Annabelle Mercier

An information retrieval model using the fuzzy proximity degree of term occurences

Based on the idea that the closer the query terms in a document are, the more relevant this document is, we propose a mathematical model of information retrieval based on a fuzzy proximity degree of term occurences. Our model is able to deal with Boolean queries, but contrary to the traditional extensions of the basic Boolean information retrieval model, it does not explicitly use a proximity operator. A single parameter allows to control the proximity degree required. With conjunctive queries, setting this parameter to low values requires a proximity at the phrase level and with high values, the required proximity can continuously be relaxed to the sentence or paragraph levels. We conducted some experiments and present the results.

ENSM-SE at INEX 2009: scoring with proximity and semantic tag information

We present in this paper some experiments on the Wikipedia collection used in the INEX 2009 evaluation campaign with an information retrieval method based on proximity. The idea of the method is to assign to each position in the document a fuzzy proximity value depending on its closeness to the surrounding keywords. These proximity values can then be summed on any range of text - including any passage or any element - and after normalization this sum is used as the relevance score for the extent. To take into account the semantic tags, we define a contextual operator which allow to consider at query time only the occurrences of terms that appear in a given semantic context.

A Method to Analyze, Diagnose and Propose Innovations for Complex Ecosystems: the InnoServ Project

Understanding and modeling complex ecosystems, where a great number of entities interact in different ways, is a great challenge in the information systems’ domain. In this context, the InnoServ project aims to understand and support innovations around fragile people considering public, private and volunteering structures. The aim of this paper is to present the ADInnov method, which facilitates the analysis, the diagnosis and the proposition of innovations for complex ecosystems. This method has been extracted in an empirically way, from the lessons learned in the InnoServ project combining different techniques such as expert interviews, goal modeling and serious games. This method could be used in other areas where it is necessary to analyze complex ecosystems. Drawing out and discussing the results of the InnoServ project, we prove the efficiency of our method.

Doubt Removal for Dependant People from Tablet Computer Usage Monitoring

This article describes an agent which detects and handle potentially abnormal situations from the monitoring of applications usage on a tablet computer. The main purpose of this agent is to improve dependent people’s safety by signaling potentially risky situations to caregivers. Indeed, such signaling can improve response time, thus reducing the consequences of such situations. The detection of abnormal situations is based on the construction of a user profile from the monitoring of used applications. When a user is inactive during a certain period of time, the recent activity is compared to the learned user’s profile to decide if this is normal or not. Once an abnormal situation has been identified, the system will try to confirm that the situation is actually abnormal by prompting the user for input. In order to be as less intrusive as possible, the input request is an application suggestion. The suggested application will be the one that is usually the most used during the time period corresponding to the inactivity. When a situation is confirmed as abnormal, the tablet agent will send an intervention request to the user’s caregivers. A simple coordination mechanism aimed at reducing redundant interventions and improving caregivers response time is proposed. The main contribution of this work is to propose a mechanism which monitors elderly people’s applications usage on a tablet computer and is therefore able to complement existing monitoring devices in the detection of abnormal situations.

Using multiagent self-organization techniques for seeking information in virtual social communities

This paper proposes an approach which aims at improving information access skills, services in large networks. It proposes a self-organized multiagent analysis of the problem allowing to reduce the number of messages transmitted for a skill search. The MWAC Multi-Wireless-Agent Communication model is extended to take some specificities of social networks into account, like the information held by the social network members and their connections with the others.The network of agents may represent a social network. Each agent holds information about members agents and is linked with other agents which represent its neighborhood. Each member builds a local network representation according to its needs. A global view of a social network is not dynamically available and furthermore not desirable. Our approach is thus based on a decentralized interrogation of the network and on organizational structures detection.With this method, a search does not need to flood all the network with the query because it uses the self-organized structure that emerges so, it can avoid a saturation in the network and it reduces the amount of transmitted messages. In order to show the interest of the approach, our proposition is validated by several simulations. The applications of this work are related to skill searching in a virtual social network.

Using Multiagent Self-organization Techniques to Improve Dynamic Skill Searching in Virtual Social Communities

This paper introduces an approach which aims at improving information access (skills, services) in large networks. A self-organized multiagent analysis of the problem, reducing the number of messages transmitted for a skill search is proposed. The MWAC (Multi-Wireless-Agent Communication) model is extended here to take some specificities of social networks into account, like the information held by the social network members and their connections with the others.

ENSM-SE at CLEF 2005: using a fuzzy proximity matching function

Starting from the idea that the closer the query terms in a document are to each other the more relevant the document, we propose an information retrieval method that uses the degree of fuzzy proximity of key terms in a document to compute the relevance of the document to the query. Our model handles Boolean queries but, contrary to the traditional extensions of the basic Boolean information retrieval model, does not use a proximity operator explicitly. A single parameter makes it possible to control the proximity degree required. We explain how we construct the queries and report the results of our experiments in the ad-hoc monolingual French task of the CLEF 2005 evaluation campaign.

Towards an Agent-based Model to Monitor Epidemics and Chronic Diseases in DR Congo

Many contagious diseases occurred around sub-Saharan countries in the last decade due to the inefficiency of health structures to anticipate disease outbreaks. In a huge poorly-infrastructured country such as The Democratic Republic of Congo (DRC) with insufficient health staff and laboratory facilities, to provide quick response to an urgent case of epidemic is challenging especially facing the development of its rural areas. As DRC’s Health System has three levels (peripheral, regional and national levels), from producing health data at peripheral to national level that takes the decision, it can take time resulting in the spread of disease. The lack of communication between health centers and laboratory facilities in the same health zone does not contribute to regional riposte. This paper proposes to face this problem using an agent-centered approach to study through simulation how to improve the process. An experiment is described by agentifying two health zones on the same regional level to show how it can reduce the decision time.. It consists of 2 peripheral coordination offices, 2 labs and 2 health zones the former with 12 health centers and the latter with 20 health zones. The interaction between these agents will provide a first model to be compared with the current system in other to reduce decision time.

Toward an Embedded Multi-agent System Methodology and Positioning on Testing

Like all systems, multi-agent systems need to be verified during the design cycle. But the test of multi-agent systems is more diffi-cult than the test of classic software. Indeed, it implies to test more than agent functionalities: individual agent behavior, inter-action between agents and global system need to be tested. Sever-al testing methods compensate these difficulties, but do not cover all the aspects of multi-agent systems. In this paper, we compare software testing and multi-agent systems testing and particularly in embedded context. Then major multi-agent system testing techniques are analyzed, with the AEIO facets for multi-agent systems. Finally, we propose a strategy to conduct the testing activity for embedded MAS.

ENSM-SE at CLEF 2006: fuzzy proxmity method with an adhoc influence function

We experiment a new influence function in our information retrieval method that uses the degree of fuzzy proximity of key terms in a document to compute the relevance of the document to the query. The model is based on the idea that the closer the query terms in a document are to each other the more relevant the document. Our model handles Boolean queries but, contrary to the traditional extensions of the basic Boolean information retrieval model, does not use a proximity operator explicitly. A single parameter makes it possible to control the proximity degree required. To improve our system we use a stemming algorithm before indexing, we take a specific influence function and we merge fuzzy proximity result lists built with different width of influence function. We explain how we construct the queries and report the results of our experiments in the ad-hoc monolingual French task of the CLEF 2006 evaluation campaign.