Some Means of Processing Electronic Text Documents

Abstract

Introduction. Digitization of legislation is an important area today, which is identified by the government as a priority. Creating digital legal documents and verifying them for compliance with the law is a necessary task in all areas of jurisprudence. This sets the task of automatic formalizing a legal document created as an arbitrary text in natural language. Purpose. Preparing a document for storage in digital format for further processing may require prior work with an original text. When using automatic means of linguistic analysis of the texts submitted in natural language, in particular, legal, which processes the text in sentences (working up the text sequentially sentence by sentence), problems of local and global nature arise. The problem of local nature is created, in particular, by the presence in the text of the sentences, which due to their considerable length are difficult to process (with the help of one or another tool of text analysis). The problem of a global nature arises when the semantic connection between the components of different sentences should be taken into account during the automatic processing of the text. The purpose of this work is to develop means for overcoming these problems. Results. A model for structuring long sentences containing enumerations as well as a method for eliminating the synonymy of object names referred to in the text, which is intended for automatic analysis, has been developed. Conclusion. Marking up sentences containing enumerations is useful, especially when the text is intended for analysis using a procedure that processes the text sentence by sentence. Structuring a sentence with an enumeration enables, on the one hand, to prepare the sentence for processing in parts, and on the other hand, not to lose the integrity of the sentence when processing in parts. In the method of eliminating the synonymy of names proposed in this paper, both the step of identifying the names of objects and the step of revealing the identity of names requires semantic analysis. To control the correctness of these steps, Oracle was introduced to improve the reliability of the result.