Béatrice Rumpler

An efficient collaborative information retrieval system by incorporating the user profile

As the volume of information augments, the importance of the Information Retrieval (IR) increases. Collaborative Information Retrieval (CIR) is one of the popular social-based IR approaches. A CIR system registers the previous user interactions to response to the subsequent user queries more efficiently. But the goals and the characteristics of two users may be different; so when they send the same query to a CIR system, they may be interested in two different lists of documents. In this paper we deal with the personalization problem in the CIR systems by constructing a profile for each user. We propose three new approaches to calculate the user profile similarity that we will employ in our personalized CIR algorithm.

Personalised information retrieval in specialised virtual libraries

The New Information Technology in Engineering Science Field (NIT‐ESF) project focuses on the virtual library’s contribution to the education and research field. Different teams are involved in this project. The team specialised in information retrieval aims to design an intelligent tutoring system capable of performing personalised information retrieval, depending on users’ interests. The proposed system is specially designed for people who have a sight deficiency. These specific persons use slow devices to access scientific textual information, so they have a critical need for systems able to retrieve relevant information quickly. For these reasons, we need intelligent systems to define and manage particular user interests. We aim to improve answer quality by personalising the retrieval and by performing user profile management. For implementation, we have based the information filtering and extraction on SMART features. Some system results are shown in this paper.

PERCIRS: a system to combine personalized and collaborative information retrieval

Purpose – This paper aims to discuss and test the claim that utilization of the personalization techniques can be valuable to improve the efficiency of collaborative information retrieval (CIR) systems.Design/methodology/approach – A new personalized CIR system, called PERCIRS, is presented based on the user profile similarity calculation (UPSC) formulas. To this aim, the paper proposes several UPSC formulas as well as two techniques to evaluate them. As the proposed CIR system is personalized, it could not be evaluated by Cranfield, like evaluation techniques (e.g. TREC). Hence, this paper proposes a new user‐centric mechanism, which enables PERCIRS to be evaluated. This mechanism is generic and can be used to evaluate any other personalized IR system.Findings – The results show that among the proposed UPSC formulas in this paper, the (query‐document)‐graph based formula is the most effective. After integrating this formula into PERCIRS and comparing it with nine other IR systems, it is concluded that th...

A graph-based profile similarity calculation method for collaborative information retrieval

Collaborative Information Retrieval (CIR) is one of the popular social-based IR approaches. A CIR system registers the previous user interactions to response to the subsequent user queries more efficiently. But CIR suffers from the personalization problem because the goals and the characteristics of two users may be different; so when they send the same query to a CIR system, they may be interested in two different lists of documents. We have developed a personalized CIR system, called PERCIRS, to solve this problem. Selecting an efficient method to calculate the similarity between user profiles is a key factor for enhancing PERCIRSs efficiency. In this paper, we propose a new graph-based method for user profile similarity calculation. Finally, by introducing an evaluation method, we will show that this new method is more efficient than the previous methods.

Virtual library for ancient manuscripts

This work is the second part of a European project called BAMBI (Better Access to Manuscripts and Browsing of Images). In the initial BAMBI project, we developed a workstation for historians, and more particularly philologists, which allows them to make transcriptions, annotation, indexing, etc. on manuscripts. In this paper, we present the design of the virtual library based on BAMBI, allowing philologists to work on ancient manuscripts within the Internet network. Our approach for the Web implementation of the BAMBI workstation is based on the reuse of the local BAMBI software and on the ActiveX approach.

Mapping visualization on-demand onto a virtual globe: an appealing complement to browser-based navigation

Current browser-based navigation is a universal and powerful tool, but lacks of three useful features: overview of the global website structure, efficient history browsing and an alternative to link-link navigation. By combining Visualization on-demand (Vizod) with an interactive virtual globe, we tackled these issues by means of multi-resolutions maps displayed according to users interactions and preferences. We provided in this way a contextual hypertext navigation, each page being assigned locations and links on top a a virtual map. We built up and performed experiments of a prototype providing a smooth, appealing and promising complement to browser-based navigation.

An Evaluation Model for Systems and Resources Employed in the Correction of Errors in Textual Documents

The wide adoption of Web 2.0 services has resulted in an increase in the amount of information produced. The quantity of errors contained in such information has grown even faster. Indeed, in traditional information production process documents were produced by professionals while in the Web context the content is generated by the users themselves. It is therefore necessary to take into account the errors particularly when such systems need to manage information of variable quality. Our state of the art leads us to identify difficulties in the comparative evaluation of error correction systems. Our proposal consists in an evaluation model for error correction systems and low-level string similarity (and distance) metrics they rely on. This model is implemented in an extensible platform providing a framework to evaluate those systems.

Une interface de programmation visuelle pour la composition de services de visualisation d'information

In this article, we are interested in information visualisations creation and sharing. Our approach is to consider information visualisation as a dataflow, issued from web services compositions which held both syntaxic and semantic rules. To ease service composition, we introduce mashviz, a visual programming interface aimed to both designers and users, to share and annote visualisations. We discuss our early created visualisations, and give clues about our next steps, such as evaluation by usage.

Towards a Leaner Evaluation Process: Application to Error Correction Systems

While they follow similar procedures, evaluations of state of the art error correction systems always rely on different resources (collections of documents, evaluation metrics, dictionaries, ...). In this context, error correction approaches cannot be directly compared without being re-implemented from scratch every time they have to be compared with a new one. In other domains such as Information Retrieval this problem is solved through Cranfield like experiments such as TREC [5] evaluation campaign. We propose a generic solution to overcome those evaluation difficulties through a modular evaluation platform which formalizes similarities between evaluation procedures and provides standard sets of instantiated resources for particular domains. While this was our main problem at first, in this article, the set of resources is dedicated to the evaluation of error correction systems. The idea is to provide the leanest way to evaluate error correction systems by implementing only the core algorithm and relying on the platform for everything else.

Towards a Better Semantic Matching for Indexation Improvement of Error-Prone (Semi-)Structured XML Documents

Documents containing errors in their textual content (which we will call noisy documents) are difficultly handled by Information Retrieval systems. The same observation is verified when it comes to (semi-)structured IR systems this paper deals with. However, the problem is even bigger when those systems rely on Semantics. In order to achieve that, they need an additional external semantic resource related to the documents collection. Then, ranking is made possible thanks to concepts comparisons allowed by similarity measures. Similarity measures assume that concepts related to the words have been identified without ambiguity. Nevertheless, this assumption can’t be made in presence of noisy documents where words are potentially misspelled, resulting in a word having a different meaning or at least in a non-word. Semantic aware (semi-)structured IR systems lay on basic concept identification but they don’t care about spelling uncertainties. As this can degrade systems results, we suggest a way to detect and correct misspelled terms which can be used in documents pre-processing of IR systems. First results on small datasets seem promising.