Advancing User Experience : Achievements in ExtremeXP's WP4

In work package 4, “Visualization of and interaction with decision support based on data analytics”, the project partners SINTEF, Athena Research Center (ARC), Activeeon (AE), Bournemouth University (BU), Charles University (CUNI), DFKI, INTERACTIVE 4D (I4D), ICCS, Intracom Telecom (ICOM), MOBY X, Universitat Politècnica De Catalunya (UPC) and Vrije Universiteit Amsterdam (VUA) cooperate on providing the best possible support and user experience for the users conducting experiments. This is done by making it easy to specify experiments (Task 4.3 and 4.4), and by making it easy to understand and interact with the results of the experiments in an engaging way for users (Task 4.1, 4.2 and 4.5).

When specifying experiments, users are given the opportunity to express their goal for an experiment in a high-level manner, called a user intent. Based on this, a complex analytics workflow can be created automatically, or specified interactively. This work (T4.3) is led by UPC, with contributions from ARC, AE, and VUA. To further ease the specification of experiments, users may receive suggestions for their intents, based on characteristics of the user and their previous experiments. This work (T4.4) is also led by UPC, with contributions from ICCS and ICOM.

When an experiment is being conducted, users are offered flexible and varying visualizations of the data used in the experiment, the results of different parts of the experiment, as well as benchmarks of the performance of different variants of the experiment. This work (T4.1) is led by ARC, with contributions from CUNI, ICOM, MOBY, and SINTEF. To further enhance the users’ understanding of the results and benchmarks, the users are offered explanations, both of individual results of different parts and variants of the experiment, and of the whole experiment. Explanations are offered both using traditional interaction mechanisms and augmented reality. Furthermore, users may also provide feedback to the specification of the experiment when interacting with the visualizations and explanations. In addition to enhancing the understanding, such explainability and feedback mechanisms also increase the trust in the experiment and its results. This work (T4.2) is led by SINTEF, with contributions from ARC, BU, and DFKI.

To ensure that users are engaged when working with complex analytics workflows, users are offered gamification mechanisms, including the use of augmented and virtual reality when appropriate. This work (T4.5) is led by I4D, with contributions from CUNI, DFKI and ICOM .

The first 15 months of the ExtremeXP project WP4 has developed many impressive results. Results from T4.1 and T4.2 include a visualization middleware to ensure high performance when working with large data sets, an explainability module that provides the most appropriate explanations of the results from an experiment, and specifications and mock-ups of a visualization dashboard, i.e., a generic module for visualizing and explaining data sets, results, and benchmark. Furthermore, a prototype implementation of this dashboard for one of the use cases, as well as mock-ups of other visualizations for selected use cases have been developed.

Results from T4.3 and T4.4 include knowledge-graph-based prototype tools for capturing user intents in an interactive way. Based on the obtained user intents, complex analytics workflows are automatically generated for the workflow engines KNIME and ProActive. Furthermore, prototype implementations of intent capturing tools tailored to selected use cases have been developed, and initial work on anticipation of user intents has been conducted.

Results from T4.5 include concepts, mockups and plans for a generic gamification framework which will be instantiated to all use cases in ExtremeXP.

In addition, WP4 has also contributed to the architecture and meta models on which ExtremeXP are based.

The ExtremeXP project is co-funded by the European Union Horizon Program HORIZON-CL4-2022-DATA-01-01, under Grant Agreement No. 101093164
© ExtremeXP 2023. All Rights Reserved – Privacy Policy
Verified by MonsterInsights