Creating a text analysis tool for socio-humanitarian research. Part 2. RSA machine and the experience of using it

The second part of the work describes the most known tools for linguistic-statistical analysis of text corpuses and introduces RSA machine - a novel text analysis tool for socio-humanitarian research. This tool works with network representation of text and allows finding the constructions with complex graph structure in texts. RSA machine implements following features: search of constructions by query, computation of frequencies and statistical characteristics for search results, corpora or individual texts, comparing texts using statistical and frequency features. This paper describes the RSA machine architecture and developing tools. We present the results of pilot research of RSA machine using 142 texts examples written by people with different psychology and demographic characteristics. Some of them (18) were diagnosed with mental disorder. The performed correlation analysis revealed some relations between extracted texts attributes (e.g. frequency of predicate types) and results of psychological analysis performed by experts.