A Methodology for Discovering Upstream and Downstream Causal Relationships in Stock Market
Harchana Bhoopathi1, B Rama2

1Harchana Bhoopathi, Department of Computer Science, Kakatiya University, Warangal, Telengana State, India.
2B Rama, Department of Computer Science, Kakatiya University, Warangal, Telengana State, India.

Manuscript received on 16 August 2019 | Revised Manuscript received on 20 August 2019 | Manuscript published on 30 August 2019 | PP: 4448-4490 | Volume-8 Issue-10, August 2019 | Retrieval Number: J10740881019/2019©BEIESP | DOI: 10.35940/ijitee.J1074.0881019
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Causal relationships between events pertaining to stock market have potential to influence the stakeholders of the companies associated with those events. Understanding causal relationships in stock markets help in making intelligent decisions. Traditional prediction approaches cannot estimate the upstream and downstream causal relationships. Therefore, it is inevitable to consider portfolios that exhibit causal relationships. Simple correlation between variables may not reflect causal relationships unless there is an event that is the result of occurrence of another event. Finding upstream and downstream causal relationships is challenging. In the literature it is found that inter-transactional details can help in finding causal relationships. Based on this idea, in this paper, we planneda methodology to mine upstream and downstream causal relationships. An algorithm by name Upstream Downstream – Causal Relationship Mining (UD-CRM) is proposed to achieve this. The framework and underlying algorithm produce specific rules that are used to conclude causal relations. Experiments are made with stock dataset using a prototype application built. The experimental results revealed that the proposed framework is useful and performance better than existing approach.
Keywords: Data mining, causal relationships, upstream and downstream causal relationships

Scope of the Article:  Data Analytics