Open AccessAutomation of the data processing procedure on the basis of a strong semantic connection
Author(s) -
Марина Лапшина,
Марина Лапшина,
Надежда Юдина,
Nadezhda Yudina,
Т. Богомолова,
Т. Богомолова,
А. Бойкова,
A. Boykova
Publication year - 2018
Publication title -
lesotehničeskij žurnal
Language(s) - English
Resource type - Journals
ISSN - 2222-7962
DOI - 10.12737/article_5a3d0376d9d021.80176334
Subject(s) - computer science , automation , task (project management) , identification (biology) , object (grammar) , process (computing) , feature (linguistics) , data mining , quality (philosophy) , similarity (geometry) , software , information retrieval , basis (linear algebra) , artificial intelligence , systems engineering , engineering , image (mathematics) , programming language , mechanical engineering , geometry , mathematics , linguistics , philosophy , botany , epistemology , biology
One of the defining tasks in automating the processing of current information obtained in the process of object research is the possibility of assigning it to specific objects located in the database (DB) of the information system. Such a task can be represented as a task of identifying the objects of analysis (monitoring) on current data. Initial information about a particular facility during the research can come from different sources and be incomplete and inaccurate. Due to identification, received data for each object are accumulated in the database for further automated analysis. But in a number of cases, the current information contains not specific names of the research objects of interest, but only some signs of their functioning: location, operating time, numerical characteristics of objects, etc. In such conditions, the identification of objects can occur on the basis of similarity of semantic features contained in the current information and attributes which have been already accumulated by the objects of research in the database. A distinctive feature of the subject area in which the analysis of objects is made is its variability; therefore, with the passage of time, the attributes serving to identify the objects may change. The addition of new features will result in a change in application software and will require the involvement of developers, which leads to additional time costs and decrease in the speed of procedure for identifying objects. This reason leads to the need to create a new approach to the automation of object monitoring, based on such requirements as high quality of the decisions made and ability to quickly change the logic of identification in the process of the system. The paper proposes the possibility of using predicate logic formulas, with the aim of making new information in the database more quickly and assigning it to a specific object, which makes it possible to develop the fastest solution.