Catálogo de publicaciones - libros

One of the major problems that might hinder the construction of the knowledge society on the information network is what I call the understanding and communication bottlenecks, which might be caused by the limitation of human cognitive capability. In this paper, I present Communicative Intelligence as a step towards solving the understanding and communication bottleneck by inventing communicative artifacts that enable people and artifacts to interact with each other in a natural fashion. I focus on conversational communications in particular, for conversation is the most natural means for communication. I believe that making conversation-rich community contributes a lot to resolve the understanding and communication bottlenecks. Intelligent media technology aims at inventing communicative artifacts which allow people and artifacts to interact with each other in a natural fashion and thereby enable conversation-rich knowledge society. It consists of five subfields: conversation measurement and analysis, conversational artifacts, conversational environment design, conversational contents, and applied conversational systems. I will overview major results obtained in each subfield.

This paper reports our research toward enhancing the ‘User Involvement’ in human-computer interaction, which is introduced to grasp the idealised state in which a natural communication between a computer and its user should be established. Moreover, we also introduce a new idea of ‘Empathy Channel’, through which humans can interact with virtual agents and objects on computers with more sense of reality. The main issue of this paper is to clarify various methods for establishing an Empathy Channel in building a human-computer interface. We believe a good design of human-computer interface with an Empathy Channel is one of the best methods to enhance the User Involvement.

Although considerable work on named-entity recognition for few major languages exists, research on this topic in the context of Slavonic languages has been almost neglected. This paper presents a rule-based named-entity recognition system for Polish built on top of SProUT, a novel multi-lingual NLP platform. We pinpoint the encountered difficulties and present some promising evaluation results.

The term was introduced as a generic term by Polkowski and Skowron in the seminal paper that laid foundations for Rough Mereology – a paradigm for Approximate Reasoning that combines ideas of Mereology – a set theory based on the notion of a part – with ideas of Rough Set Theory and Fuzzy Set Theory; in particular, its basic predicate of rough inclusion is a rendering of the notion of being a part to a degree. Rough Mereology is an approach towards constructing reasoning schemes that take into account uncertainty of either knowledge or concepts used in reasoning. This abstract reasoning methodology is therefore a constituent of the vast field of Cognitive Technologies (styled also Artificial Intelligence).

It is well–known that mereological theories of objects have been applied in Spatial Reasoning – reasoning about uncertainty in spatial contexts. The majority of theories based on mereology and applied in reasoning about spatial objects stem from the idea of A. N. Whitehead, viz., Mereology Theory based on the predicate of being connected.

In this article, we give a survey of the current state of the art in spatial reasoning based on constructs of Rough Mereology. We include here theoretical results – some of them already shown in earlier works – that witness applicability of constructs based on rough inclusions in spatial reasoning as well as we mention recent works on practical applications to real–world robot navigation.

This paper discusses a concept of an intelligent self-configuration of sensors’ clusters in a grid network as an overlay on contemporary grid technologies for processing of the sensory information locally and collaboratively. An approach for information exchange among sensors with utilisation of sensor agents for later service discovery in sensor mesh is proposed. The main task of sensor agents is to localise themselves in information space from the point of view of captured sensory data and not only a geographical location, which is often unknown to the sensors. The experiments were conducted in a large scale lecture room, where several microphones and cameras were installed in order to capture students’ audiovisual activities.

In the context of intelligent communication for distance learning, the paper describes the 3D face modelling framework based on advanced face recognition descriptor for a 2D facial image and on a 3D eigenfaces concept.

We present a graphically embodied animated agent (a virtual speaker) capable of reading a plain English text and rendering it in a form of speech accompanied by the appropriate facial gestures. Our system uses a lexical analysis of an English text and statistical models of facial gestures in order to automatically generate the gestures related to the spoken text. It is intended for the automatic creation of the realistically animated virtual speakers, such as newscasters and storytellers and incorporates the characteristics of such speakers captured from the training video clips. Our system is based on a visual text-to-speech system which generates a lip movement synchronised with the generated speech. This is extended to include eye blinks, head and eyebrow motion, and a simple gaze following behaviour. The result is a full face animation produced automatically from the plain English text.

This paper describes a medium, called , for helping children to understand contents from the Web. It works by transforming electronic contents into an easily understandable “storybook world”. In this world, easy-to-understand contents are created generated by creating 3D animations that include contents and metaphors, and by using a child-parent model with dialogue expression and a question-answering style comprehensible to children.

This paper describes our attempt to build a communicative medium for capturing and re-experiencing conversations situated in the real space. We first show a system that captures and interprets conversation scenes by ubiquitous sensors. Based on the system, we present three approaches to visualise and facilitate users to access the extracted conversation scenes, i.e., chronological summarisation of videos, spatio-temporal collage of videos, and ambient sound display.

The present paper aims to analyse some difficult problems in man-machine dialogues and to draft solutions based on the idea of Dialogue Processing Memories. The idea behind the well known technique of translation memories is to economise effort by re-use of previously translated fragments. This could be adapted to other text processing domains, in particular to the field of man-machine dialogue. Simulation of correct human behaviour could be based on memorising the dialogue fragment or, more adequately, the results of its processing. The mechanism of extraction from the memory should be triggered by the same element which had earlier triggered the memorisation mechanism. Each case of making such use of Dialogue Processing Memory may be stored for possible future use with a memory refreshment/oblivion mechanism.