Jaime A. Guzmán

Peniel method for the ultrasonic material identification in robots

In this paper is addressed the development of an ultrasonic sensor for the material identification. This sensor had to be appropriate, in terms of computational cost, in order to be implemented in small robots. This implementation was achieved by a novel signal processing method called Peniel. The associated electronic circuit and the algorithm of this method have as main strengths their circuital simplicity and the low computational cost, respectively. These two characteristics let the sensor to be implemented in two robots of the low cost robotic kit TEAC2H-RI, which was constructed as part of the master thesis of one the authors. The developed system was proved with acrylic, aluminium and glass samples. In the performed experiments those materials were identified with 100% accuracy in all of ten trials. According to the literature review, it is the first time which is achieved the implementation of the ultrasonic material identification in a robotic system.

A Case-Based Multi-Agent and Recommendation Environment to Improve the E-Recruitment Process

The current growth of information and communication technologies has promoted the development of tools in order to facilitate the process of e-recruitment; benefiting both recruiters as jobseekers. This paper presents a case-based Multi-Agent System which aims at integrating an ontology in order to select and to recommend adapted jobseekers to the recruiter job postings or vice versa. For this reason, the ontology considers the HR-XML standard for map-ping CVs in order to standardize the knowledge representation. The MAS was designed following the Prometheus Methodology and then a prototype has been implemented. A case study was performed within a testing phase in order to validate our work. As a result of this phase, we can prove the effectiveness of using this kind of technologies in the e-recruitment process.

Trajectory reconstruction using personalized routing over a graph

Location based services provide information enriched with geo-located data. However, the detailed trajectory of a moving object is lost when low-sampling data are present. Previous works have been developed to explore trajectories using location history logged by users in order to reconstruct the trajectory of a moving object. Those methods can be considered as reconstruction or imputation processes. In this paper, we reconstruct trajectories using the personalization features of routing theory based on criterion decision and over a graph. In addition, we consider the trajectory reconstruction only in a constrained environment, a road network.

An Analysis of a Methodology that Transforms the Entity-Relationship Model into a Conceptual Model for a Graph Database

The graph databases (GDB) have gained a lot of importance in the last years; this is due to the necessity to store and manage very large volumes of data whose natural structure is a graph. However, nowadays there do not exist conceptual models widely accepted to represent a GDB. This fact implies that the analysts are guided considering their experience and best practices. There have been proposed different conceptual models for GDB; in this paper, we analyze a methodology that generates a conceptual model for a GDB from the entity-relationship (E-R) model. We explore several limitations of this methodology and offer some ideas for solving them.

Extracción de información desde la web para identificar acciones de un modelo de dominio en planificación automática

Automated Planning (AP) is the discipline of Artificial Intelligence whose aim is to produce a sequence of actions in order to achieve a specific goal and requires the inflow of a pre-defined action model. However, to start the construction of this model is a difficult task even for experts. This paper proposes the extraction of information from the Web, in order to identify those elements that correspond to the action model for AP tasks, by analyzing a number of Web pages that contain data of already existing programs and extracting from them a set of tools which would allow to solve any given problem. The data obtained will be processed using Natural Language Processing (NLP) tools, in order to identify a set of actions that are part of an AP domain model. The proposed system reached an average accuracy of 89.87% and save all actions taken on ontology to form a Knowledge Base (BC), which allows later use for other domains of PA. This article presents the results of a partial research about the uses of extraction tools, pre-processing, identification of components and storage in the ontology.

Un Modelo para la Identificación de Elementos KAOS (Especificación Automática de Adquisición de Conocimientos) a partir de la Especificación de Requisitos en Lenguaje Natural

Resumen En este articulo, se propone un modelo para obtener de manera automatica los elementos basicos del diagrama de objetivos de KAOS (objetivo, agente y entidad) y minimizar la ambiguedad semantica de tipo polisemica en la obtencion de dichos elementos. Ademas, se valida el modelo propuesto mediante casos de estudio. El diagrama de objetivos de KAOS (Especificacion Automatica de Adquisicion de Conocimientos; del ingles Knowledge Acquisition Automated Specification) suele ser utilizado en la fase de definicion y analisis de la ingenieria de software para determinar los procesos que se efectuan en la organizacion que solicita el software. En los diferentes trabajos publicados en la literatura cientifica que usan el diagrama de objetivos de KAOS subsisten algunos problemas causados por la elaboracion manual del diagrama por parte del analista (confusiones, subjetividad, ambiguedad, entre otros). Con la metodologia propuesta los elementos basicos del diagrama de objetivos de KAOS se obtienen de manera automatica evitando errores como los mencionados.

Prediction of Failure Risk Through Logical Decision Trees in Web Service Compositions

In a service composition, the Quality of Services can be useful to identify those hidden data for a traditional composition; they can be a decisive factor for determining the behavior of future compositions since they allow evaluating risks resulting from reasons totally dependent on both the service environment and/or the composition system. Importance of this data is reflected on the way they are obtained, estimated, and applied to a composition. This paper has specifically studied the following three characteristics: availability, reactivity of services in periods of time, and management of beliefs to determine influence of services composition and to determine failure risk in such a composition through machine learning.