Christian Sallaberry

On the evaluation of Geographic Information Retrieval systems: Evaluation framework and case study

Search engines for Digital Libraries allow users to retrieve documents according to their contents. They process documents without differentiating the manifold aspects of information. Spatial and temporal dimensions are particularly dismissed. These dimensions are, however, of great interest for users of search engines targeting either the Web or specialized Digital Libraries. Recent studies reported that nearly 20% queries convey spatial and temporal information in addition to topical information. These three dimensions were referred to as parts of “geographic information.” In the literature, search engines handling those dimensions are called “Geographic Information Retrieval (GIR) systems.” Although several initiatives for evaluating GIR systems were undertaken, none was concerned with evaluating these three dimensions altogether. In this article, we address this issue by designing an evaluation framework, usefulness of which is highlighted through a case study involving a test collection and a GIR system. This framework allowed the comparison of our GIR system to state-of-the-art topical approaches. We also performed experiments for measuring performance improvement stemming from each dimension or their combination. We show that combining the three dimensions yields improvement in effectiveness (+73.9%) over a common topical baseline. Moreover, rather than conveying redundancy, the three dimensions complement each other.

La composante temps dans l'information géographique textuelle : Extraction et recherche d'information dans des fonds documentaires patrimoniaux numérisés

Les travaux presentes dans ce papier se situent dans le cadre de la recherche d’information geographique. Notre objectif est d’exploiter les fonds documentaires patrimoniaux dont le contenu est fortement attache au territoire (information geographique). L’information geographique peut etre declinee selon le triptyque espace, temps, phenomene. Ce papier se focalise plus particulierement sur le volet temps. Nous proposons tout d’abord un modele decrivant les expressions temporelles a annoter. Puis, nous presentons une methodologie d’annotation semantique pour l’indexation automatique de ces expressions. Enfin, nous evoquerons la recherche d’information geographique d’abord selon les criteres spatial, thematique et temporel respectivement, puis nous combinons ces trois criteres.

From Electronic Documents to Problem-based Learning Environments: an ongoing Challenge for Educational Modeling Languages

This work reports an evaluation of two educational modeling languages for the design of environments dedicated to problem-based learning (PBL) situations. Firstly, we assess the expressivity of the IMS-LD language against the requirements of the Smash case-study. Considering the limits of the IMS-LD language, we present some capabilities of the CPM language to describe other views representing contextual didactic choices (actions and events) that IMS-LD cannot afford. An analysis of the details of these CPM views has shown that they refer to learning/tutoring actions to be performed from the semantics of the electronic documents exploited in the Smash situation. Since it is a widely shared tendency to specify learning activities that take advantage of the information embedded in documents, we present some techniques, methods and tools to engineer together the educational capabilities of a corpus of electronic documents and the dynamics of learning activities enhancing the value of such a corpus. The only documents addressed by these research works are those embedding spatial and temporal information

Profiling cooperative problem-based learning situations

Computer-based learning aims in essence to imitate cognitive behaviors, assist learners and teachers in pedagogical exchanges and monitor their overall cooperation based on learning paradigms. In this respect, this paper is concerned with problem-based learning (PBL). Designing PBL platforms comes up against the lack of reusable domain-specific or educational software assets. To avoid such a problem, metamodeling (via a UML profile) is considered in this paper as a reliable way for instantiating PBL situation models. A semantics and a notation are in particular associated with the proposed metamodel in order to provide rigor but also flexibility in pedagogical scenario elaboration. An example in particular illustrates the applicability of the offered educational modeling language (EML).

Normalizing Spatial Information to Improve Geographical Information Indexing and Retrieval in Digital Libraries

Our contribution is dedicated to geographic information contained in unstructured textual documents. The main focus of this article is to propose a general indexing strategy that is dedicated to spatial information, but which could be applied to temporal and thematic information as well. More specifically, we have developed a process flow that indexes the spatial information contained in textual documents. This process flow interprets spatial information and computes corresponding accurate footprints. Our goal is to normalize such heterogeneous grained and scaled spatial information (points, polylines, polygons). This normalization is carried out at the index level by grouping spatial information together within spatial areas and by using statistics to compute frequencies for such areas and weights for the retrieved documents.

GeoStream: Spatial Information Indexing Within Textual Documents Supported by a Dynamically Parameterized Web Service

Cultural heritage content is everywhere on the web: digital libraries, archives, and portals of museums or galleries. Cultural heritage document collections are characterized by contents related to a territory and its lands history. In this context, the GeoTopia project is supported by the CNRS-TGE-Adonis and focuses on archive data sharing and interpretation. It consists in a Content Management System (CMS) that aims to manage a repository of multimedia digital documents: it exploits information like origin, theme, period, area, etc. to index and/or query documents. Our contribution is dedicated to spatial information contained in non structured textual documents. More specifically, we have developed a process flow that can extract the spatial information contained in textual documents. This process flow indexes spatial information and computes precise geolocalized representations. We propose to encapsulate it into the GeoStream specific web service and to make its behavior dynamically customizable for easier integration into such platforms used for the management of cultural heritage electronic documents

Spatial Information Management within Digital Libraries

Local cultural heritage document collections are characterized by contents strongly attached to a territory and its land history. Our contribution aims at enhancing such a content retrieval process efficiently each time a query includes geographic criteria. We propose a unified model for a formal representation of geographic information. This geographic model allows spatial features to be described independently of their representation mode (textual, graphical) in the documents. We have developed a prototype implementing geographic information extraction (IE) and geographic information retrieval (IR) processes. We process geographic IE with semantic techniques combined to classic IE approaches. Then, we implemented geographic IR with intersections researching algorithms: these algorithms search for all geocoded entities in the documents collections indexes which intersect any entity in the users request. This paper focuses on IR and visualization proposals relying on the geospatial characteristics of cultural heritage corpora.

Model Driven Development of Cooperative Problem-Based Learning Situations Implementing Tools for Teachers and Learners from Pedagogical Models

Model Driven Development is a promising research area for the production of systems dedicated to Human Learning. Our research focuses on developing tools and techniques for producing specific models for cooperative Problem-Based Learning (PBL). Our works led us to propose a UML profile dedicated to such learning situations modeling. The profile is built on the CPM (Co-operative Problem-based learning Meta-model) meta-model which provides the designers with a set of concepts allowing to semi-formally describe a learning scenario and to document it. CPM is thus a language which is both useful for the designers team and for the developers of computerized systems while implementing all or part of a learning scenario. In this paper, we focus on the added value of the models produced from the CPM language. We pinpoint the advantage for tutors and learners using systems implemented from models of a cooperative learning scenario. Finally, we briefly present the Learning Management System (LMS) and the technical architecture from which we have implemented educational tools based on such human learning situations.

Using the Semantics of Texts for Information Retrieval: A Concept- and Domain Relation-Based Approach

Our hypothesis is that assessing the relevance of a document with respect to a query is equivalent to assessing the conceptual similarity between the terms of the query and those of the document. In this article, we therefore propose a method of calculating conceptual similarity. Our information retrieval strategy is based on exploring an ontology and domain relations between concepts marked by verbal forms. Our approach overall is implemented by a prototype and the results obtained are evaluated. We thus show that a semantic IR system based on concepts improves recall with respect to a classic IR system and that a semantic IR system based on concepts and domain relations improves precision with respect to IR based on concepts alone.

Toward the Spatial and Temporal Management of Documents: The GéoTopia Platform

Abstract GeoTopia is an online platform designed to promote the publication, distribution, exchange, discussion, and analysis of archive data. As an original feature, it incorporates functions for the spatial and temporal management of relatively informal documents of complex structure, such as images, drawings, and texts. A map interface allows the archiving, consultation, and handling of georeferenced documents by means of a zoom bar and a timeline. Display by map juxtaposition facilitates the comparison of two documents. Two Web services with adjustable parameters, GeoStream and TempoStream, have been developed to aid users with the spatial and temporal indexing of the texts that they integrate into the GeoTopia platform. The result of the spatial indexing is displayed on the base map in the form of thumbnail images showing the spatial references extracted from the text, which the user can validate, either wholly or in part.