Role
UX Designer
Tools
Sketch, InVision
Time
1 Month
Summary
When we started work on scite_ they were past the start up phase, but not gaining the traction with their users they wanted to see. scite_'s mission was to combine a natural language processing A.I. with a network of experts to evaluate the veracity of scientific work to become a powerful time saver for those conducting scientific research.
01.Discovery
Overview
scite_ was created to aid those who conduct scientific research. It does this by using A.I. to classify papers into either supporting, mentioning and contradicting, saving researchers precious time.
The Challenge
When users of scite_ understand what scite_ can do, they immediately see the benefit and power of such a platform. However, not too many users really understood what scite_ was, and how it worked. Also, once they did use it they never came back to platform after the initial use.
Problem Statement
How might we create a trusted resource for researchers at every stage of their research journey?
02.Research
Methods
-
User Interviews
-
Usability Testing
-
Competitive Analysis
-
Comparative Analysis
-
User Journey
Insights
-
User's wanted access to a breadth of articles.
-
A search engine that allowed for custom search results.
-
Ability to save and organize research.
-
Trust and confidence in scite_ brand
03.Design
Overview
From our research we were able to identify 3 features that we could design for that would help users during their research.
Advanced Search
To give user's more control over search results, filters such as Authors, Journals, Languages, Field of Study and Year Published were added as these were the most common and important to users.
Sorting by Relevancy
We also identified a need to sort the results by relevancy to the user. This included Most Recent, Supporting, Contradicting and Mentioning.
Unique Classifications
We also added scite_'s matrix to the results page. As this was the first place user's interacted with scite_ we wanted to showcase scite_'s unique classifications front and center.
User Library
We introduced the idea of a user library, where user's would be able to save, categorize and add custom labels to papers. Saving them the time and hassle of cross-referencing and loosing sources.
Badge
scite_ has a unique set of metrics that we wanted to showcase. From our research we realized that the badge could address two user concerns.
1. Build trust and confidence in brand
2. Give scite_ users a quick and familiar way to view unique information
04.Outcomes
Feedback
Since our work on scite_ they have had a tremendous uptick in active users. They have partnered up with various data and research companies to make their A.I. even more robust, and have expanded their features to allow for efficient and seamless research.
If I had more time on the project, I did like their approach of making the A.I. even more powerful and adding additional features. I would still play around with ideas for showcasing their unique classification system.
Prototype
