Validating an Incremental Rule Management Approach for Financial Market Data Pre-processing

Driven by the growth and availability of vast amounts of financial market data, financial studies are becoming increasingly of interest to finance researchers. However, financial market data is normally huge in amount and with data quality issues especially time-related ones, which renders it extremely difficult to generate reliable results and get interesting insights. Data pre-processing is hence necessary to control data quality and have raw data standardised, which is often achieved by using bespoke or commercial tools. In this paper, we first define ACTER criteria (automatability, customisability, time-handleability, evolvability and repeatability) to assess a financial market data pre-processing system. Then we update our previously proposed system (EP-RDR), which uses an incremental rule management approach for building and conducting user-driven event data analysis, with some new features to make it more suitable for financial market data pre-processing. Finally, we apply the ACTER criteria on an EP-RDR prototype as well as some other existing tools in the context of two real-life financial study scenarios to compare the desirability of these tools for financial market data pre-processing.