The Internet of Things is a paradigm that interconnects several smart devices through the internet to provide ubiquitous services to users. This paradigm and Web 2.0 platforms generate countless amounts of textual data. Thus, a significant challenge in this context is automatically performing text classification. State-of-the-art outcomes have recently been obtained by employing language models trained from scratch on corpora made up from news online to handle text classification better. A language model that we can highlight is BERT (Bidirectional Encoder Representations from Transformers) and also DistilBERT is a pre-trained smaller general-purpose language representation model. In this context, through a case study, we propose performing the text classification task with two previously mentioned models for two languages (English and Brazilian Portuguese) in different datasets. The results show that DistilBERT’s training time for English and Brazilian Portuguese was about 45% faster than its larger counterpart, it was also 40% smaller, and preserves about 96% of language comprehension skills for balanced datasets.

In the traditional approach for centrality measures, also known as sociocentric, a network node usually requires global knowledge of the network topology in order to evaluate its importance. Therefore, it becomes difficult to deploy such an approach in large-scale or highly dynamic networks. For this reason, another concept known as egocentric has been introduced, which analyses the social environment surrounding individuals (through the ego-network). In other words, this type of network has the benefit of using only locally available knowledge of the topology to evaluate the importance of a node. It is worth emphasizing that in this approach, each network node will have a sub-optimal accuracy. However, such accuracy may be enough for a given purpose, for instance, the vehicle selection mechanism (VSM) that is applied to find, in a distributed fashion, the best-ranked vehicles in the network after each topology change. In order to confirm that egocentric measures can be a viable alternative for implementing a VSM, in particular, a case study was carried out to validate the effectiveness and viability of that mechanism for a distributed information management system. To this end, we used the egocentric betweenness measure as a selection mechanism of the most appropriate vehicle to carry out the tasks of information aggregation and knowledge generation. Based on the analysis of the performance results, it was confirmed that a VSM is extremely useful for VANET applications, and two major contributions of this mechanism can be highlighted: (i) reduction of bandwidth consumption; and (ii) overcoming the issue of highly dynamic topologies. Another contribution of this work is a thorough study by implementing and evaluating how well egocentric betweenness performs in comparison to the sociocentric measure in VANETs. Evaluation results show that the use of the egocentric betweenness measure in highly dynamic topologies has demonstrated a high degree of similarity compared to the sociocentric approach.

With the increase of vehicles in large urban centers, there is also an increase in the number of traffic jams and accidents on public roads. The development of a proper Intelligent Transport System (ITS) could help to alleviate these problems by assisting the drivers on route selections to avoid the most congested road sections. Therefore, to improve on this issue, this work proposes an architecture to aid an ITS to detect, analyze, and classify the traffic flow conditions in real time. This architecture also provides a control room dashboard to visualize the information and notify the users about the live traffic conditions. To this end, the proposed solution takes advantage of computer vision concepts to extract the maximum information about the roads to better assess and keep the drivers posted about the traffic conditions on selected highways. The main contribution of the proposed architecture is to perform the detection and classification of the flow of vehicles regardless of the luminosity conditions. In order to evaluate the efficiency of the proposed solution, a testbed was designed. The obtained results show that the accuracy of the traffic classification rate is up to 90% in daylight environments and up to 70% in low light environments when compared with the related literature.

Transport authorities are employing advanced traffic management system (ATMS) to improve vehicular traffic management efficiency. ATMS currently uses intelligent traffic lights and sensors distributed along the roads to achieve its goals. Furthermore, there are other promising technologies that can be applied more efficiently in place of the abovementioned ones, such as vehicular networks and 5G. In ATMS, the centralized approach to detect congestion and calculate alternative routes is one of the most adopted because of the difficulty of selecting the most appropriate vehicles in highly dynamic networks. The advantage of this approach is that it takes into consideration the scenario to its full extent at every execution. On the other hand, the distributed solution needs to previously segment the entire scenario to select the vehicles. Additionally, such solutions suggest alternative routes in a selfish fashion, which can lead to secondary congestions. These open issues have inspired the proposal of a distributed system of urban mobility management based on a collaborative approach in vehicular social networks (VSNs), named SOPHIA. The VSN paradigm has emerged from the integration of mobile communication devices and their social relationships in the vehicular environment. Therefore, social network analysis (SNA) and social network concepts (SNC) are two approaches that can be explored in VSNs. Our proposed solution adopts both SNA and SNC approaches for alternative route-planning in a collaborative way. Additionally, we used dynamic clustering to select the most appropriate vehicles in a distributed manner. Simulation results confirmed that the combined use of SNA, SNC, and dynamic clustering, in the vehicular environment, have great potential in increasing system scalability as well as improving urban mobility management efficiency.