Search

The rapid growth of the Internet and widespread use of mobile computing and telephony has resulted in a broad range of ways to communicate and access information. Data and telecommunications convergence promises a wide range of possible solutions that will increase productivity, reduce costs, and provide new opportunities and revenues for enterprises. However, such converged telecommunications and data services have been largely isolated to static environments where fixed Personal Computers (PC) and an Internet connection are used in conjunction with various software tools to simulate a pseudo converged session. Generally, data presented on the internet and in enterprise applications is not available on voice networks and vice-versa. Due to the versatile nature of today's enterprise, a feature-rich, flexible, adaptive and widely accessible convergence solution is required. This paper presents an Interactive Voice Response (IVR) driven approach that uses Converged voice access to data (CVAD) services. The CVAD solution offers enhanced service effectiveness, flexibility and convenience for professionals on the move.

Until recently, most data interoperability techniques involved central components, e.g., global schemas or ontologies, to overcome semantic heterogeneity for enabling transparent access to heterogeneous data sources. Today, however, with the democratization of tools facilitating knowledge elicitation in machine-processable formats, one cannot rely on global, centralized schemas anymore as knowledge creation and consumption are getting more and more dynamic and decentralized. Peer Data Management Systems (PDMS) implementing semantic overlay networks are a good example of this new breed of systems eliminating the central semantic component and replacing it through decentralized processes of local schema alignment and query processing. As a result semantic interoperability becomes an emergent property of the system. In this talk we provide examples of both structural and dynamic aspects of such emergent semantics systems based on semantic overlay networks. ?From the structural perspective we can show that the typical properties of self-organizing networks also appear in semantic overlay networks. They form directed, scalefree graphs. We present both analytical models for characterizing those graphs and empirical results providing insight on their quantitative properties. Then we present semantic gossiping, a model for the dynamic reorganization of semantic overlay networks resulting from information propagation through the network and local realignment of semantic relationships. The techniques we apply in that context are based on belief propagation, a distributed probabilistic reasoning technique frequently encountered in self-organizing systems. Finally we will give a quick glance on how this techniques can be implemented at the systems level, based on a peer-to-peer systems approach.

Part 1: Artificial Intelligence Techniques in Bio Medicine; International audience; Accurately detecting psychological stress in time is a significant issue in the modern stressful society, especially for adolescents who are not mature enough to cope with pressure well. Micro-blog offers a new channel for teens’ stress detection, since more and more teenagers nowadays prefer to express themselves on the lively virtual social networks. Previous work mainly rely on tweeting contents to detect tweeters’ psychological stress. However, a tweet is limited to 140 characters, which are too short to provide enough information to accurately figure out its tweeter’s stress. To overcome the limitation, this paper proposes to leverage details of social interactions between tweeters and their following friends (i.e., time-sensitive comment/response actions under a tweet) to aid stress detection. Experimental results through a real user study show that time sensitivity of comment/response acts plays a significant role in stress detection, and involving such interaction acts can improve the detection performance by 23.5% in F-measure over that without such interactions.

Part 2: Artificial Intelligence for Knowledge Management; International audience; Companies today have several major issues while managing information. Many subsidiaries and departments have developed their own Data Management which has led to a multitude of Operational Databases and sometimes a multitude of Data Marts, policies and processes. Thus, these systems lack sustainability because they are not dynamic and not self-organizing, and so they do not adapt to the continuous needs arising from evolution that the companies experience. The Dynamic Data Mart architecture is built around 6 main functions, namely the 3Ms (Data Mining, Data Marshalling and Data Meshing) and the 3Rs (Recommendation, Reconciliation and Representation), which will address the aforementioned problems. Once the totality of the data have been loaded into a single Data Warehouse, the Dynamics Data Marts address these problems by mining the user’s behavior and the user’s decision making processes and continuously and automatically adapting the Data Mart to the needs of the users. Dynamic Data Marts create adapted dimensions, facts, data associations and views and then automatically find the ones that are not used anymore. These latter are then automatically dropped by the system, or can be presented to the IT manager if needed for validation of their removal.

Many applications take advantage of both ontologies and the Linked Data paradigm to characterize various kinds of resources. To fully exploit this knowledge, measures are used to estimate the relatedness of resources regarding their semantic characterization. Such semantic measures mainly focus on specific aspects of the semantic characterization (e.g. types) or only partially exploit the semantics expressed in the knowledge base. This article presents a framework for defining semantic measures to compare instances defined within an RDF knowledge base. A special type of measure, based on the representation of an instance through projections, is detailed and evaluated through its use in a music band recommender system.

International audience; In the Trusted Architecture for Securely Shared Services (TAS3) EC FP7 project we have developed a method to provide semantic support to the process modeler during the design of secure business process models. Its supporting tool, called Knowledge Annotator (KA), is using ontology-based data matching algorithms and strategy in order to infer the recommendations the best fitted to the user design intent, from a dedicated knowledge base. The paper illustrates how the strategy is used to perform the similarity (matching) check in order to retrieve the best design recommendation. We select the security and privacy domain for trust policy specification for the concept illustration. Finally, the paper discusses the evaluation of the results using the Ontology-based Data Matching Framework evaluation benchmark.

International audience; Artful processes are informal processes typically carried out by those people whose work is mental rather than physical (managers, professors, researchers, engineers, etc.), the so called “knowledge workers”. In this paper we propose the MailOfMine approach, to automatically build, on top of a collection of email messages, a set of workflow models that represent the artful processes laying behind the knowledge workers activities.

Section: Business process repositories; International audience; As organizations reach higher levels of Business Process Management maturity, they tend to accumulate large collections of process models. These repositories may contain thousands of activities and be managed by different stakeholders with varying skills and responsibilities. However, while being of great value, these repositories induce high management costs. Thus, it becomes essential to keep track of the various model versions as they may mutually overlap, supersede one another and evolve over time. We propose an innovative versioning model, and associated storage structure, specifically designed to maximize sharing across process models and process model versions, reduce conflicts in concurrent edits and automatically handle controlled change propagation. The focal point of this technique is to version single process model fragments, rather than entire process models. Indeed empirical evidence shows that real-life process model repositories have numerous duplicate fragments. Experiments on two industrial datasets confirm the usefulness of our technique.

The original publication is available at www.springerlink.com; International audience; Retrieving and composing individuals' capabilities are the matter of several research fields, like competence-based management, human resources management, enterprise knowledge management systems, knowledge representation systems, etc. This work focuses on capability representation, discovery and composition in heterogeneous, and possibly distributed, knowledge representation environments. We define a capability representation language in Description Logics, and we propose approaches and algorithms for capability management and discovery. In addition, we outline how our proposals are implemented in a mediator-based prototype system.

In this paper, we report the findings we gathered in designing and implementing a system that provides a distributed massive simulation environment. Massive Battle is a system capable of simulating historical battles for the purpose of learning and to carry out historical researches (e.g. what-if scenarios). We present a distributed implementation of Massive Battle and some early tests. We report and discuss some analysis of the problems related to the workload distribution in this particular environment. We report how is possible to measure a better load balancing by adopting a more general scheme of computation that generalize the assignments that each peer has to complete together with simulation.