MULTIFOR: A Multidisciplinary Framework for the identification of new target markets by open time-series data and deep learning methods
Fig.1 - MULTIFOR framework abstract view |
MULTIFOR framework principles, are the following:
- Robust PESTEL analysis, as a means of selecting the appropriate indicators (features) that are necessary to carry out the Business Internationalization,
- A set of predefined indicators for European countries
- Open data exploitation.
- Implementation of a LSTM RNN approach,
- Non-stationary time-series forecasting.
Fig.2 - MULTIFOR Web Service |
The MULTIFOR Web app, as demonstrated in Figure’s 3 site-map, consists of the three following pages:
Fig.3 - Site-map |
- Map Page, which is the home page, and contains an interactive map of the European countries included in the project, as illustrated in the screenshot of Figure 4.
Fig.4 - Map page |
- PESTEL Analysis page, which contains an interactive pie chart with the PESTEL analysis headers, revealing their included indicators in a list whenever hovered, as displayed in Figure 5.
Fig.5 - PESTEL-Analysis page - Charts Page, a screenshot of which is depicted in Figure 6, contains the main functionality, developed for the main interaction with the user for time-series analysis, compare and forecasting.
Fig.6 - Chart page