Through semantic labelling we enrich structured information from sources such as HTML pages, tables, or JSON files, with labels to integrate it into a local ontology. This process involves measuring some features of the information and then finding the classes that best describe it. The problem with current techniques is that they do not model relationships between classes. Their features fall short when some classes have very similar structures or textual formats. In order to deal with this problem, we have devised TAPON: a new semantic labelling technique that computes novel features that take into account the relationships. TAPON computes these features by means of a two-phase approach. In the first phase, we compute simple features and obtain a preliminary set of labels (hints). In the second phase, we inject our novel features and obtain a refined set of labels. Our experimental results show that our technique, thanks to our rich feature catalogue and novel modelling, achieves higher accuracy than other state-of-the-art techniques.
Autores: Daniel Ayala / Inma Hernandez / David Ruiz / Miguel Toro /
Palabras Clave: Information Integration - Machine Learning - Semantic labelling
Business Process Management Systems (BPMS) are increasingly used to support service composition, typically working with executable BP models that involve resources, which include both automatic services and services provided by human resources. The appropriate selection of human resources is critical, as factors such as workload or skills have an impact on work performance. While priorities for automatic services are intensively researched, human resource prioritization has been hardly discussed. In classical workflow management, only resource assignment at BP design time to select potential performers for activities, and resource allocation at run time to choose actual performers, are considered. There is no explicit consideration of prioritizing potential performers to facilitate the selection of actual performers. It is also disregarded in professional solutions.
In this paper, we address this research gap and provide two contributions: (i) we conceptually define prioritized allocation based on preferences; and (ii) we propose a concrete way in which preferences over resources can be defined so that a resource priority ranking can be automatically generated. Our solution builds on the adaptation of a user preference model developed for the discovery and ranking of semantic web services called SOUP  to the domain at hand. As a proof of concept, we have extended the resource management tool CRISTAL (http://www.isa.us.es/cristal) with the SOUP component , using RAL  for resource selection. 1. J. M. García, D. Ruiz, and A. R. Cortés, «A Model of User Preferences for Semantic
Services Discovery and Ranking,» in ESWC (2), pp. 114, Springer, 2010. 2. J. M. García, M. Junghans, D. Ruiz, S. Agarwal, and A. R. Cortés, «Integrating
semantic Web services ranking mechanisms using a common preference model,» Knowl.-Based Syst., vol. 49, pp. 2236, 2013. 3. C. Cabanillas, M. Resinas, and A. Ruiz-Cortés, «Defining and Analysing Resource Assignments in Business Processes with RAL,» in ICSOC, vol. 7084, pp. 477486, Springer, 2011.
This work was published in ICSOC 2013, vol. 8274, 374-388. It was partially supported by the EU-FP7, the EU Commission, the Spanish and the Andalusian R&D&I programmes (grants 318275, 284860, TIN2009-07366, TIN2012-32273, TIC-5906).
Autores: Cristina Cabanillas / José María García / Manuel Resinas / David Ruiz / Antonio Ruiz-Cortés / Jan Mendling /
There exist many available service ranking implementations, each one providing ad hoc preference models that offer different levels of expressiveness. Consequently, applying a single implementation to a particular scenario constrains the user to define preferences based on the underlying formalisms. Furthermore, preferences from different ranking implementation’s model cannot be combined in general, due to interoperability issues. In this article we present an integrated ranking implementation that enables the combination of three different ranking implementations developed within the EU FP7 SOA4All project. Our solution has been developed using PURI, a Preference-based Universal Ranking Integration framework that is based on a common, holistic preference model that allows to exploit synergies from the integrated ranking implementations, offering a single user interface to define preferences that acts as a fa¸cade to the integrated ranking implementation.
Autores: José María García / David Ruiz / Antonio Ruiz-Cortés /
Palabras Clave: Preference Models - Ranking Tools - Semantic Web Services - Systems Integration
In the context of knowledge graphs, the task of completion of relations consists in adding missing triples to a knowledge graph, usually by classifying potential candidates as true of false. Creating an evaluation dataset for these techniques is not trivial, since there is a large amount of variables to consider which, if not taken into account, may cause misleading results. So far, there is not a well defined workflow that identifies the variation points when creating a dataset, and what are the possible strategies that can be followed in each step. Furthermore, there are no tools that help create such datasets in an easy way. To address this need, we have created AYNEC-DataGen, a customisable tool for the generation of datasets with multiple variation points related to the preprocessing of the original knowledge graph, the splitting of triples into training and testing sets, and the generation of negative examples. The output of our tool includes the evaluation dataset, an optional export in an open format for its visualisation, and additional files with metadata. Our tool is freely available online.
Autores: Daniel Ayala / Agustin Borrego / Inma Hernandez / Carlos R. Rivero / David Ruiz /
Palabras Clave: Evaluation - Graph refinement - Knowledge Graph - Tool