Catálogo de publicaciones - libros

This article will describe the Web Modelling Language (WebML), a notation for visually designing intelligent Web application at the conceptual level. All the concepts of WebML are specified both graphically and in the XML standard. WebML defines four orthogonal dimensions: structural model, hypertext model (splits on composition model and navigational model), presentation model and personalisation model. All models enable a high-level intelligent approach to designing and maintaining Web site.

This paper describes a text understanding system for conversational agents. The system resolves zero, direct and indirect anaphors in Japanese texts by integrating two sorts of linguistic resources: a hand-annotated corpus with various relations and automatically constructed case frames. The corpus has relevance tags which consist of predicate-argument relations, relations between nouns and coreferences, and is utilised for learning parameters of the system and testing it. The case frames are indispensable knowledge both for detecting zero/indirect anaphors and estimating appropriate antecedents. Our preliminary experiments showed promising results.

Algebra of ordered fuzzy numbers (OFN) is defined to handle with fuzzy inputs in a quantitative way, exactly in the same way as with real numbers. Additional two structures: algebraic and normed (topological) are introduced to define a general form of defuzzyfication operators. A useful implementation of a Fuzzy Calculator allows counting with the general type membership relations.

We present a new approach to the grid technology which is based on updateable views. Views are used in two ways: (1) as wrappers of local servers which adopt local schemata to the federated database requirements; (2) as a facility for intelligent data integration and transformation into a canonical form according to the federated database schema. Views deliver virtual updatable objects to global clients. Objects can be associated with methods which present the procedural part of remote services, like e.g., in Web Services. The fundamental quality of the approach is transparency of servers: the user perceives the distributed data/service environment as an integrated virtual whole. Such a quality is achieved by applications based on CORBA. We attempt to achieve a higher level of transparency by providing means for integrating horizontal and vertical fragmentations of data and by taking into account various forms of data redundancy. The approach is based on a very simple and universal architecture and on the stack-based approach (SBA) to object-oriented query languages.

The paper presents a concept of interactive system Vis3D for realistic real terrain visualisation based on Geographical Information Systems (GIS) data. Taking account of complexity and inaccuracy of geographical data this system has to work in two phases: off-line and on-line. The former off-line phase should convert data from GIS format to a form adapted for rendering algorithms (the intermediate form of data is called Terrain Model for Visualisation TMV). Using this form the latter on-line phase can generate a realistic image efficiently. This paper describes both phases, indicates difficulties in their realisation and shows the ways for overcoming them.

This paper presents basic conceptions and classification of a data acquisition system for multimedia and robotic applications. A DAQ system designed for humanoid head robot built at PJIIT Robotics and Multiagent Systems Laboratory is described. The main goal of the system was to provide robust data acquisition architecture that would reliably provide data to processing modules. A real time operating system QNX Neutrino was used together with a dedicated hardware driver to provide a low overhead, high throughput data acquisition system. The driver architecture was determined by the environment it was operating in and software it interfaced with.

SpeeCon Polish Speech Database was collected within the framework of the SpeeCon project partially sponsored by the EC (IST-1999-10003). The database contains two sets of data, which comprise 550 adults’ recording sessions and 50 sessions from children, respectively. The adult speakers were recorded in various environments: offices, living rooms, cars and public places. Recordings contain free spontaneous speech passages, elicited spontaneous speech, phonetically compact words and sentences, general-purpose words and phrases, specific application words and utterances. One of the most important problems in the construction of the database is to define bases for multi-level transcription composed of several tiers. They could be grouped into three classes – linguistic, symbolic and physical representation. The orthographic transcription is applied to the sentence, phrase and word tiers, symbolic transcription related to grammar and articulation – to part of speech, phoneme and syllabic tiers and mnemonics – to the description of some characteristic of the measurable physical data. The paper presents the rules applied to text, speech and noise transcriptions and remarks on pronunciation varieties found in the database. The final part of the paper discusses the problem of the lexicon creation, which is an alphabetically ordered list of distinct lexical items occurring in the recorded corpus. The Polish lexicon has been built up by various methods, including hand-annotation and generation by rule with subsequent manual check.

The paper presents a method for intelligent automatic processing of medical reports. First, we extract single pieces of information using SProUT (a general-purpose Information Extraction platform), and then, externally merge the results in order to obtain a detailed formalised description of the reports.

In this paper, we will present an evaluation experiment of Stream Public Opinion Channel (SPOC) from contents creators perspective. This system allows a contents creator to publish programs using not only multimedia but also conversational information in sync with interface agent and multimedia. Contents creator of SPOC needs to make a story with consideration of agent ç-’ actions and their effect on audience. Therefore, it is important to evaluate the effect of interface agents on contents creators. Results of the experiment showed that contents creators positively accepted the SPOC in subjective evaluation, but they could not successfully embody what they want to express as a story. Additionally, we found correlations between the impressions of interface agent and factors which would form users behavioural intention. We need to measure not only subjective evaluation but also objective indicator in evaluation.

Character agents have become more popular in the Internet due to its attractiveness. This paper proposes an agent animation generation system which automatically selects agent behaviours as well as highlighting animations to emphasise the agent actions. In order to produce appropriate animations according to the content of agent’s spoken message, our system, first analyses the message text using natural language processing engine, and then selects animations based on the linguistic information calculated by the engine.