Evolutionary Visual Exploration: Evaluation of an IEC Framework for Guided Visual Search

Abstract

We evaluate and analyse a framework for evolutionary visual exploration (EVE) that guides users in exploring large search spaces. EVE uses an interactive evolutionary algorithm to steer the exploration of multidimensional data sets toward two-dimensional projections that are interesting to the analyst. Our method smoothly combines automatically calculated metrics and user input in order to propose pertinent views to the user. In this article, we revisit this framework and a prototype application that was developed as a demonstrator, and summarise our previous study with domain experts and its main findings. We then report on results from a new user study with a clearly predefined task, which examines how users leverage the system and how the system evolves to match their needs. While we previously showed that using EVE, domain experts were able to formulate interesting hypotheses and reach new insights when exploring freely, our new findings indicate that users, guided by the interactive evolutionary algorithm, are able to converge quickly to an interesting view of their data when a clear task is specified. We provide a detailed analysis of how users interact with an evolutionary algorithm and how the system responds to their exploration strategies and evaluation patterns. Our work aims at building a bridge between the domains of visual analytics and interactive evolution. The benefits are numerous, in particular for evaluating interactive evolutionary computation (IEC) techniques based on user study methodologies.

How Domain Experts Structure Their Exploratory Data Analysis: Towards a Machine-Learned Storyline

“…Sessions were organised in two parts: (i) a training part to teach participants how to use our tool, similar to the training task we used in previous studies [4]; (ii) an ideation part where participants brainstormed about the functionalities our SPLOM-tool should have to better support sensemaking of exploration history. We used affinity diagramming and thematic analysis to organise those ideas into the following high-level user requirements, ordered by how frequently they were mentioned by our participants: 1 In terms of support for storytelling and authoring, participants mentioned creating automatically a storyboard of past exploration and annotating it, for example, by allowing users to tag places where the exploration branched out.…”

Multi-Criteria Reverse Engineering for Food: Genesis and Ongoing Advances

“…To this end, the EvoGraphDice [97] visualization tool was used to explore the Pareto front of a wine fermentation model [98]. In this case, domain experts were interested in finding fermentation control strategies to obtain a target aromatic composition, while minimizing the amount of energy required to regulate the optimum temperature.…”

Evaluation of Interactive Machine Learning Systems

“…Automatic dimension reduction techniques, such as principle component analysis and multidimensional scaling, reduce the search space, but often are difficult to understand [42], or require the specification of objective criteria to filter views before exploration. Other techniques guide the exploration towards the most interesting areas of the search space based on information learned during the exploration, which appears to be more adapted to the free nature of exploration [6,10]. In our previous work on guided exploratory visualization [6,7,37,12,13], we tried to address the problem of how to efficiently explore multidimensional datasets characterised by a large number of projections.…”

“…Other techniques guide the exploration towards the most interesting areas of the search space based on information learned during the exploration, which appears to be more adapted to the free nature of exploration [6,10]. In our previous work on guided exploratory visualization [6,7,37,12,13], we tried to address the problem of how to efficiently explore multidimensional datasets characterised by a large number of projections. We proposed a framework for Evolutionary Visual Exploration (EVE, Figure 1) that combines visual analytics with stochastic optimisation by means of an Interactive Evolutionary Algorithm (IEA).…”

“…Cortellessa and Cesta [18] found that the quantitative evaluation of mixed-initiative systems tend to focus either on problem-solving performance of the human and what they call the artificial solver, or the quality of interaction looking at user requirements and judgment of the system. This statement also applies to iML systems, where current evaluations tend to be either algorithm-centered to study the computational behaviour of the system, or human-centered focusing on the utility and effectiveness of the application for end-users [6,8,7].…”