The necessity of content recognition in digital documents is everincreasing in the financial sector. Extracted data is used for fundamental analysis, modelling and portfolio selection. In the most prominent markets, there is a wide array of available sources to obtain the data, such as SEC filings easily. However, it is not so in markets with less investor interest, such as the CEE region or Latin America. Often, the only sources containing the data are primary reports by the company itself. Scarce secondary sources may provide data of dubious reliability. This leads to an excessive workload for analysts, implying the necessity to adapt existing intelligent methods for processing financial data.
This work is licensed under a Creative Commons Attribution 4.0 International License.