PROJECT NAME: ResearchSpace
ROLE: UX/UI Designer
DURATION: 1 year (Sep 2021-Sep 2022)
Team Members:
- Kim Martin (Research Board Chair)
- Susan Brown (Project Manager)
- Robin Bergart (UX Team Lead)
- Jordan Lum (UX Designer)
- Nem Brunell (Content Designer)
- Evan Rees (UX/UI Designer)
The Linked Infrastructure for Networked Cultural Scholarship (LINCS) enables researchers to create interlinked and contextualized online data about culture to benefit scholars and the public.
LINCS’ s mission is to make cultural data more readily available, shareable, searchable, and reusable. As part of this infrastructure project, the UX team investigated several different platforms and their affordances. .
LINCS Tube Map: This illustration is meant to show users the flow of data through the LINCS ecosystem of tools, and the breakdown of functions for each.
RESEARCHSPACE OVERVIEW
ResearchSpace enables linked data storage, review, publication, and discovery.
With its origins in the field of museum studies and, from its conception, working with data from the British Museum, ResearchSpace allows researchers to create knowledge maps (interactive networks of entities and their relationships, as seen in the visualization directly below) and view datasets in multiple structured views to be able to see connections they otherwise may not have.
ResearchSpace: Visualization of the process and purpose of creating a knowledge map, from the British Museum (https://researchspace.org/)
Through multiple rounds of user testing and iterations, our team researched how to best utilize and adapt existing ResearchSpace UI/information architecture for Digital Humanities (DH) purposes.
RESEARCHSPACE WITHIN LINCS
ReseachSpace supports two main types of users by creating two different ResearchSpace environments to cater to both types of users:
- ResearchSpace: Provides access to published LINCS data and all the features of the tool with the exception of data creation and modification
- ResearchSpace (Review): Provides access to published and pre-published LINCS data, and lets users with a researcher account make changes to their own data and curate project metadata.
Prior to LINCS’ adoption of the interface, ResearchSpace was used by the BritishMuseum for two other main projects: Late Hokusai, focusing on the life and artwork of Katsushika Hokusai, and Amara West, focusing on archaeological investigations in the Nile Valley.
Amara West ResearchSpace
Late Hokusai ResearchSpace
https://latehokusai.researchspace.org/resource/rsp:Start
THE PROBLEM
Adapting an existing UI designed for a relatively small dataset within one topic poses many benefits but also complications.
LINCS' goal was to repurpose the tool to house and connect hundreds of datasets from scholars worldwide required assessing pain points for users and areas of the infrastructure that were static, and others that could be changed.
Presenting findings of cross-analysis between LINCS and INKE tools at the INKE Conference in January 2023
Research and Insights
RESEARCH CONTEXT
As LINCS ResearchSpace continued to evolve from its Minimum Viable Product (MVP) to pre-publication versions with increasing functions and datasets, our resources grew and we conducted extensive user research on ResearchSpace to flush out major pain points within the interface.
Through guiding and observing participants through a series of scenarios and tasks, such as creating a knowledge map based on particular entities found in datasets accessible in ResearchSpace (Figure 1), our team was able to assess the overall usability of the site and whether the functionality of ResearchSpace was apparent to users.
Figure 1: Knowledge Map
KNOWLEDGE MAP
A. Entity card depicting basic details regarding an entity (data point in a data set).
B. CIDOC-CRM connection between entities visualized as a line showing the directional correlation between them
C. Left side panel access for a list of classes and instances (occurrences and classifications of entities within a data set).
D. Access bar for quick functions (expand, clear all, auto-layout, export to PNG/SVG, zoom)
E. Connections box containing a list of CIDOC-CRM connections specific to the selected entity
F. Clipboard side panel to allow users to save entities from a variety of datasets to a clipboard to drag and drop into a custom Knowled
METHODS
Through a think-aloud process, our team collected feedback on ResearchSpace in two separate parts: the first focused on searching and browsing the datasets and saving entities for future use (Figure 2), and the second round focused on the data visualization elements such as the knowledge map (Figure 1).
Figure 2: Examples of searched entities in ResearchSpace
FINDINGS AND ANALYSIS
Each round revealed valuable insights regarding what was working best with users, and what limitations required attention. Users predominantly found the idea of the tool useful, agreeing that being able to visualise their data and how it connects to others’ data within ResearchSpace was similar to a mind map (P2, round 2).
LIMITATIONS WITH EXISTING UI
However, limitations in the coding environment for ResearchSpace made implementation of feedback difficult due to complex changes that needed to occur on the backend.
Further, having its origins in museum studies, the language used on the site that categorizes entities by their CIDOC-CRM ontological class was confusing to users who were not versed in LOD or ontologies, shown in the figure to the left.
Design and IA Changes
After multiple rounds of user testing, adjustments to iconography and layout to reduce user pain points and clarify overall functionality.
ICONOGRAPHY AND LAYOUT
Key areas for pain points for users that we found were:
- language of the top-level knowledge map bar (Figure 1, point D), like “forced layout” was seen as confusing and the iconography generally unintuitive (P1, round 2).
- the clipboard and side panels (Figure 1, points C and F) should remain open by default rather than closed were also suggested, and were easy to implement by the dev team.
Thinking Frame Nav Bar
The language of ‘force layout’ combined with the ‘snowflake’ icon was unintuitive to users, so changes were made to clarify functionality with ‘auto arrange’.
Entity Card Design
Entity cards were redesigned to have a simpler, cleaner look, and include images, with CTA’s for finding more info about the entity (explore) and to create a knowledge map of the entity (map).
CONCLUSION
Overall, while users found navigational and aesthetic elements of the ResearchSpace UI confusing, they valued the purpose of the overall project and the function of the tool, stating it would allow them to “see a constellation of entities and connections” (P2, round 2).
Through further user testing to occur this winter, focusing more on the side panels in the knowledge map section to determine better terminology and layout for users to understand, we will continue to discover the benefits that implementing an interface from a different field affords, and find ways to make it further afar.