Shvorob I. Methods and means of extraction and analysis of poorly structured text data based on a document-oriented graph

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0418U001342

Applicant for

Specialization

  • 10.02.21 - Структурна, прикладна та математична лінгвістика

15-03-2018

Specialized Academic Board

Д 35.052.05

Lviv Polytechnic National University

Essay

The dissertation solved the problem of developing technologies for extraction, storage, processing and analysis of semistructured data. The notion of a document-oriented graph for the presentation of semistructured text-to-speech texts was introduced, which enabled the use of graph theory to establish links between elements of the document and determine the relationship between the document and the template. For the first time, a method of initial analysis of data has been developed, which allows to partially structure the natural language text for its further elaboration.

Files

Similar theses