Adaptable Text Corpus Development for Specific Linguistic Research

Thesis deals with automated corpus development techniques for specific linguistic research purposes. Different corpora and corpus toolboxes available today have been overviewed. Corpus tool functionality for processing custom and private text purposes and automated corpora building system capabilities has been classified. Lack of analytical means which decrease effectiveness, multifunctionality and adaptability in overviewed corpora has been illustrated (impossibility of FrameNet usage, lack of detailed statistics and other relevant issues). Main stages of source data text processing have been reviewed. Natural language processing and natural language understanding methods and tools for adaptable corpora building have been examined and justified. General workflow and design of adaptable text corpus software tool are suggested. Adaptable text corpus software prototype for specific linguistic research purposes has been developed. Big Data and text search engine approaches have been implemented in the prototype.