Autonomous Acquisition of Natural Situated Communication

An important part of human intelligence, both historically and operationally, is our ability to communicate. We learn how to communicate, and maintain our communicative skills, in a society of communicators – a highly effective way to reach and maintain proficiency in this complex skill. Principles that might allow artificial agents to learn language this way are in completely known at present – the multi-dimensional nature of socio-communicative skills are beyond every machine learning framework so far proposed. Our work begins to address the challenge of proposing a way for observation-based machine learning of natural language and communication. Our framework can learn complex communicative skills with minimal up-front knowledge. The system learns by incrementally producing predictive models of causal relationships in observed data, guided by goal-inference and reasoning using forward-inverse models. We present results from two experiments where our S1 agent learns human communication by observing two humans interacting in a realtime TV-style interview, using multimodal communicative gesture and situated language to talk about recycling of various materials and objects. S1 can learn multimodal complex language and multimodal communicative acts, a vocabulary of 100 words forming natural sentences with relatively complex sentence structure, including manual deictic reference and anaphora. S1 is seeded only with high-level information about goals of the interviewer and interviewee, and a small ontology; no grammar or other information is provided to S1 a priori. The agent learns the pragmatics, semantics, and syntax of complex utterances spoken and gestures from scratch, by observing the humans compare and contrast the cost and pollution related to recycling aluminum cans, glass bottles, newspaper, plastic, and wood. After 20 hours of observation S1 can perform an unscripted TV interview with a human, in the same style, without making mistakes.

Generating Linked-Data based Domain-Specific Sentiment Lexicons from Legacy Language and Semantic Resources

We present a methodology for legacy language resource adaptation that generates domain-specific sentiment lexicons organized around domain entities described with lexical information and sentiment words described in the context of these entities. We explain the steps of the methodology and we give a working example of our initial results. The resulting lexicons are modelled as Linked Data resources by use of established formats for Linguistic Linked Data (lemon, NIF) and for linked sentiment expressions (Marl), thereby contributing and linking to existing Language Resources in the Linguistic Linked Open Data cloud.

Parameter-based mechanism for unifying user interaction, applications and communication protocols

In the smart building control industry, creating a platform to integrate different communication protocols and ease the interaction between users and devices is becoming increasingly important. BATMP is a platform designed to achieve this goal. In this paper, the authors describe a novel mechanism for information exchange, which introduces a new concept, Parameter, and uses it as the common object among all the BATMP components: Gateway Manager, Technology Manager, Application Manager, Model Manager and Data Warehouse. Parameter is an object which represents a physical magnitude and contains the information about its presentation, available actions, access type, etc. Each component of BATMP has a copy of the parameters. In the Technology Manager, three drivers for different communication protocols, KNX, CoAP and Modbus, are implemented to convert devices into parameters. In the Gateway Manager, users can control the parameters directly or by defining a scenario. In the Application Manager, the applications can subscribe to parameters and decide the values of parameters by negotiating. Finally, a Negotiator is implemented in the Model Manager to notify other components about the changes taking place in any component. By applying this mechanism, BATMP ensures the simultaneous and concurrent communication among users, applications and devices.

Capacity Building through education, research and collaboration: AFRICA BUILD, an eHealth Case Study

AFRICA BUILD (AB) is a Coordination Action project under the 7th European Framework Programme having the aim of improving the capacities for health research and education in Africa through Information and Communication Technologies (ICT). This project, started in 2012, has promoted health research, education and evidence-based practice in Africa through the creation of centers of excellence, by using ICT,?know-how?, eLearning and knowledge sharing, through Web-enabled virtual communities.

Physical activity assessment in the general population; instrumental methods and new technologies.

La medición objetiva del movimiento humano y la cuantificación del gasto energético debido a la actividad física es una necesidad identificada tanto en investigación como en clínica. Los métodos de referencia validados y bien definidos (el agua doblemente marcada, la calorimetría directa, la calorimetría indirecta) son caros y prácticamente se limitan a la investigación en el laboratorio. Por lo tanto, en los últimos años, se han desarrollado diferentes dispositivos de medición objetiva que son apropiados para los estudios de campo y clínicos. No hay ningún estándar de oro entre ellos, ya que todos tienen limitaciones. Los podómetros son ligeros, poco costosos, cuentan los pasos y aportan información sobre la actividad física total, pero no sobre el comportamiento y los patrones de actividad física. Los acelerómetros son caros, aportan información sobre patrón, frecuencia e intensidad de la actividad física, pero no sobre el tipo de actividad física. Los podómetros y acelerómetros únicamente recogen información sobre el movimiento del movimiento corporal, pero la validez en la estimación del gasto energético es limitada. La monitorización de la frecuencia cardíaca relaciona intensidad del ejercicio con gasto de energía, pero no aporta información sobre la actividad física. Los dispositivos GPS son portátiles, relativamente asequibles, no invasivos y recogen distancia, velocidad y elevación con hora y lugar exactos, pero quizás estén limitados para la evaluación de movimientos cortos de alta intensidad y elevado gasto energético. Los dispositivos de última generación combinan acelerometría con la medición de variables fisiológicas, comparten las ventajas de los dispositivos individuales y son más precisos. Para el cálculo del gasto energético se aplican algoritmos específicos de la actividad incluidos en el software del fabricante que pueden afectar a los resultados. La mayoría de los dispositivos estiman con mayor precisión el gasto energético a intensidades ligeras y moderadas, pero subestiman el gasto a intensidades muy ligeras y de mayor intensidad.