellings, and accents. Furthermore, event detection algorithms must be capable of processing diverse types of content, including text, images, videos, and geolocations. Our findings indicate that event detection algorithms capable of handling different data formats, languages, and platforms remain largely unavailable. With today's technology, everyone's online presence provides access to a massive pool of data that can be utilized for a variety of purposes, ranging from assessing market patterns to comprehending a population's overall emotional condition. Text and sentiment analysis is lot easier than it was a few years ago, thanks to technological advancements and natural language processing tools. Each tweet's label indicates a different type of disaster-related data that may be used in a number of ways during an emergency response. When someone tweets a warning about an oncoming catastrophe or tragedy, and our BERT models detect it instantly, we may respond as soon as possible, perhaps saving lives. BERT is a free natural language processing (NLP) machine learning tool.