Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
2 results
Search Results
Item Restricted Crisis Detection from Arabic Social Media(University of Birmingham, 2023-09-12) Alharbi, Alaa; Lee, MarkSocial media (SM) streams such as Twitter provide large quantities of real-time information about emergency events from which valuable information can be extracted to enhance situational awareness and support humanitarian response efforts. The timely extraction of crisis-related SM messages is challenging as it involves processing large quantities of noisy data in real time. Supervised machine learning classifiers are challenged by out-of-distribution learning when classifying unseen (new) crises due to data variations across events. Besides that, it is impractical to label training data from each novel and emerging crisis since obtaining sufficient labelled data is time-consuming and labour-intensive. This thesis addresses the problem of Twitter crisis classification using supervised learning methods to identify crisis-related data and categorising them into different information types in the multi-source (training data from multiple events) setting. Due to Twitter’s ubiquity during emergency events in the Arab world, the current research focuses on Arabic Twitter content. We have created and published a large-scale Arabic Twitter corpus of crisis events. The corpus has been analysed and manually labelled. Analysing the content includes investigating the main information categories of conversations posted during a range of crisis events using natural language processing techniques. Building these resources is considered one of this thesis’s contributions. The thesis also investigates the generalisation performance of different supervised classical machine learning and deep learning approaches trained on out-of-crisis data to classify unseen crises. We find that deep neural networks such as LSTM and CNN outperform the classical machine learning classifiers such as support vector machines and decision trees. We also evaluate different architectures of deep neural networks and several pre-trained text representations (embeddings) learnt from vast amounts of unlabelled text. Results show that BERT-based models are more robust to out-of-distribution target events and remarkably outperform other models on the information classification task. Experiments show that the performance of BERT-based classifiers can be enhanced when training on similar data. Thus, the last contribution of the present study is to propose an instance distance-based data selection approach for adaptation to improve classifiers’ performance under a domain shift. Using the BERT embeddings, the method selects a subset of multi-event training data that is most similar to the target event. Results show that fine-tuning a BERT model on a selected subset of data to classify crisis tweets outperforms a model that has been fine-tuned on all available source data.15 0Item Restricted Information Integrity: From a Lens of Explainable AI With Cultural and Social Behaviors(2023-08-11) Alharbi, Raed; Thai, My TThe rapid development of Artificial intelligence (AI), such as machine learning (ML) and deep neural networks (DNNs), has changed the way information is processed and used. However, along with these advancements, challenges to information integrity have emerged. The widespread dissemination of misinformation through digital platforms, coupled with the lack of transparency in black-box ML models, has raised concerns about the reliability and trustworthiness of informa- tion to expert users (ML developers) and non-expert users (end-users). Unfortunately, employing eXplainable Artificial Intelligence (XAI) approaches on real-world applications to improve the trustworthiness of DNNs models is still far-fetched and not straightforward. Motivated by these observations, this thesis concentrates on two directions. • Misinformation Mitigation. In the first direction, we leverage XAI techniques to mitigate misinformation through three main approaches: evaluating the trustworthiness of fake news detection models from a user perspective, studying the influence of social and cultural behavior on misinformation propa- gation, and analyzing the diffusion of descriptive norms in social media networks to promote positive norms and combat misinformation. • Developing Advanced ML Models. In the second direction, we turn our attention to developing ML models from two aspects. The first aspect exploits XAI behaviors to provide a new method to simultaneously preserve the performance and explainability of student models, which in their primitive form provide little transparency. In the second aspect, we develop the Temporal graph Fake News Detec- tion Framework (T-FND), which effectively captures heterogeneous and repetitive charac- teristics of fake news behavior.22 0