BigDataGrapes D5.3 - Trust-aware Decision Support System

In deliverable 5.3 of work package 5, we demonstrate a trust-aware decision support that uses visualisation techniques to explain the influence of input (predictor) variables on prediction outcomes. Research has shown that prediction models currently employed in agricultural decision support systems (DSS) remain opaque to users and hidden behind the software. This blackbox nature can often lead to users not trusting the system’s decisions especially when the system fails to provide meaningful explanations. Previous work has expressed that explaining a model’s predictions is an important approach for earning users’ trust. Visualisation is a powerful technique to address this problem and can effectively communicate uncertainty emerging from both data and prediction models.

To demonstrate our decision support system, we used a wine quality dataset and linear regression, building on top of deliverable 4.3 where WP4 partners (CNR) used the same dataset to demonstrate a linear regression model. The dataset is based on red variants of the Portuguese "Vinho Verde" wine and contains 1599 instances of 11 physicochemical (inputs) variables and a sensory (the output) variable which is wine quality. A literature review was conducted to better understand visualisation techniques that have previously been developed based on this dataset. Based on the findings from this review, we selected two unique visualisations as a starting point, which are parallel coordinates and a waterfall plot. The two visualisations can clearly illustrate multidimensional data and the influence of each input variable by allowing us to arrange the variables accordingly (e.g. left to right and top to bottom respectively; see Figure 3). In this document, we present our decision support system with a substantial focus on the visualisation components and prediction model.

This document is structured as follows. Chapter 1 lays out an introduction to the deliverable describing the existing work and motivations. In Chapter 2, the visualisation components and prediction model of our DSS are described in detail together with their development framework. In Chapter 3, we provide a usage manual with instructions on how to reuse the components. This document concludes with Chapter 4 where a summary of the deliverable is underlined.

As the components are updated in the future with added visualisations and models, we will keep this document up to date accordingly. To keep track of such changes, please use this link: https://goo.gl/7zba2b.