Juan M. Cigarrán

Browsing Search Results via Formal Concept Analysis: Automatic Selection of Attributes

This paper presents the JBraindead Information Retrieval System, which combines a free-text search engine with online Formal Concept Analysis to organize the results of a query. Unlike most applications of Conceptual Clustering to Information Retrieval, JBraindead is not restricted to specific domains, and does not use manually assigned descriptors for documents nor domain specific thesauruses. Given the ranked list of documents from a search, the system dynamically decides which are the most appropriate attributes for the set of documents and generates a conceptual lattice on the fly. This paper focuses on the automatic selection of attributes: first, we propose a number of measures to evaluate the quality of a conceptual lattice for the task, and then we use the proposed measures to compare a number of strategies for the automatic selection of attributes. The results show that conceptual lattices can be very useful to group relevant information in free-text search tasks. The best results are obtained with a weighting formula based on the automatic extraction of terminology for thesaurus building, as compared to an Okapi weighting formula.

Automatic selection of noun phrases as document descriptors in an FCA-Based information retrieval system

Automatic attribute selection is a critical step when using Formal Concept Analysis (FCA) in a free text document retrieval framework. Optimal attributes as document descriptors should produce smaller, clearer and more browsable concept lattices with better clustering features. In this paper we focus on the automatic selection of noun phrases as document descriptors to build an FCA-based IR framework. We present three different phrase selection strategies which are evaluated using the Lattice Distillation Factor and the Minimal Browsing Area evaluation measures. Noun phrases are shown to produce lattices with good clustering properties, with the advantage (over simple terms) of being better intensional descriptors from the users point of view.

Multimatch-multilingual/multimedia access to cultural heritage

Cultural heritage content is everywhere on the web, in contexts such as digital libraries, audiovisual archives, and portals of museums or galleries, in multiple languages and multiple media. MultiMatch, a 30 month specific targeted research project under the Sixth Framework Programme, plans to develop a multilingual search engine designed specifically for the access, organisation and personalised presentation of cultural heritage digital objects.

A step forward for Topic Detection in Twitter

We propose a novel approach based on Formal Concept Analysis for Topic Detection.Our proposal overcomes traditional problems of the clustering and classification techniques.We analyse the parameters involved in the process in a Twitter-based framework.We propose a topic selection methodology based on the stability concept.We overcome the state-of-the-art results for the task. The Topic Detection Task in Twitter represents an indispensable step in the analysis of text corpora and their later application in Online Reputation Management. Classification, clustering and probabilistic techniques have been traditionally applied, but they have some well-known drawbacks such as the need to fix the number of topics to be detected or the problem of how to integrate the prior knowledge of topics with the detection of new ones. This motivates the current work, where we present a novel approach based on Formal Concept Analysis (FCA), a fully unsupervised methodology to group similar content together in thematically-based topics (i.e., the FCA formal concepts) and to organize them in the form of a concept lattice. Formal concepts are conceptual representations based on the relationships between tweet terms and the tweets that have given rise to them. It allows, in contrast to other approaches in the literature, their clear interpretability. In addition, the concept lattice represents a formalism that describes the data, explores correlations, similarities, anomalies and inconsistencies better than other representations such as clustering models or graph-based representations. Our rationale is that these theoretical advantages may improve the Topic Detection process, making them able to tackle the problems related to the task. To prove this point, our FCA-based proposal is evaluated in the context of a real-life Topic Detection task provided by the Replab 2013 CLEF Campaign. To demonstrate the efficiency of the proposal, we have carried out several experiments focused on testing: (a) the impact of terminology selection as an input to our algorithm, (b) the impact of concept selection as the outcome of our algorithm, and; (c) the efficiency of the proposal to detect new and previously unseen topics (i.e., topic adaptation). An extensive analysis of the results has been carried out, proving the suitability of our proposal to integrate previous knowledge of prior topics without losing the ability to detect novel and unseen topics as well as improving the best Replab 2013 results.

Evaluating hierarchical clustering of search results

We propose a goal-oriented evaluation measure, Hierarchy Quality, for hierarchical clustering algorithms applied to the task of organizing search results -such as the clusters generated by Vivisimo search engine-. Our metric considers the content of the clusters, their hierarchical arrangement, and the effort required to find relevant information by traversing the hierarchy starting from the top node. It compares the effort required to browse documents in a baseline ranked list with the minimum effort required to find the same amount of relevant information by browsing the hierarchy (which involves examining both documents and node descriptors).

Linked Data-based Conceptual Modelling for Recommendation: A FCA-Based Approach

In a recommendation task it is crucial to have an accurate content-based description of the users and the items. Linked Open Data (LOD) has been demonstrated as one of the best ways of obtaining this kind of content. The main question is to know how useful the LOD information is in inferring user preferences and how to obtain it. We propose a novel approach for Content Modelling and Recommendation based on Formal Concept Analysis (FCA). The approach is based in the modelling of the user and content related information, enriched with LOD, and in a new algorithm to analyze the models and recommend new content. The framework provided by the ESWC 2014 Recommendation Challenge is used for the evaluation. The results are within the average range of other participants, but further work has to be carried out to refine the approach using LOD information.

Multimedia retrieval in a medical image collection: results using modality classes

The effective communication between user and systems is one main aim in the Multimedia Information Retrieval field. In this paper the modality classification of images is used to expand the user queries within the ImageCLEF Medical Retrieval collection provided by organizers. Our main contribution is to show how and when results can be improved by understanding modality-related challenges. To do so, a detailed analysis of the results of the experiments carried out is presented and the comparison between these results shows that the improvement using modality class query expansion is query-dependent.

Enhancing higher education experience: The eMadrid initiative at UNED university

In this paper we focus on the achievements of eMadrid initiative in some fields of technology-enhanced learning, mainly involving the improvement of the mechanisms for open educational content retrieval from Internet, considering Internet resources as potential learning objects. Also we facilitate the integration of remote laboratories and external tools in virtual campuses architectures supporting enriched capabilities and describe a way to cluster and identify learner weaknesses using a learning analytics approach in combination with the item response theory.

Declarative Mark-Up Language as a Tool for Developing Educational Hypermedia

Web ever-growing generalization and its use in educational settings is impelling the development of high-quality educational hypermedia applications. Hypermedia systems complexity has promoted the use of abstract hypermedia models to capture the most relevant characteristics of these systems. Current models put the stress on hypermedia components and its hyperlinking structure (e.g. Dexter Model), on-screen presentation (e.g. Amsterdam Model), or on the adaptation of the contents to specific users (e.g. adaptive hyperbooks). These models usually do not address specific aspects that are crucial in the design and maintenance of educational hypermedias, such as the structure and content of the individual components, the structural relations between them, and the conceptual model or educational strategy underlying the application. In our work, we use a descriptive specification, based on the Standard General Mark-up Language (SGML), to simplify the design, development and maintenance of educational hypermedia. We use an SGML based specification for modeling two main aspects of a hypermedia educational application: a) the organization and structure of the application contents, and b) a description that explicitly captures relevant design decisions about presentation and educational features. This kind of description, a “document type definition” in SGML terminology, provides clear and useful design documents that easily relate the work of designers and programmers partially achieving software and platform independence. Also, the specific information about contents simplifies its adaptation to individual users based on an explicit user model. This approach has been used in the development of a hypermedia system designed for teaching text comprehension in a foreign language.

Integration of Formal Concept Analysis in a Knowledge-Based Assistant

To build a knowledge base for an intelligent help system, as Aran, is an essential and difficult phase. In Aran, a knowledge based assistant to help users in the use of UNIX systems, we integrate the traditional manual approach of conceptual information structuring with a complementary one based on Formal Concept Analysis (FCA). FCA allow us to obtain the domain formal concepts (semi) automatically and to organise the information around them.