Week 10: Cleaning the UI

After a meeting with our faculty advisor, we settled on beginning the process of a user study. But, before we can progress with that we have to focus on cleaning up the UI to make sure it is user friendly and consistent, adding more datasets, and establishing a process for our user study.

For the frontend, we compiled a list of things we needed to accomplish before having a testable website. And so we began implementing and crossing tasks off of the list.

[ ] Switching out json printing view with table designed from scratch.
[ ] When removing object from the “black box” selection div. Make sure to remove that object in the “black box” once placed in rank div to represent that the user changed that object.
[ ]  Fix Pairwise view for ranking. Remove “Enter object here” on  rank div. Add icons for ui.
[ ] Shuffle button
[ ] Search through objects in build
[ ] Submit button in the right of build view and search bar in the middle of the build view.
[ ] Black box border for div containing all objects unranked should have margins on side
[ ] Bigger and centered text for data objects in build view
[ ] Footer should be slightly larger and consistent on all pages.
[ ] Data Objects in build view should have different font to stand out from other fonts on the page. Can add any labels/color to make it easier to distinguish on build view
[ ] Categories need names for all build views
[ ] Building Dataset view UI using bootstrap

As for the user study, we focused on goals. We wanted to answer the questions:
- Is this ranking tool useful?
- Among the presented three methods to rank (pairwise, list comparison, and categorical), which one is the most preferred by users?
- How intuitive is the interface of this web application?
- Who needs this ranking tool the most?

Picking out new datasets was a difficult task. Because the ranking algorithm expects a clean dataset and does not accept non-numerical attributes, a lot of the free data found online was difficult to clean up and mold in to work with our script. Also, since some parts of the code was refactored, I updated our paper for the overall project to match up with what we have.

Comments

Popular posts from this blog

Week 32: Datasets

Week 13: Receiving Feedback

Week 31: Fair Ranking Search