A financial crime analysis methodology for financial discussion boards using information extraction techniques

Lee, Pei Shyuan (2018) A financial crime analysis methodology for financial discussion boards using information extraction techniques. Doctoral thesis (PhD), Manchester Metropolitan University.

Preview

Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (13MB) | Preview

Abstract

Financial discussion boards (FDBs) have been widely used for a variety of financial knowledge exchange activities through the posting of comments. Popular public FDBs are prone to be used as a medium for spreading misleading financial information due to having larger audience groups. Moderation of posted content heavily relies on manual tasks. Unfortunately, the daily comments volume received on popular FDBs realistically prevents human moderators or relevant authorities from proactively monitoring and moderating possibly fraudulent FDB content as it is extremely time-consuming and expensive to manually read all the content. This thesis presents a financial crime analysis methodology (which is comprised of novel forward analysis and novel backward analysis methodologies) implemented in a template-based Information Extraction (IE) prototype system, namely FDBs Miner (FDBM). The methodologies aim to detect potentially illegal Pump and Dump (P&D) activities on FDBs with the integration of per minute share prices in the detection process. This integration can reduce false positives during the detection as it categorises the potentially illegal comments into different risk levels for investigation purposes. P&D is a well-known financial crime that happens through different methods including FDBs. P&D happens when fraudsters deceive investors into buying stocks by spreading misleading information. FDBM extracts a company’s ticker symbol (i.e. a unique symbol that represents and identifies each listed company on the stock market), comments and share prices from FDBs based in the UK for experimental purposes. Results from both forward and backward analysis experiments show that the two novel methodologies can aid relevant authorities in the detection of potentially illegal activities on FDBs. Semantic Textual Similarity (STS) experiments have also shown that the approach could be adopted in the process of detecting potentially illegal activities on FDBs.