Wyniki wyszukiwania - BazTech

1

Distinguishing between paradigmatic semantic relations across word classes : human ratings and distributional similarity

Schulte im Walde Sabine

Journal of Language Modelling

|

2020

|

Vol. 8, No. 1

53--101

EN

This article explores the distinction between paradigmatic semantic relations, both from a cognitive and a computational linguistic perspective. Focusing on an existing dataset of German synonyms, antonyms and hypernyms across the word classes of nouns, verbs and adjectives, we assess human ratings and a supervised classification model using window-based and pattern-based distributional vector spaces. Both perspectives suggest differences in relation distinction across word classes, but easy vs. difficult class-relation combinations differ, exhibiting stronger ties between ease and naturalness of class-dependent relations for humans than for computational models. In addition, we demonstrate that distributional information is indeed a difficult starting point for distinguishing between paradigmatic relations but that even a simple classification model is able to manage this task. The fact that the most salient vector spaces and their success vary across word classes and paradigmatic relations suggests that combining feature types for relation distinction is better than applying them in isolation.

2

Antyscam : Practical Web Spam Classifier

Luckner Marcin, Gad Michał, Sobkowiak Paweł

International Journal of Electronics and Telecommunications

|

2019

|

Vol. 65, No. 4

713--722

EN

To avoid of manipulating search engines results by web spam, anti spam system use machine learning techniques to detect spam. However, if the learning set for the system is out of date the quality of classification falls rapidly. We present the web spam recognition system that periodically refreshes the learning set to create an adequate classifier. A new classifier is trained exclusively on data collected during the last period. We have proved that such strategy is better than an incrementation of the learning set. The system solves the starting–up issues of lacks in learning set by minimisation of learning examples and utilization of external data sets. The system was tested on real data from the spam traps and common known web services: Quora, Reddit, and Stack Overflow. The test performed among ten months shows stability of the system and improvement of the results up to 60 percent at the end of the examined period.

3

Klasyfikacja sekwencji wizyjnych do określonych Uogólnionych Klas Użytkowania na podstawie wielkości celu i poziomu oświetlenia

Witkowski M., Leszczuk M., Dziech A.

Przegląd Telekomunikacyjny + Wiadomości Telekomunikacyjne

|

2012

|

nr 8-9

646-650

PL

Przedstawiono badania mające na celu opracowanie algorytmów automatycznie klasyfikujących sekwencje wizyjne do wybranych Uogólnionych Klas Użytkowania (Generalized Use Classes - GUC) na podstawie wielkości obiektu i poziomu oświetlenia. Opisano badania przeprowadzone na ekspertach, za pomocą specjalnie stworzonego interfejsu oraz analizę uzyskanych wyników. Szczegółowo omówiono proponowany algorytm klasyfikujący sekwencje wizyjne oraz poddano go ocenie. Prace opisane w artykule były realizowane w ramach projektu INDECT.

EN

This paper presents the research devoted to development of algorithms for automatic classification of video sequences into specified Generalized Use Classes of target size and lighting level. The experiment conducted on the experts with a specially created interface, and the analysis of the obtained results are described first. Next, the developed algorithm that classifies video sequences, and its evaluation arę discussed. The research described in this paper was conducted under the INDECT project.

4

Matrix replica method and artificial neural networks as a component of condition assessment of materials for the power industry

Sroka M., Zieliński A.

Archives of Materials Science and Engineering

|

2012

|

Vol. 58, nr 2

130--136

EN

Purpose: use of matrix replica method and artificial neural networks in condition assessment of materials working under creep conditions. Demonstration of scanning electron microscopy usefulness for correct assessment of the occurrence of internal damages. Design/methodology/approach: For material after long-term service the microstructural investigations were carried out on metallographic specimens and matrix replicas using light and scanning electron microscope. A computer program was used for material condition assessment. Findings: The way of material condition assessment by matrix replica method using a computer program as an example of the application of computer materials science methods was presented. The correct development of methodology for assessment of the degree of internal damages in steel working under creep conditions was found a result of verification. Critical comments on condition assessment of structure observed using light microscopy were presented. Practical implications: The presented method can be used for evaluation and qualification of structural changes in power station boiler components operating in creep conditions. Originality/value: The presented results of changes in the mechanical properties, structure and in the precipitation processes are applied to evaluation the condition of the elements in further industrial service.

5

Optymalizacja zróżnicowania wartości wymiarów w systemach modułowych konstrukcji

Gendarz P.

Mechanik

|

2010

|

R. 83, nr 1/S

43-50

PL

Najwyższą formą uporządkowania rodziny konstrukcji są systemy modułowe konstrukcji. W celu ich racjonalnego tworzenia proponuje się, by zastosować optymalizację zróżnicowania wartości wymiarów, z zachowaniem relacji sprzężeń między współdziałającymi elementami. Podstawą jej realizacji jest automatyczna klasyfikacja (Cluster analyse). Spośród różnych metod automatycznej klasyfikacji wybrano hierarchiczną i iteracyjną, których celem jest utworzenie optymalnego układu klas. Kryterialny wybór reprezentantów klas pozwala tworzyć systemy modułowe konstrukcji złożonych środków technicznych ze wspomaganiem komputerowym.

EN

Basic problem of design in a way of production process is to assure serial manufacture. New requirements for construction process are oriented on using the newest solutions for design and also on putting the widest spectrum needs for specified class of technical means. It is possible by using modulus systems of design. In this written scientific issue you can find a trial of showing how to assure serial manufacture through aggregation of measurements values. Aggregation influences the amount of design; it means production of the biggest number of composition limited by design modulus.

6

Application of Monte Carlo Preliminary Image Analysis and Classification Method for Automatic Reservation of Parking Space

Okarma K., Lech P.

Machine Graphics and Vision

|

2009

|

Vol. 18, No. 4

439-452

EN

The paper presents a Monte Carlo-based method for preliminary image analysis and classification applied to parking space reservation. The proposed approach can be used especially for large parking areas with high density of parking places, where automatic recognition of some geometrical parameters of the entering vehicles can be helpful for optimised usage of the available parking area. After performing real-time classification of the entering vehicles, the shortest path towards the reserved parking place can be presented for the driver e.g. on some display screens. This requires appling a fast classification method which can be easily implemented in embedded systems instead of some typical classification algorithms characterised by relatively high computational complexity. The Monte Carlo approach presented in the paper can be used as an efficient solution of this problem.

7

An automatic method to identify human alternative movements: application to the ingress movements

Ait El Menceur M. O., Pudlo P., Debril J. F., Lepoutre F. X.

Journal of Automation Mobile Robotics and Intelligent Systems

|

2009

|

Vol. 3, No. 3

64-69

EN

Alternative movement techniques identification is primordial in many studies of biology, medicine and ergonomics. Few studies in literature propose quantitative techniques to achieve this purpose. Park et al. [ W. Park, B.J. Martin, S. Choe, D.B. Chaffin, M.P. Reed, "Representing and identifying alternative movement techniques for goal-directed manual tasks", Journal of Biomechanics, vol. 38 , 2005, pp. 519-527] proposed a biomechanical index based technique named JCV. This method finds its limits when dealing with three-dimensional complex movements. In the present study we propose a modification of this method so that it will be applied to complex movements. We consider a non-habitual end effector (Midpoint between Hips). We obtain other indices, 3BJCV, to which a non-supervised clustering technique is applied. We have applied our method to the ingress movements of 37 young and elderly subjects with or without prosthesis entering in a minivan vehicle. Our method allows the identification of the two big classes of ingress movements observed by Ait El Menceur et al. [M-O. Ait El Menceur, P. Pudlo, P. Gorce, A. Thévenon, and F-X, Lepoutre, "Alternative Movement Identification in the Automobile Ingress and Egress for Young and Elderly Population with or without Prostheses", International Journal of Industrial Ergonomics, vol. 38, 2008, pp. 1078-1087].

8

Classifying images of vehicles with the use of mathematical model ling

Piecha J., Wróbel K.

Magazyn Autostrady

|

2006

|

Nr 4

92--96

EN

The paper presents a method of recognizing and classifying images of vehicles with the use of a group of features selected on the basis of mathematical models by Hough [5].

9

International Comparison of Occupational Safety and Health Research - A Review Based on Published Articles

Brueggmann M., Roetting M., Luczak H.

International Journal of Occupational Safety and Ergonomics

|

2001

|

Vol. 7, No. 4

387--401

EN

A comparison of international literature in occupational safety and health (OSH) research over the years from 1980 to 1998 was conducted. The comparison is based on the different languages of the over 35,000 articles, which were collected from different databases. The distribution of languages in the analyzed data has to be considered carefully with respect to the original purpose of the data collection and the specific role of the international publication language, English. The comparison shows that in several aspects OSH research does not differ in different publication languages. In some methodological aspects, and in specific objects of research, differences between articles published in different languages can be found.