Towards a model-driven engineering approach of data mining

Nowadays, data mining is based on low-level specications of the employed techniques typically bounded to a specic analysis platform. Therefore, data mining lacks a modelling architecture that allows analysts to consider it as a truly software-engineering process. Here, we propose a model-driven approach based on (i) a conceptual modelling framework for data mining, and (ii) a set of model transformations to automatically generate both the data under analysis (via data-warehousing technology) and the analysis models for data mining (tailored to a specic platform). Thus, analysts can concentrate on the analysis problem via conceptual data-mining models instead of low-level programming tasks related to the underlying-platform technical details. These tasks are now entrusted to the model-transformations scaffolding.

Integrating the development of data mining and data warehouses via model-driven engineering

Data mining is one of the most important analysis techniques to automatically extract knowledge from large amount of data. Nowadays, data mining is based on low-level specifications of the employed techniques typically bounded to a specific analysis platform. Therefore, data mining lacks a modelling architecture that allows analysts to consider it as a truly software-engineering process. Bearing in mind this situation, we propose a model-driven approach which is based on (i) a conceptual modelling framework for data mining, and (ii) a set of model transformations to automatically generate both the data under analysis (that is deployed via data-warehousing technology) and the analysis models for data mining (tailored to a specific platform). Thus, analysts can concentrate on understanding the analysis problem via conceptual data-mining models instead of wasting efforts on low-level programming tasks related to the underlying-platform technical details. These time consuming tasks are now entrusted to the model-transformations scaffolding. The feasibility of our approach is shown by means of a hypothetical data-mining scenario where a time series analysis is required.

Uso de la Web Social para fomentar el aprendizaje autónomo [póster]

Póster presentado en las X Jornadas de Redes de Investigación en Docencia Universitaria, Alicante, 16-17 junio 2012.

Uso de la Web Social para fomentar el aprendizaje autónomo

Comunicación presentada en las X Jornadas de Redes de Investigación en Docencia Universitaria, Alicante, 16-17 junio 2012.

Uso de redes sociales para mejorar el rendimiento de los alumnos con diferentes estilos de aprendizaje

Debido a los cambios que el Espacio Europeo de Educación Superior introduce al potenciar las horas de trabajo no presencial, se hacen necesarios nuevos mecanismos para posibilitar una mejor comunicación y cooperación en el proceso de aprendizaje. Las redes sociales, como Facebook, pueden suministrar estos mecanismos, pero su uso satisfactorio para la docencia puede verse afectado en gran medida por el estilo de aprendizaje de los alumnos. Este artículo plantea la necesidad de estudiar la influencia de los diferentes estilos de aprendizaje en la docencia no presencial mediante el uso de redes sociales con el fin de incrementar el rendimiento de los alumnos. Cabe destacar que este artículo describe el proyecto “Las redes sociales y su relación con los estilos de aprendizaje” a realizar dentro del programa de Redes de Investigación en Docencia Universitaria del Instituto de Ciencias de la Educación de la Universidad de Alicante.

Influencia de los estilos de aprendizaje en el uso de redes sociales para docencia

Comunicación presentada en las IX Jornadas de Redes de Investigación en Docencia Universitaria, Alicante, 16-17 junio 2011.

Influencia de los estilos de aprendizaje en el uso de redes sociales para docencia [póster]

Póster presentado en las IX Jornadas de Redes de Investigación en Docencia Universitaria, Alicante, 16-17 junio 2011.

Las redes sociales y su relación con los estilos de aprendizaje

Memoria Redes ICE 2011

Memoria de la red AprenWeb: uso de la Web Social para fomentar el aprendizaje autónomo

Memoria Redes ICE 2012

Open business intelligence: on the importance of data quality awareness in user-friendly data mining

Citizens demand more and more data for making decisions in their daily life. Therefore, mechanisms that allow citizens to understand and analyze linked open data (LOD) in a user-friendly manner are highly required. To this aim, the concept of Open Business Intelligence (OpenBI) is introduced in this position paper. OpenBI facilitates non-expert users to (i) analyze and visualize LOD, thus generating actionable information by means of reporting, OLAP analysis, dashboards or data mining; and to (ii) share the new acquired information as LOD to be reused by anyone. One of the most challenging issues of OpenBI is related to data mining, since non-experts (as citizens) need guidance during preprocessing and application of mining algorithms due to the complexity of the mining process and the low quality of the data sources. This is even worst when dealing with LOD, not only because of the different kind of links among data, but also because of its high dimensionality. As a consequence, in this position paper we advocate that data mining for OpenBI requires data quality-aware mechanisms for guiding non-expert users in obtaining and sharing the most reliable knowledge from the available LOD.

Hacia la consideración de aspectos de calidad de datos en procesos de minería: el caso de las técnicas de clasificación

El éxito en la búsqueda de conocimiento a partir de grandes cantidades de datos radica en la calidad de los mismos. Hasta ahora los aspectos de calidad de los datos se han enfocado principalmente a la limpieza de los datos: detección de duplicados, valores atípicos, perdidos, incompletos o conflictos en instancias, entre otros. En este trabajo se presenta un caso de estudio que nos ha permitido determinar ciertos aspectos de calidad que pueden mejorar la expectativa de éxito en el análisis evitando resultados erróneos, incorrectos o poco fiables. Este es un primer paso hacia la consideración de manera sistemática y estructurada de criterios de calidad específicos para minería de datos que ayude al minero de datos en sus objetivos.

Towards a reverse engineering approach for guiding user in applying data mining

Comunicación presentada en las XVI Jornadas de Ingeniería del Software y Bases de Datos, JISBD 2011, A Coruña, 5-7 septiembre 2011.

Creación automática de sistemas de búsqueda de respuestas en dominios restringidos

Los sistemas de búsqueda de respuestas (BR) se pueden considerar como potenciales sucesores de los buscadores tradicionales de información en la Web. Para que sean precisos deben adaptarse a dominios concretos mediante el uso de recursos semánticos adecuados. La adaptación no es una tarea trivial, ya que deben integrarse e incorporarse a sistemas de BR existentes varios recursos heterogéneos relacionados con un dominio restringido. Se presenta la herramienta Maraqa, cuya novedad radica en el uso de técnicas de ingeniería del software, como el desarrollo dirigido por modelos, para automatizar dicho proceso de adaptación a dominios restringidos. Se ha evaluado Maraqa mediante una serie de experimentos (sobre el dominio agrícola) que demuestran su viabilidad, mejorando en un 29,5% la precisión del sistema adaptado.

Does Müller cell differentiation occur prior to the emergence of synapses in embryonic turtle retina?

Müller cells are the main glial cells in the retina, and are related to plexiform layer activity. Recent studies have demonstrated that Müller cells are involved in the synaptic conservation, plasticity, development and metabolism of glutamate. During turtle retinal development, layers, cells and synapses appear at different times. The aim of this research is to study the emergence of Müller cells during embryonic development and their relationship with the synaptogenesis. The authors used retinas from Trachemys scripta elegans embryos at stages S14, 18, 20, 23, and 26. Some retinas were processed with immunocytochemistry in order to detect the presence of glutamine synthetase in Müller cells, which was used as a marker of these cells. Other retinas from the same stages were processed for ultrastructural studies. Samples were observed in confocal and transmission electron microscopes, respectively. The present results show that glutamine synthetase expression in Müller cells occurs at S18, before the emergence of the retinal layers and the early synapses.

On the nature of candidate luminous blue variables in M 33

Context. Luminous blue variables (LBVs) are a class of highly unstable stars that have been proposed to play a critical role in massive stellar evolution as well as being the progenitors of some of the most luminous supernovae known. However the physical processes underlying their characteristic instabilities are currently unknown. Aims. In order to provide observational constraints on this behaviour we have initiated a pilot study of the population of (candidate) LBVs in the Local Group galaxy M 33. Methods. To accomplish this we have obtained new spectra of 18 examples within M 33. These provide a baseline of ≥ 4 yr with respect to previous observations, which is well suited to identifying LBV outbursts. We also employed existing multi-epoch optical and mid-IR surveys of M 33 to further constrain the variability of the sample and search for the presence of dusty ejecta. Results. Combining the datasets reveals that spectroscopic and photometric variability appears common, although in the majority of cases further observations will be needed to distinguish between an origin for this behavour in short lived stochastic wind structure and low level photospheric pulsations or coherent long term LBV excursions. Of the known LBVs we report a hitherto unidentified excursion of M 33 Var C between 2001-5, while the transition of the WNLh star B517 to a cooler B supergiant phase between 1993−2010 implies an LBV classification. Proof-of-concept quantitative model atmosphere analysis is provided for Romano’s star; the resultant stellar parameters being consistent with the finding that the LBV excursions of this star are accompanied by changes in bolometric luminosity. The combination of temperature and luminosity of two stars, the BHG [HS80] 110A and the cool hypergiant B324, appear to be in violation of the empirical Humphreys-Davidson limit. Mid-IR observations demonstrate that a number of candidates appear associated with hot circumstellar dust, although no objects as extreme as η Car are identified. The combined dataset suggests that the criteria employed to identify candidate LBVs results in a heterogeneous sample, also containing stars demonstrating the B[e] phenomenon. Of these, a subset of optically faint, low luminosity stars associated with hot dust are of particular interest since they appear similar to the likely progenitor of SN 2008S and the 2008 NGC 300 transient (albeit suffering less intrinsic extinction). Conclusions. The results of such a multiwavelength observational approach, employing multiplexing spectrographs and supplemented with quantitative model atmosphere analysis, appears to show considerable promise in both identifying and characterising the physical properties of LBVs as well as other short lived phases of massive stellar evolution.