Detection of temporality at discourse level on financial news by combining Natural Language Processing and Machine Learning

• DOI: 10.1016/j.eswa.2022.116648
• Publication Date: 2022-02-24T18:28:26Z
• Authors: Silvia García-Méndez; Francisco de Arri...; Ana Barros-Vila; Francisco J. Gonz...
• Keywords: Temporality

Finance-related news such as Bloomberg News, CNN Business and Forbes are valuable sources of real data for market screening systems. In news, an expert shares opinions beyond plain technical analyses that include context political, sociological cultural factors. the same text, often discusses performance different assets. Some key statements mere descriptions past events while others predictions. Therefore, understanding temporality in a text is essential to separate information from We propose novel system detect finance-related at discourse level combines Natural Language Processing Machine Learning techniques, exploits sophisticated features syntactic semantic dependencies. More specifically, we seek extract dominant tenses main statements, which may be either explicit or implicit. have tested our on labelled dataset annotated by researchers with knowledge field. Experimental results reveal high detection precision compared alternative rule-based baseline approach. Ultimately, this research contributes state-of-the-art identifying predictive financial decision making.