Program for Construction of Training Treebank for System of Multipurpose Analysis of Russian-Language Text Messages

<p>Program for Construction of Training Treebank for System of Multipurpose Analysis of Russian-Language Text Messages</p><p>The major objectives of this work are to analyze and compare existing Natural Language Processing tools, including Russian language, and to develop the program, which will utilize these tools. &nbsp;The main purpose of the program is to construct training treebanks and allow experts to annotate nodes of the semantic dependency tree of the database sentences.</p><p>The low number and poor quality of the existing tools for the analysis of Russian text make this research relevant.</p><p>First chapter covers the existing Russian corpus, instruments for morphological analysis and syntactic analysis of text. This chapter also outlines main ways to create new instruments. The conducted experiments and comparison of the existing tools put forward suggestions on which of these tools to use during the development of the program.</p><p>The second chapter describes the technologies used during the development of the program and explains the reason why they were chosen. In addition, this chapter highlights the architecture of the program and its pros and cons. This part also describes data models that lay behind the program.</p><p>Number of pages: 69; pictures: 5; tables: 6; appendices: 3; references: 25</p><p>Keywords: morphological analysis, syntactic analysis, parsing, corpus linguistics, machine learning.</p>