Towards a Benchmarking Framework for Financial Text Mining
Caslav Bozic, Ryan Riordan, Detlef G. Seese, Christof Weinhardt · 2010
Different data mining methods for financial texts and various sentiment measures are described in the existing literature, without common benchmarks for comparing these approaches. The framework proposed in this paper and the corresponding implemented system facilitate combining more sources of financial data into comprehensive integral dataset. The use of the dataset is then illustrated by analyzing the candidate measures by estimating parameters of regression on different returns and other financial indicators that can be defined using system’s novel data transformation approach.