Open AccessExperimental investigations of text graph representation model efficiency for borrowings detection
Author(s) -
Kuropiatnyk Olena
Publication year - 2020
Publication title -
ìnformacìjnì tehnologìï v metalurgìï ta mašinobuduvannì
Language(s) - English
Resource type - Journals
ISSN - 2708-0102
DOI - 10.34185/1991-7848.itmm.2020.01.049
Subject(s) - constructive , representation (politics) , graph , computer science , base (topology) , software , theoretical computer science , artificial intelligence , algorithm , data mining , natural language processing , mathematics , programming language , mathematical analysis , process (computing) , politics , political science , law
The paper deals with investigation of time and functional efficiency of the developed software system for text borrowings detection. Base of this system is constructive-synthesizing text graph representation model. The experiment revealed a linear relationship between the time of the text borrowing check operation and the size of the text base for comparison. The conducted experiments showed that there is a difference in the results of the checking text documents by the developed system and analogue. For its estimation the degree of functional efficiency advantage is calculated by similarly to S-estimation of efficiency of algorithms. The reasons for the difference in results are identified. Attention is drawn to obstacles in analogues use for experimental efficacy investigations.