Social media occupies an important place in people’s daily lives where users share various contents and topics such as thoughts, experiences, events and feelings. The massive use of social media has led to the generation of huge volumes of data. These data constitute a treasure trove, allowing the extraction of high volumes of relevant information particularly by involving deep learning techniques. Based on this context, various research studies have been carried out with the aim of studying the detection of mental disorders, notably depression and anxiety, through the analysis of data extracted from the Twitter platform. However, although these studies were able to achieve very satisfactory results, they nevertheless relied mainly on binary classification models by treating each mental disorder separately. Indeed, it would be better if we managed to develop systems capable of dealing with several mental disorders at the same time. To address this point, we propose a well-defined methodology involving the use of deep learning to develop effective multi-class models for detecting both depression and anxiety disorders through the analysis of tweets. The idea consists in testing a large number of deep learning models ranging from simple to hybrid variants to examine their strengths and weaknesses. Moreover, we involve the grid search technique to help find suitable values for the learning rate hyper-parameter due to its importance in training models. Our work is validated through several experiments and comparisons by considering various datasets and other binary classification models. The aim is to show the effectiveness of both the assumptions used to collect the data and the use of multi-class models rather than binary class models. Overall, the results obtained are satisfactory and very competitive compared to related works.

Currently, advanced technological hardware can offer mobile devices which fits in the hand with a capacity to consult documents at anytime and anywhere. Multiple user context constraints as well as mobile device capabilities may involve the adaptation of multimedia content. In this article, the authors propose a new graph-based method for adapting multimedia documents in complex situations. Each contextual situation could correspond to a physical handicap and therefore triggers an adaptation action using ontological reasoning. Consequently, when several contextual situations are identified, this leads to multiple disabilities and may give rise to inconsistency between triggered actions. Their method allows modeling relations between adaptation-actions to select the compatible triggerable ones. In order to evaluate the feasibility and the performance of their proposal, an experimental study has been made on some real scenarios. When tested and compared with some existing approaches, their proposal showed improvements according to various criteria.

In pervasive systems, contextual constraints can hinder the smooth execution of multimedia documents. Several adaptation approaches have been proposed that operate as client-side, server-side, proxy-based or peer-to-peer applications. Most of these approaches rely on centralized mechanisms that adapt content at a single point. While centralized mechanisms offer several advantages, they also introduce significant drawbacks related to resource overhead. In this work, we propose a novel hybrid approach for multimedia document adaptation that efficiently combines multiple adaptation categories, such as client-side, server-side, proxy-based, and peer-to-peer, to alleviate some of their limitations. The main objective is to improve the quality of service in adaptation processes. The proposed process can operate on different devices, enabling self-reconfiguration by seamlessly switching from one adaptation category to another, depending on the computation context of the users, i.e., the device capabilities. To validate our proposal, we focus on three key aspects: first, real-world prototyping applied to the adaptation of multimedia content for e-learning materials; second, an assessment of performance in terms of response time, considering the number of adaptation requests and the available computing resources of each device; and third, a comparison with related work. The results show that as the number of adaptation requests increases, our approach outperforms traditional methods that rely on a single location. Specifically, it achieves response time reductions of up to 53.3% over the client/proxy/server hybrid scheme, 66.6% over the client/proxy hybrid scheme, and 74.3% over the client-only scheme. These findings are highly satisfying and encouraging, demonstrating that distributed computation can effectively optimize resource consumption, reduce response time, and maintain high scalability and robustness.