Texts have become an essential spatial data resource in recent years. One leading task to manage the texts’ conveyed spatial data is the Spatial role labeling (SpRL). The latter intends the extraction of formal spatial knowledge from text. Instead of treating the text as a straightforward sequence of words, we incorporate syntactic dependencies to identify entities uttering spatial semantics. First, we investigate the impact of dependency-based word embeddings in SpRL. Then, we propose a dependency-guided LSTM-CRF deep learning model to exploit syntactic relationships among words. Then, we enhance these dependencies features with POS tags and CNN-based character-level representations. Experiments are performed on the standard SpRL-2012 and SpRL-2013 datasets. The experimental results show that the proposed model outperforms other machine learning approaches. The findings show the importance of taking dependencies into account in SpRL tasks.

In recent years, there has been a growing interest in the digitalization of medical records and health data. This makes it possible to build sophisticated systems that facilitate rapid data retrieval, processing and analysis. An example of such systems is Question-Answer ones. It can help domain experts efficiently retrieve the required data by asking questions in natural language. Question-answering systems based on Knowledge Graphs have found applications in various fields. Our work aims to build a question-answering chatbot for electronic medical records. Thus, healthcare professionals can ask questions in natural language, and the chatbot will iteratively build and execute appropriate SPARQL queries over a given Knowledge Graph containing digital medical records to return relevant data.

The Internet of Things (IoT) has connected the physical and digital worlds, resulting in a pervasive and self-organized network. Yet, due to the configurable model's behavioral constraints, ambiguity emerges during the design stage while selecting the most relevant composition plan. The constraints placed on the composition plan model restrict its behavior, leading to a single desired composition plan while eliminating the undesired ones. In this paper, we offer an entropy-based uncertainty metric to assess the predictability and efficiency of IoT object composition plans. Our method assesses the degree of uncertainty associated with a specific composition plan and provides insights into optimizing the performance of the IoT system.

The Web of Things (WoT) connects physical devices to the web using standard protocols. However, uncertainty in WoT service compositions may lead to critical problems in real- world applications. For example, in a smart hotel system, a user wishing to control the room temperature based on data from multiple sensors may get incorrect service compositions when using inaccurate or incomplete data, which can be dangerous in safety-critical situations. To address this issue, we propose a probabilistic approach that represents uncertain WoT services using Thing Description (TD) by including Quality of Thing (QoT) properties, interactions, and behaviors. Our approach computes the uncertainty of each node used in the composition process and proposes, using a WoT probabilistic algebra, an HTTP GET method that takes into account uncertain input and calculates the confidence degree of outputs when invoking WoT services. Our approach can improve the reliability of WoT services in uncertain environments and can be applied in various domains to create more robust and safer applications.

Commonly spatial data are stored in the Geographic Database (GDB) which is the backbone of Geographic information System (GIS). Setting up such database is a tedious and cost task. Moreover, the overwhelming amount of spatial information, particularly in the textual data is growing continuously. Hence, efforts were devoted to extract such data convoyed in the text streams freely available. In this context, we address this issue by proposing a novel hybrid approach that combines several Natural Language Processing (NLP) techniques, rules, and gazetteers to extract spatial named entities and retrieve relationships amongst them.

The widespread availability of tabular datasets on the internet has facilitated the adoption of technologies for information retrieval and prediction of missing or obscured data. However, a significant proportion of these datasets suffer from poor quality issues, such as misspelled or absent values, incomplete metadata. Furthermore, the manual collection and integration of data from various sources heavily rely on expert users’ knowledge, making the process laborious and time-consuming. Hence, there is a pressing need to streamline the collection and linking of data from multiple sources, ensuring that datasets are readily prepared for the intended analysis. This article introduces an approach that addresses these challenges by offering several key functionalities. Firstly, it facilitates data annotation to address concerns related to misspelled and incomplete metadata. Secondly, it enables data repair to handle missing values within the dataset. Lastly, it provides data augmentation capabilities, allowing the dynamic addition of meaningful columns and their corresponding cell values. The effectiveness of this approach has been evaluated using benchmark datasets with promising results in terms of evaluation metrics.

Résumé—Every day, road accidents cause human and material losses and become one of the world’s most serious problems. The National Observatory for Road Safety in Tunisia (NORS) has reported that the number of accidents has significantly increased year over year. The severity of the current road safety situation in Tunisia is highlighted by the increase in accidents from 5, 089 in 2021 to 5, 715 in 2022. In order to fight this phenomenon and improve the road safety situation in Tunisia, we propose a road classification approach based on road traffic severity analysis. Our proposal offers a new methodology for classifying roads according to their dangerousness based on the accident history. Our goal is to help NORS to orient their actions and improve road safety.

Since it is widely conveyed on the Web, tabular data is mostly organized in a table structure. Indeed, this data format is widely used in multiple scenarios on the web and even within data storage entities. In addition, tabular data is a source of information that deserves to be interpreted and exploited. In this framework, efforts to extract meaningful information from tabular data based on semantic approaches, such as an ontology or a knowledge graph, are commonly referred to as Semantic Table Interpretation (STI). In this paper, we present an interactive tool that deploys a mapping approach to overcome possible semantic gaps in tabular data against a Knowledge Graph. Indeed, the ultimate goal of this tool is to provide an application and adaptable framework that combines search and filtering services associated with text preprocessing techniques. The experimental evaluation was conducted under the SemTab challenge and yielded encouraging and promising results regarding its performance and rankings.

Road insecurity is a real scourge; road accidents cause more and more human and material losses worldwide, however, inequalities in the distribution of damages. On the one hand, a report published by the WHO estimated that some 1.3 million people are killed and 50 million injured in road accidents worldwide each year. On the other hand, a report published by the National Observatory of Road Safety in Tunisia showed that 1007 people were killed and 6757 injured in Tunisia in 202. Certainly, the statistics allow estimating the level of road insecurity at the global and national levels. To improve the current situation of road safety in Tunisia we propose a new mobile crowdsourcing platform called R-safety which allows road users to report in real-time their road safety feeling indexes. It also allows users to report violations they witness and interact with the NORS. R-safety also aims to inform citizens about risk areas through a vulnerability map that divides the areas into five types from the least dangerous to the most dangerous. In our approach, we are interested in improving the current situation of road safety by integrating the opinions of road users in the decision support tool of the national observatory of road safety. Thus, integrating citizens opinions into the information system would serve to have new sources of incoming data to improve decisions. This research work responds to real needs discussed with the members responsible for the studies and research of the NORS. We are evaluating our proposal through a questionnaire with 524 participants. This questionnaire helps us to choose among the proposed safety indicators those that affect road safety according to the citizen's opinion. This result confirms the importance of our crowdsourcing service.

The number of accidents is increasing over time, which has a social and economic impact. According to the World Health Organization (WHO) road accidents are now the main cause of death among 5-29 year old’s. In addition, Every day, roads cause roughly 40 serious and minor injuries and 4 fatalities in Tunisia, warns the National Observatory of Road Safety(NORS). The statistics show the seriousness of the current situation of road safety at the global and national level. In order to enhance road safety in Tunisia we propose a new system baptized R-Secure based on data from our crowdsourcing platforms. Our system has two functionalities, the first one consists in collecting rich road anomaly data such as images and GPS location from citizens and road safety experts. And the second is to analyze the information gathered and to inform resident’s about the risk areas through a vulnerability map of the city.

In this article, we present Kepler-aSI, a matching approach to overcome possible semantic gaps in tabular data by referring to a Knowledge Graph. The task proves difficult for the machines, which requires extra effort to deploy the cognitive ability in the matching methods. Indeed, the ultimate goal of our new method is to implement a fast and efficient approach to annotate tabular data with features from a Knowledge Graph. The approach combines search and filter services combined with text pre-processing techniques. The experimental evaluation was conducted in the context of the SemTab 2021 challenge and yielded encouraging and promising results referring to its performance and ranks held.

Human activity recognition has increasingly received attention in recent years to track regular activities of people. Existing activity recognition approaches considerably contributed to the analysis of human behavior. However, they still confront numerous issues related to the variability of activities performed by people within dynamic environments. Generally, this variability renders the used training or ontology models with predefined activities unsuitable. Therefore, creating an activity recognition approach that is able to leverage dynamically new and unknown activities at runtime becomes important. In this paper, we propose a novel knowledge-driven activity recognition framework using smartphone. This framework envisions taking a knowledge-driven approach to reinforce the recognition accuracy and people's quality of life in the context of dynamic environments at runtime. More specifically, we propose an ontology-based context evolution along with a dynamic decision-making, so that new and unknown performed activities can be accurately recognized. Furthermore, we use a public activity recognition dataset to demonstrate the effectiveness of the proposed framework and show its advantage over a data-driven baselines in terms of accuracy. Experimental results reveal that our framework not only reinforces the accuracy, but also enables an effective activity learning when facing unknown activities at runtime.

Language model pretraining has yielded significant results in diverse natural language processing tasks. RoberTa, an efficient method for pretraining self-supervised NLP systems, is a good example. Our hypothesis in this paper is that the performance of Spatial Role Labeling (SpRL) can be improved by combining static word vectors and bags of features with RoberTa vectors. Furthermore, we show that our method is successful in several SpRL datasets.

Summary In the era of Internet Technology (IT), uncertainty management is a challenge in many fields. These include e‐commerce, social and sensor networks, scientific data production and mining, object tracking, data integration, geo‐located services, and recently Internet and Web of Things. Due to the uncertain data published on the web, web resources are diverse. Hence, identical resources could be available from heterogeneous platforms and heterogeneous resources could represent the same objects. These resources are hugely heterogeneous, conflict, inconsistent, or have incompatible formats. This uncertainty is inherently related to many facts, such as information extraction and integration. Hence, with web resources proliferation on the web, referencing through the uncertain web has become increasingly difficult. The traditional techniques used for the classical web could not handle uncertain navigation. Generally, it's implicitly represented, decided randomly, or even neglected. Harnessing these uncertain resources to their full potential in order to handle the uncertain navigation, raises major challenges that relate to each phase of their life cycle: creation, representation, and navigation. In this article, we establish a probabilistic approach to model and interpret uncertain web resources. We present operators to compute response uncertainty. Finally, we create algorithms in order to validate resources and achieve uncertain hypertext navigation.

In several real-world applications, extracting spatial semantics from text is critical. Spatial Role Labeling (SpRL) introduces a language-independent annotation scheme used in these applications, particularly for reasoning purposes. This paper proposes, first of all, a transfer learning method with a word embeddings-based approach for SpRL. Then, we enhance the word vectors with POS tags and CNN-based character-level representations. Finally, we propose a Residual BiLSTM CRF deep learning model to identify the spatial roles. The experimental results on two datasets: SemEval-2012 and SemEval-2013 Task 3, show that the proposed model outperforms other machine learning approaches.

In this era of big data, as relational databases are inefficient, NoSQL databases are a workable solution for data storage. In this context, one of the key issues is the veracity and therefore the data quality. Indeed, as with classic data, geospatial big data are generally fuzzy even though they are stored as crisp data (perfect data). Hence, if data are geospatial and fuzzy, additional complexities appear because of the complex syntax and semantic features of such data. The NoSQL databases do not offer strict data consistency. Therefore, new challenges are needed to be overcome to develop efficient methods that simultaneously ensure the performance and the consistency in storing fuzzy geospatial big data. This paper presents a new methodology that tackles the storage issues and validates the fuzzy spatial entities' consistency in a document-based NoSQL system. Consequently, first, to better express the structure of fuzzy geospatial data in such a system, we present a logical model called Fuzzy GeoJSON schema. Second, for consistent storage, we implement a schema-driven pipeline based on the Fuzzy GeoJSON schema and semantic constraints.

International audience

Internet of Things (IoT) and Blockchain (BC) is an innovative paradigm that is gaining ground in smart environments. In intelligent classrooms, IoT makes our exchange easier with the prominent advent of smart devices, connected objects and sensors. However, the important research direction in this kind of IoT-Based Ubiquitous Learning Environment (ULE) is security and privacy that remaining essential challenges. Previously, we exposed the initial architecture of ULE based on BC technology and the educational services that can be delivered via this platform. In this paper, we investigate deeper and we highlight the main component of our ULE known as integrated IoT-ubiquitous platform using BC. The collection of data exchanged across devices is determined by the miner that preserves security using transactions that trace communications. Finally, in this study we demonstrate our preliminary experimental results that show the effectiveness of the proposed decentralized platform which is more secure by analyzing confidentiality, integrity, and availability.

Real-time spatial applications have become more and more important. These applications are innovating in the way we live. As a result, there is a huge amount of real-time spatial data generated everyday, accessed simultaneously by two types of transactions, Update transactions and User transactions (continuous requests). In real-time spatial Big Data, the performance can be increased by allowing concurrent execution of transactions. This activity is called concurrency control. The concurrency control algorithm must be used to ensure serializability of transaction scheduling and to maintain data consistency.Several works have been done in this area, but without holding into account the existence of a huge volume of data. In this paper, we apply the technique of imprecise computation for real-time spatial nested transactions. We consider that imprecise transaction consist that a User transaction is decomposed logically into a one mandatory sub-transaction and one or more optional sub-transactions and an Update transaction consists only of a single mandatory sub-transaction. We propose an improvement of an existing Two-Shadow Speculative Concurrency Control (SCC-2S) with priority with the use of the imprecise real-time spatial transaction. Our main objectives are: to guarantee the data freshness, to enhance the deadline miss ratio even in the presence of conflicts and unpredictable workloads and finally to satisfy the requirements of users by the improving of the quality of service (QoS).

A gradual evolution of the Internet, allowing it to extend beyond the electronic world to the physical world by interconnecting various devices and sensors that can communicate with each other and share data. The field of healthcare monitoring Systems have experienced significant changes using this promising new technology, called Internet of Things (IoT). In this paper, we propose a new platform for health care, which we have called the 'Semantic Medical IoT platform'. This platform is designed primarily for the semantization of the Internet of things in the medical and healthcare field. It proposes solutions for the problems of the interoperability of medical devices, the integration of massive and heterogeneous data, and the personalized visualization of these data. The platform also provides various services, like: Multi-type functional communication service, the significant exploitation of the localization feature provided by the connected objects, the simplification of medical texts for the patients and the effective integration of the social media technology. The SM-IoT Platform also defines new contract-based security policies to ensure the confidentiality of patient's health information.

Today, the event detection activity is becoming increasingly more complex with the emergence of the Big Data phenomenon. More specifically, different social networks and microblogging services are leading to the massive explosion of data describing the real world. Multiple conducted works highlight the role of textual data as the simplest form to report actualities and to describe real-time events. But, the variety of the proposed solutions with high complexity degrees reveals big challenges that are hidden behind the nature of those data. A concentrated study around some established works for event detection over textual data published on social networks allows us to meet limits of proposed solutions. These limits are the main areas of research leading to the proposal of a new approach for a real-time localization and detection of events over textual data posted on social networks. This approach takes benefit from a set of Apache platforms for efficient treatment of the huge streams of data.

Ubiquitous learning environments have an increasing trend considering the huge number of connected smart devices dedicated to educational services. Ubiquitous learning (U-learning) provides to students the possibility to learn at anyplace and anytime within the collaborative environment using interactive multimedia system that enables effective communication among teacher and learners. The architecture of ubiquitous learning environment (ULE) still suffers from the problems of vulnerability. Blockchain (BC) technology recently explored to provide much more privacy and security using essentially peer-to-peer (P2P) networks has a significant role in the development of decentralized topologies. This paper expounds a novel BC-based architecture for ULE that preserves the benefits of security and privacy. The architecture offers new opportunities to design secured collaborative learning system. It provides data exchange with BC using trust methods within the decentralized topology. The evaluation of this implemented system demonstrates its efficiency while it delivers security and privacy for ULE.

Today, numerous technological advances in electronic and information technologies are rapidly transforming our modern life. Industries are being completely transformed and health systems are not oblivious to these changes. All these new technologies are opening up a wide range of new opportunities and challenges for researchers, physicians and patients. The emergent paradigm of Internet of Things is one of the promising technologies introduced into the world of health care. In fact, IoT has been widely applied to interconnect available medical devices and sensors which allow patients to take control of their health condition in real time, also physicians to accurately remotely monitor the health of their patients. The ultimate goal of achieving high quality of healthcare practices depends on the ability to effectively integrate data incoming from heterogeneous sources, share the collected data while keeping their security and privacy, use powerful data analytics tools to extract useful information from these data, and the ability to have an expressive and personalized visualization. In this paper, we propose the SM-IoT platform, an IoT-based platform for intelligent and personalized healthcare, dedicated to patients, as well as caregivers. The aim of this platform is to improve the remote patient monitoring and promote healthcare services. SM-IoT platform is able to collect data from heterogeneous information sources, integrate them by using a flexible semantic web, store them in the cloud for further analysis, visualized these data with user-friendly interfaces and facilitate their sharing by taking into account their privacy aspect.

Microblogs such as Twitter are characterized by the richness and recency of information shared by their users during major events. However, it is very challenging to automatically mine for information or for users sharing certain information due to the huge variety of unstructured stream of data shared in such microblogs. This work proposes a ranking and classification model for identifying users sharing useful information during a specified event. The model is based on a novel set of features that can be computed in real time. These features are designed such that they take into account both the on and off-topic activities of a user. Once users are characterized by a feature vector, supervised machine learning tool is trained to classify users as either prominent or not. Our model has been tested on data shared during a flooding disaster event and performed very well. The achieved results show the effectiveness of the proposed model for both the classification and ranking of prominent users in such events, and also the importance of the adjustment of the on-topic features by the off-topic ones when describing users' activities.

During specific real-world events, some users of microblogging platforms could provide exclusive information about those events. The identification of such prominent users depends on several factors such as the freshness and the relevance of their shared information. This work proposes a probabilistic model for the identification of prominent users in microblogs during specific events. The model is based on learning and classifying user behavior over time using Mixture of Gaussians Hidden Markov Models. A user is characterized by a temporal sequence of feature vectors describing his activities. The features computed at each time-stamp are designed to reflect both the on- and off-topic activities of users, and they are computationally feasible in real-time.

Geographic Information System (GIS) is a computer system designed to capture, store, manipulate, analyze, manage, and present all types of spatial data. Spatial data, whether captured through remote sensors or large scale simulations becomes more and big and heterogenous. As a result, structured data and unstructured content are simultaneously accessed via an integrated user interface. The issue of real-time and heterogeneity is extremely important for taking effective decision. Thus, heterogeneous real-time spatial data management is a very active research domain nowadays. Existing research are interested in querying of real-time spatial data and their updates without taking into account the heterogeneity of real-time geospatial data. In this paper, we propose the use of the real-time Spatial Big Data and we define a new architecture called FCSA-RTSBD (Feedback Control Scheduling Architecture for Real-Time Spatial Big Data). The main objectives of this architecture are the following: take in account the heterogeneity of data, guarantee the data freshness, enhance the deadline miss ratio even in the presence of conflicts and finally satisfy the requirements of users by the improving of the quality of service (QoS).

The spatial character intrinsic to the routing field requires the integration of geographic information systems (GIS) and optimization approaches to handle spatial and non-spatial data in transportation applications. Motivated by the need to better support decision making in logistic area, we develop an interactive spatial decision support system (SDSS) for solving the vehicle routing problems by coupling the simulated annealing method with Quantum GIS (QGIS). In this paper, the evoked variants of VRP are detailed and formulated mathematically. The SDSS architecture is designed for the VRPs showing the interaction of GIS and SA approach according to the tight coupling strategy. A VRP variant termed the Open VRP (OVRP) is selected to show the system effectiveness. The computational performance of the SDSS for the OVRP, based on a set of benchmark instances, turned out to be effective on both computation time and solution quality.

Given the exponential growth of online information, one of the primary difficulties facing users is the information overload. Huge amounts of these electronic information are mainly encapsulated in text documents. So, the researchers are striving to develop robust methods to provide users and managers with prominent solutions needed for text analysis. This allows maximizing profits by saving time and money devoted to manage the increasing amount of information. In this context and among others we find text segmentation and theme identification. In this paper and by resting on the existing well known approaches and advances in the text segmentation and theme identification fields, we propose new solutions to reach the objectives behind achieving these two techniques more accurately.

Knowledge discovery in databases is used to discover useful and understandable knowledge from large databases. A process of knowledge discovery consists of two steps, the data mining step and the evaluation step. In this paper, evaluating and ranking the interestingness of summaries generated from databases, which is a part of the second step, is studied using diversity measures. Sixteen previously analyzed diversity measures of interestingness are used along with three not previously considered ones, brought from different well-known areas. The latter three measures are evaluated theoretically according to five principles that a measure must satisfy to be qualified acceptable for ranking summaries. A theoretical correlation study between the eight measures that satisfy all five principles is presented based on mathematical proofs. An empirical evaluation is conducted using three real databases. Then, a classification of the eight measures is deduced. The resulting classification is used to reduce the number of measures to only two, which are the best over all criteria, and that produce non-similar results. This helps the user interpret the most important discovered knowledge in his decision making process.

The paper presents a methodology for e-commerce site conception, made up of three fundamental parts: a market analysis allowing to better learn the site's visitors profile in order to provide them with the appropriate customer experience, a benchmarking aimed at identifying the best practises among the best-in-class sites selling similar products and the site architecture part, which takes into account the findings of the previous phases in order to provide the users with tailored functionalities. The conceptual approach was applied to the case of the merchant site of AtlasGeoLearn/spl copy/ an electronic encyclopaedia focused on culture and tourism in Tunisia.

Data warehouses have appeared in answer to the new requirements of database technologies. With them appeared some automated techniques like knowledge discovery in databases (KDD), data mining (DM) and online analytical processing (OLAP), which have registered a great demand in relational, transactional and finally in geographical databases. We give a survey of the state of the art of this research field. We mention the principles of data warehouses defining the context of geographical information, and present models used to build the data warehouse. Then, the tools permitting exploration of data warehouses and considered as tools for decision making are introduced. An evaluation of the technique of generalization is carried out using incendiary data. On the other hand, we expose a suggestion to manage incendiary data by cubic representation.