Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
POLISH SEMANTIC PARSER
No full textAmount of information transferred by computers grows very rapidly thus outgrowing the avera- ge man’s capability of reception. It implies Computer programs increase in the demand for which would be able to perform an introductory classification or even selection of information directed to a particular receiver. Due to the complexity of the problem, we restricted it to understanding short ncwspaper notes. Among many conceptions formulated so far, the conceptual dependency worked out by Roger Schank has been chosen. It is a formal language of description of the se- mantics o f pronouncement integrated with a text understanding algorithm.Substantial part of each text iransformation system is a semantic parser of the Polish language. It is a module, which as the first and the only one has an access to the text in the Polish language. It plays the role of an element, which fmds relations between words of the Polish language and the formal registration. It translates sentences written in the language used by people into the langua ge oftheory.The presented structure of knowledge units and the shape of understanding process algorithms are universal by virtue of the theory. On the other hand the defined knowledge units and the rules used in the algorithms are only examples because they are constructed in order to understand short newspaper notes
TRENDS IN MODERN EXCEPTION HANDLING
Get PDFException handling is nowadays a necessary cornponent of error proof Information systems. The paper presents overview of techni ues and models of exception handling, problems con- nected with them and potential Solutions. The aspects of implementation of propagation mechanisms and exception handling, their effect on semantics and genera program efhcJen- cy are also taken into account. Presented mechanisms were adopted to modern programming languages. Considering design area, formal methods and formal verihcation o f program pro- perties we can notice exception handling mechanisms are weakly present what makes a field for futur research
FITTING OF TONAL CURVE AND BALANCING OF GRAY LEVELS IN CONTRAST EXPANSION AND PRELIMINARY IMAGING OF STRUCTURES IN THE ANALYSIS
No full textThe article presents a group of transformations which medical pictures can be subjected to, in the aim of enhancing their contrast and better distinguishing and diagnosis of organs shown in this pictures. In this work the analysis was based on images acąuired by various techniąues, such as roentgenography (RTG), magnetic resonance (MRI), classic Computer tomography (CT) and ultrasonography (USG). Contrast expansion in these images was performed using the histogram mcthod of balancing of gray levels and the method based on the modification of tonal curve. Such transformations were necessary during preliminary processing of these images, especially to adjust their quality for next steps of the analysis in case of automatic medical diagnosis, and also to enable their visua! evaluation by specialists and diagnostic interpretation of images of or gans and their structural features. The studies show that the method based on the modification of tonal curve yields better results in case of images acąuired by roentgenography and magnetic re sonance imaging. On the other hand, histogram method is morę suitable for contrast expansion in CT and USG images
Survey of Scientific Document Summarization Techniques
No full textThe number of scientic or research papers published every year is growing at an exponential rate, which has led to an intensive research in scientic document summarization. The different methods commonly used in automatic text summarization are discussed in this paper with their pros and cons. Commonly used evaluation techniques and datasets in this field are also discussed. Rouge and Pyramid scores of the different methods are tabulated for easy comparison of the results
Tunnel Parsing with counted repetitions
Get PDFThe article describes a new and efficient algorithm for parsing, called Tunnel Parsing, that parses from left to right on the basis of a context-free grammar without left recursion and rules that recognize empty words. The algorithm is applicable mostly for domain-specific languages. In the article, particular attention is paid to the parsing of grammar element repetitions. As a result of the parsing, a statically typed concrete syntax tree is built from top to bottom, that accurately reflects the grammar. The parsing is not done through a recursion, but through an iteration. The Tunnel Parsing algorithm uses the grammars directly without a prior refactoring and is with a linear time complexity for deterministic context-free grammars
NATURAL SOLVERS IN PROBLEMS OF SEARCHING FOR THE BEST SOLUTION
No full textIn the paper we present a new method, which can be used as a natural solver for searching the best solution in the multidimensional and multimodal parameter space. The method is based ona well-known simulation techniąue, i.e., molecular dynamics. To show advantages and disadvanta- ges of the particie method in comparison to the standard genetic algorithm, we analyse efficiency of the methods in finding the global minimum of multi-dimensional and multi-modal test-bed functions and we calculate the evaluation indices. We analyse also the ways the solution space is explored and the parameters of algorithms adjusted. The optimal heuristics are proposed. The tests carried out show that the choice of the most appriopriate optimization method depends on type of a problem considered. We show that the particie method is morę efficient for finding the optimal solution for multi-modal problems with distinct global extreme, while the genetic algo rithm is better for deceptive functions with several locals extreme, which are placed far away from the global optimum. This comes from the different ways in which the particie method and genetic algorithm explore the solution space. The particie method can be used for initial analysis of functions, which character is unknown
TF-IDF Inspired Detection for Cross-Language Source Code Plagiarism and Collusion
Get PDFSeveral computing courses allow students to choose which programming language they want to use for completing a programming task. This can lead to cross-language code plagiarism and collusion, in which the copied code file is rewritten in another programming language. In response to that, this paper proposes a detection technique which is able to accurately compare code files written in various programming languages, but with limited effort in accommodating such languages at development stage. The only language-dependent feature used in the technique is source code tokeniser and no code conversion is applied. The impact of coincidental similarity is reduced by applying a TF-IDF inspired weighting, in which rare matches are prioritised. Our evaluation shows that the technique outperforms common techniques in academia for handling language conversion disguises. Further, it is comparable to those techniques when dealing with conventional disguises
Compression of Convolutional Neural Network for Natural Language Processing
No full textConvolutional Neural Networks (CNNs) were created for image classification tasks. Quickly, they were applied to other domains, including Natural Language Processing (NLP). Nowadays, the solutions based on artificial intelligence appear on mobile devices and in embedded systems, which places constraints on, among others, the memory and power consumption. Due to CNNs memory and computing requirements, to map them to hardware they need to be compressed.This paper presents the results of compression of the efficient CNNs for sentiment analysis. The main steps involve pruning and quantization. The process of mapping the compressed network to FPGA and the results of this implementation are described. The conducted simulations showed that 5-bit width is enough to ensure no drop in accuracy when compared to the floating point version of the network. Additionally, the memory footprint was significantly reduced (between 85% and 93% comparing to the original model)
THE CONSTRAINT DRIVEN METHOD FOR AMBIGUITY RESOLUTION
No full textThcrc are different types of ambiguity in natural language, which causcs different problems in natural language Processing. This paper describcs the ambiguity of inflection form, cg. kamień (the stonc), which is the form of Moninative Singular or Accusativc Singular. The paper also proposes the constraint drivcn method for rcsolving that type of ambiguity
MODEL OF INTERNET SYSTEM CLIENT SERVICE
Get PDFAt present, service of different kinds of internet clients involves necessity of defining their behaviour in various situations. To determine optimal Solutions it is purposeful to know and to describe objectively their behaviour by means of mathematical models. In this article there are presented existing Solutions o f the discussed problem and the own model o f service of internet clients is proposed