Open Access

Katsiaryna Stalpouskaya, Christian Baden

• Text is one of the key sources of information for social sciences and humanities which, with the rise and development of computational technologies, has been mostly available via digital libraries, archives and websites. It enables researchers to increasingly deal with large scale text corpora that require the use of advanced software tools to process them and extract information. Computational linguistics - a discipline that has emerged on the border of computer science, linguistics and statistics - has achieved certain results in automated text analysis and information extraction, e.g., tools for part-of-speech tagging, grammar parsing, semantic role labelling, sentiment analysis and anaphora resolution have been developed and successfully used in many scientific projects. However, there still exists a gap between technology available and the needs of social sciences: named entity recognizers are incapable of identifying actors, sentiment analysis just provides the overall mood of an expression but is not able to identify the evaluation of information by the utterer, topic modeling tools can only assign a topic to a document, but fall short of measuring its frame.