Abstract: In this paper, we shed light on the use of two types of fuzzy genetic algorithms, which 1 stand out from the literature for the innovative ideas behind them. One is the Gendered Fuzzy 2 Genetic Algorithm, where the crossover mechanism is regulated by the gender and the age of the 3 population to generate offspring through proper fuzzy rules. The other one is the Elegant Fuzzy 4 Genetic Algorithm, where the priority of the parent genome is updated based on the child’s fitness. 5 Both algorithms present a significant computational burden. To speed up the computation, we 6 propose to adopt a nearest-neighbour caching strategy. We performed several experiments using 7 first some well-known benchmark functions, trying different types of membership functions and 8 logical connectives. Afterwards, some additional benchmarks were retrieved from the literature for 9 a fair comparison against published results obtained by means of former variants of fuzzy genetic 10 algorithms. A real-world application problem, retrieved from the literature and dealing with rice 11 production, was also tackled. All the numerical results show the potential of the proposed strategy.

Keywords: Nearest neighbour, Caching, Fuzzy rules, Fitness, Priority

DOI: 10.3390/a17120549, Multidisciplinary Digital Publishing Institute

Article BibTeX

Abstract: The topic of persuasion in online conversations has social, political and security implications; as a consequence, the problem of predicting persuasive comments in online discussions is receiving increasing attention in the literature. Following recent advancements in graph neural networks, we analyze the impact of conversation structure in predicting persuasive comments in online discussions. We evaluate the performance of artificial intelligence models receiving as input graphs constructed on the top of online conversations sourced from the “Change My View” Reddit channel. We experiment with different graph architectures and compare the performance on Graph Neural Networks, as structure-based models, and Dense Neural Networks as baseline models. Experiments are conducted on two tasks: 1) Persuasive Comment Detection, aiming to predict which comments are persuasive, and 2) Influence Prediction, aiming to predict which users are persuasive. The experimental results show that the role of the conversation structure in predicting persuasiveness is strongly dependent on its graph representation given as input to the Graph Neural Network. In particular, a graph structure linking only comments belonging to the same speaker in the conversation achieves the best performance in both tasks. This structure outperforms both the baseline model, which does not consider any structural information, and structures linking different speakers’ comments with each other. Specifically, the F1 score of the best performing model is 0.58, which represents an improvement of 5.45% over the baseline model (F1 score of 0.55) and 7.41% over the model linking different speakers’ comments (F1 score of 0.54).

Keywords: Social Media Persuasion, Persuasive Comment Detection, Influence Prediction, Text Data

DOI: 10.1007/s12652-024-04841-8, Springer-Verlag GmbH

Article BibTeX

Abstract: In our interconnected world, the dissemination of misinformation has emerged as a crucial and pressing challenge. Social media platforms and technological advancements facilitate the proliferation of false information, thereby leading to significant repercussions on societal, political, and economic fronts. Recent research suggests using Graph Neural Networks (GNNs) to represent relationships among network actors and consequent prediction activities like pinpointing influential nodes and detecting communities. This work exploits a GNN to make link predictions on a graph representing information about misinformation tweets, their authors, and their spread. The objective is to comprehensively investigate the specific attributes of online pathways that compel users to share and amplify inaccurate information. In this sense, starting from an existing dataset of misinformation tweets, the proposed approach first applies an explainability method to each prediction, then, through frequent itemset mining, tries to detect patterns among collected explanations. Results of qualitative and quantitative research questions mainly demonstrate the contribution of interpersonal aspects to misinformation tweets spreading. To the best of our knowledge, this is the first approach exploiting a combination of Explainable Artificial Intelligence (xAI) and Data Mining to GNNs in fake news spreading analysis for prevention and mitigation purposes.

Keywords: Cognitive Security, xAI, GNN, Information Disorder

DOI: 10.1007/978-3-031-64779-6_13, Springer Nature AG

BibTeX

Abstract: Fake news, which can be defined as intentionally and verifiably false news, has a strong influence on critical aspects of our society. Manual fact-checking is a widely adopted approach used to counteract the negative effects of fake news spreading. However, manual fact-checking is not sufficient when analysing the huge volume of newly created information. Moreover, the number of labeled datasets is limited, humans are not particularly reliable labelers and databases are mostly in English and focused on political news. To solve these issues state-of-the-art machine learning models have been used to automatically identify fake news. However, the high amount of models and the heterogeneity of features used in literature often represents a boundary for researchers trying to improve model performances. For this reason, in this systematic review, a taxonomy of machine learning and deep learning models and features adopted in Content-Based Fake News Detection is proposed and their performance is compared over the analysed works. To our knowledge, our contribution is the first attempt at identifying, on average, the best-performing models and features over multiple datasets/topics tested in all the reviewed works. Finally, challenges and opportunities in this research field are described with the aim of indicating areas where further research is needed.

Keywords: Fake News Detection, Content Based Fake News Detection, Content-Based Features

DOI: 10.1016/j.neucom.2023.02.005, Elsevier Ltd.

BibTeX

Abstract: Student Performance Prediction (SPP) models and tools are useful for quickly identifying at-risk students in online courses and enable the provi-sion of personalized learning plans and assistance. Additionally, they give educators and course managers the information they need to recognize the programs that require improvement. High accuracy is essential for such tools but understanding the reasons of their predictions is equally im-portant to ensure fairness and build trust in their adoption. Although many SPP models and tools have been proposed so far by different researchers, very few of them take explainability into account. This research proposes an SPP approach that is both effective and explainable. Based on demo-graphic, administrative, engagement, and intra-course outcome data, it ena-bles the prediction of student performance in terms of success/failure and final grade. It supports multiple machine learning models and includes post-hoc techniques for explainability capable of justifying the behavior of the whole system as well as its individual predictions.

Keywords: Learning Analytics, Educational Data Mining, Student Performance Predic-tion, Explainable Artificial Intelligence

DOI: 10.1007/978-3-031-41637-8_52, Springer Nature AG

BibTeX