This project presents a web-based Visual Analytics approach which helps to better understand the Draco (https://uwdata.github.io/draco/) visualization recommender systems' knowledge base.
Draco has been developed as an automated visualization recommendation system formalizing design knowledge as logical constraints in ASP (Answer-Set Programming). With an increasing set of constraints and incorporated design knowledge, even visualization experts lose overview in Draco and struggle to retrace the automated recommendation decisions made by the system. We propose a Visual Analytics (VA) approach to visualize and analyze Draco’s constraints. Our VA approach is supposed to enable visualization experts to accomplish identified tasks regarding the knowledge base and support them in better understanding Draco's internal rules:
We extend the existing data extraction strategy of Draco with a data processing architecture capable of extracting features of interest from the knowledge base. A revised version of the ASP grammar provides the basis for this data processing strategy. The resulting incorporated and shared features of the constraints are then visualized using a hypergraph structure inside the radial-arranged constraints of the elaborated visualization. The hierarchical categories of the constraints are indicated by arcs surrounding the constraints. Our approach is supposed to enable visualization experts to interactively explore the design rules' violations based on highlighting respective constraints or recommendations.
You can view an online demo here: http://vizrec.bernhardpointner.com/
This work will be presented at VIS 2023 conference.
Please cite the following paper when using the code:
Johanna Schmidt, Bernhard Pointner, and Silvia Miksch (2023). Visual Analytics for Understanding Draco’s Knowledge Base. In Proceedings of VIS 2023. To appear in IEEE Transactions on Visualization and Computer Graphics.
Automated visualization recommendation systems should be capable of creating charts of extensible specifications. ML based approaches are therefore, limited to a small predefined number of possible specifications and chart types. Vega-Lite is such an declarative specification language for creating various types of visualizations and specification, even with basic interaction possibilites. So, the vision is to extend the formal model of Draco to cover all possible visualizations of vega-lite and rank them by criteria of expressivness and effectiveness.
Because Draco's knowledge base is based on an incomplete set of rules, design experts and researchers have to extend this model to enhance Draco's recommendation capabilities. Extending this model requires the understanding of the syntax of these constraints and the structure and content of the current implemented knowledge base. As the the learning curve for this task is quite high, there is a need of an interactive system to search in this knowledge base, filter it, interact with it and showing coherences and dependencies.
This project was generated with Angular CLI.
Please note:
- Newer Angular/Cli versions newer than 14.3.0 (as defined in the
package.jsonfile) are not supported unlessangular-resize-eventsupports it. - The parameter
"allowSyntheticDefaultImports": truehad to be added totsconfig.json(vega/vega-lite#4461, vega/vega-embed#151). - We deliver Draco as a pre-build library (
./lib/draco-core-0.0.6-prebuild.tgz) due to conflicts with newer versions of Vega-Lite. We will go back to using the npm version once the references have been updated.
NodeJS version 18 and Angular/Cli version 15.2.9 need to be installed.
- Install NodeJS v18 (https://nodejs.org/en) - we used version 18.17.1 | 64bit.
This installation was tested on Ubuntu 22.04. The installation of NodeJS version 18 was done according to the NodeJS manual.
- Install curl, if not available
sudo apt-get install curl - Prepare apt-get sources
sudo apt-get update sudo apt-get install -y ca-certificates curl gnupg sudo mkdir -p /etc/apt/keyrings curl -fsSL https://deb.nodesource.com/gpgkey/nodesource-repo.gpg.key | sudo gpg --dearmor -o /etc/apt/keyrings/nodesource.gpg NODE_MAJOR=18 echo "deb [signed-by=/etc/apt/keyrings/nodesource.gpg] https://deb.nodesource.com/node_$NODE_MAJOR.x nodistro main" | sudo tee /etc/apt/sources.list.d/nodesource.list sudo apt-get update - Install NodeJS v18
sudo apt-get install nodejs -y - Install Angular/Cli v15.2.9
sudo npm install -g @angular/[email protected]
- Install NodeJS v18 from the NodeJS website (use node-v18.17.1.pkg)
- Install Angular/Cli v15.2.9
sudo npm install -g @angular/[email protected]
The source code is available here on Github.
Install Git, if not available.
Download Git for Windows and install it. A graphical user interface like TortoiseGit is recommended.
sudo apt-get install git
- Install MacPorts if not available
- Install Git using MacPorts
sudo port install git
Clone the repository
git clone https://github.com/vrvis/vis_draco_kb
If Git is not installed or should not be used, it is possible to download the source code as a ZIP file from Github and store it in a local folder.
In the folder of the project (where the package.json file is located), run npm install for installing all required modules
npm install
This will create a folder node_modules with all libraries files and a file package-lock.json.
It is necessary to manually delete all vega dependencies in ./node_modules/draco-core/node_modules by deleting all folders that start with vega-*. Unfortunately, Draco has not been updated yet to support the newest version of Vega.
The application can be run as a local development server. This is suggested for local development purposes, but not for production purposes.
- Use Angular to start a local development server
Autocompletion can be, but sharing usage data does not have to be enabled, when asked.
ng serve - Navigate your browser to http://localhost:4200/
Run Angular build
ng build
or use npm
npm run build
The build artifacts will be stored in the ./dist directory.
If you plan to host the website under a specific domain, you will probably have to add base-href to the command. For example, in our case, we used
ng build --base-href='https://vrvis.github.io/projects/vis_draco_kb/'