Use of domain ontology for homonymy clarification into the natural language texts

• Дата: 2018
• Автори: Lesko, O.N., Rogushina, J.V.
• Формат: Стаття
• Мова: rus
• Опубліковано: Інститут програмних систем НАН України 2018
• Теми: homonymy; morphological analysis; syntactical analysis; natural language processing; ontology; UDC 004.853; 004.55
• Онлайн доступ: https://pp.isofts.kiev.ua/index.php/ojs1/article/view/320
• Назва журналу: Problems in programming

The article analyses the clarification of various types of homonymy that can be executed without use of semantic information, but only on the basis of syntactic rules. This analysis shows how features of the syntactic structures of legislative and academic texts allow to reduce the number of formal rules required for parsing. A minimal set of syntactic rules necessary for the automatic analysis of such texts is proposed. A method of homonymy clarification in natural language business, scientific and legal text documents is developed. Proposed method does not require the use of a large number of syntactic rules and marked-up texts. Such specificity greatly simplifies the implementation and reduces the time required for creation and markup of text corpora. This result is achieved by use of domain ontology, and by the specifics of syntactic structures of business, scientific and legal documents. In addition, we demonstrate how the use of domain ontology allows to simplify the analysis of the test documents. As opposed to other systems of automatic processing of natural language texts that use domain ontology for semantic analysis too the domain ontology is used to highlight terms in the text and further morphological information of each word in wordy terms.Problems in programming 2017; 2: 61-71