Scientific Journals and Yearbooks Published at SAS

Article List

Computing and Informatics


Volume 29, 2010, No. 2

Content:


  Identification of Technical Journals by Image Processing Techniques
P.-Ch. Wen, L.-L. Wang

Journal identification, automatic journal registration, hidden Markov model, layout structure, journal title

The emphasis of this study is put on developing an automatic approach to identifying a given unknown technical journal from its cover page. Since journal cover pages contain a great deal of information, determining the title of an unknown journal using optical character recognition techniques seems difficult. Comparing the layout structures of text blocks on the journal cover pages is an effective method for distinguishing one journal from the other. In order to achieve efficient layout-structure comparison, a left-to-right hidden Markov model (HMM) is used to represent the layout structure of text blocks for each kind of journal. Accordingly, title determination of an input unknown journal can be effectively achieved by comparing the layout structure of the unknown journal to each HMM in the database. Besides, from the layout structure of the best matched HMM, we can locate the text block of the issue date, which will be recognized by OCR techniques for accomplishing an automatic journal registration system. Experimental results show the feasibility of the proposed approach.

Computing and Informatics. Volume 29, 2010, No. 2: 163-182.

 
  An Efficient Genetic Algorithm for Solving the Multi-Level Uncapacitated Facility Location Problem
M. Marić

Facility location, genetic algorithms, evolutionary approach

In this paper a new evolutionary approach for solving the multi-level uncapacitated facility location problem (MLUFLP) is presented. Binary encoding scheme is used with appropriate objective function containing dynamic programming approach for finding sequence of located facilities on each level to satisfy clients' demands. The experiments were carried out on the modified standard single level facility location problem instances. Genetic algorithm (GA) reaches all known optimal solutions for smaller dimension instances, obtained by total enumeration and CPLEX solver. Moreover, all optimal/best known solutions were reached by genetic algorithm for a single-level variant of the problem.

Computing and Informatics. Volume 29, 2010, No. 2: 183-201.

 
  Application of Weighted Voting Taggers to Languages Described with Large Tagsets
M. Kuta, J. Kitowski, W. Wojcik, M. Wrzeszcz

Part-of-speech tagging, combination tagger, weighted probability distribution voting tagger, TagPair tagger

The paper presents baseline and complex part-of-speech taggers applied to the modified corpus of Frequency Dictionary of Contemporary Polish, annotated with a large tagset. First, the paper examines accuracy of 6 baseline part-of-speech taggers. The main part of the work presents simple weighted voting and complex voting taggers. Special attention is paid to lexical voting methods and issues of ties and fallbacks. TagPair and WPDV voting methods achieve the top accuracy among all considered methods. Error reduction 10.8 % with respect to the best baseline tagger for the large tagset is comparable with other author's results for small tagsets.

Computing and Informatics. Volume 29, 2010, No. 2: 203-225.

 
  Dialogue Act Recognition Approaches
P. Kral, Ch. Cerisara

Bayesian approaches, dialogue act, lexical information, prosody, syntactic information

This paper deals with automatic dialogue act (DA) recognition. Dialogue acts are sentence-level units that represent states of a dialogue, such as questions, statements, hesitations, etc. The knowledge of dialogue act realizations in a discourse or dialogue is part of the speech understanding and dialogue analysis process. It is of great importance for many applications: dialogue systems, speech recognition, automatic machine translation, etc. The main goal of this paper is to study the existing works about DA recognition and to discuss their respective advantages and drawbacks. A major concern in the DA recognition domain is that, although a few DA annotation schemes seem now to emerge as standards, most of the time, these DA tag-sets have to be adapted to the specificities of a given application, which prevents the deployment of standardized DA databases and evaluation procedures. The focus of this review is put on the various kinds of information that can be used to recognize DAs, such as prosody, lexical, etc., and on the types of models proposed so far to capture this information. Combining these information sources tends to appear nowadays as a prerequisite to recognize DAs.

Computing and Informatics. Volume 29, 2010, No. 2: 227-250.

 
  Higher-Order Attribute Semantics of Flat Declarative Languages
P. Grigorenko, E. Tyugu

Higher-order attribute models, flat languages, attribute semantics of declarative languages, synthesis of programs, domain specific languages

A technique is described that provides a convenient instrument for implementation of semantics of simple declarative languages called flat languages. Semantics of a specification is defined in the paper as a set of programs derivable for solvable goals. We introduce higher-order attribute models that include more control information than conventional attribute models and explain the algorithm for dynamic evaluation of attributes on these models. A visual tool CoCoViLa is briefly described as an instrument for implementing attribute semantics of flat languages.

Computing and Informatics. Volume 29, 2010, No. 2: 251-280.

 
  Action Recognition Using Visual-Neuron Feature of Motion-Salience Region
N. Li, D. Xu, L. Liu

Action recognition, shape-based neurobiological approach (SBNA), motion-salience region, visual-neuron template, visual-neuron feature, visual cortex

This paper proposes a shape-based neurobiological approach for action recognition. Our work is motivated by the successful quantitative model for the organization of the shape pathways in primate visual cortex. In our approach the motion-salience region (MSR) is firstly extracted from the sequential silhouettes of an action. Then, the MSR is represented by simulating the static object representation in the ventral stream of primate visual cortex. Finally, a linear multi-class classifier is used to classify the action. Experiments on publicly available action datasets demonstrate the proposed approach is robust to partial occlusion and deformation of actors and has lower computational cost than the neurobiological models that simulate the motion representation in primate dorsal stream.

Computing and Informatics. Volume 29, 2010, No. 2: 281-301.

 
  Improving Computer Based Speech Therapy Using a Fuzzy Expert System
O.A. Schipor, S.G. Pentiuc, M.D. Schipor

Computer based speech therapy, fuzzy expert system, personalized therapy, exercises for dyslalia

In this paper we present our work about Computer Based Speech Therapy systems optimization. We focus especially on using a fuzzy expert system in order to determine specific parameters of personalized therapy, i.e. the number, length and content of training sessions. The efficiency of this new approach was tested during an experiment performed with our CBST, named LOGOMON.

Computing and Informatics. Volume 29, 2010, No. 2: 303-318.

 
  The Data Fusion Grid Infrastructure: Project Objectives and Achievements
L. Hluchy, N. Kussul, A. Shelestov, S. Skakun, O. Kravchenko, Y. Gripich, P. Kopp, E. Lupian

Environmental applications, Earth science, research infrastructure, grid computing, Earth observations

This paper describes the objectives and achievements of the project "Data Fusion Grid Infrastructure'' jointly supported by INTAS, the Centre National d'Etudes Spatiales (CNES) and the National Space Agency of Ukraine (NSAU). Within the project, a Grid infrastructure has been developed that integrates the resources of several geographically distributed organizations. The use of Grid technologies is motivated by the need to make computations in the near real-time for fast response to natural disasters and to manage large volumes of satellite data. We show the use of developed Grid infrastructure for a number of applications that heavily rely on Earth observation (EO) data. These applications include: numerical weather prediction (NWP), flood monitoring, biodiversity assessment, and crop yield prediction.

Computing and Informatics. Volume 29, 2010, No. 2: 319-334.