Design Journey
Define
Problem Statement
This 8-week project evaluated the metadata and taxonomy of Emily Carr University’s master’s thesis e‑collection, originally built on Islandora 7 using MODS (Metadata Object Description Schema). The existing hierarchical schema could not adequately represent the diverse, interdisciplinary nature of creative research. As a result, metadata was often incomplete, taxonomy inconsistent, and discoverability poor for users.
Goal
With the university’s planned migration to Islandora 8 and Drupal, the project aimed to: a) Identify gaps in the existing Information Architecture, b) Design a user‑centered metadata and taxonomy system, and c) Propose an implementation roadmap for Islandora 8.
User Persona
Empathy Mapping
Comparative Analysis
The comparative analysis examined the structure of master’s thesis metadata, across Canadian universities. This included how fields such as title, author, subject, genre, degree, and usage rights were applied and displayed. I also evaluated how each institution implemented faceted search filters, navigation menus, and user interfaces to support discovery.
Exisiting Information Architecture
I conducted a detailed review of the metadata and taxonomy used for both the Global navigation and Main content. This involved analyzing how thesis records were described, what metadata fields were applied, and how facets such as date, topic, genre, area, and document type were structured. The goal was to understand how well these elements supported discoverability and whether the existing taxonomy aligned with users’ mental models and research behaviours.
Problem Analysis
For this project, I conducted user research using a blended spaces approach, which explores how digital and physical environments intersect in real research behaviour.
Instead of relying only on questionnaires, I observed students and researchers in a physical library, watching how they browsed for theses, navigated shelves, cross-referenced catalogue records. I paid close attention to how they walked through the taxonomy and the vocabulary they used when making selections.
1) Topic is the primary entry point.
People think in terms of topics rather than time when browsing.
2) Users often tried to find theses by Advisore.
Students are often suggested to look for a thesis based on certain Advisors, hence name of the faculty members is an iiiiiimportant entry point that the taxonomy does not currently support.
3) Confusion between "Area" and "Topic".
Area is often recognized as Program and Topic is understood as Keywords.
4) Vocabulary gaps limited discovery.
The taxonomy doesn’t reflect user language such as Topic is often recognized as keywords.
5) Browsing behaviour was playful and exploratory.
Some users touched multiple terms in a row, discussed them with a friend, then walked to different shelves before rrrreturning to pick another topic.
6) Users skim the main content section and miss key details.
The main content section is not optimized for scanning. The long description makes it harder for users to quickly evaluate RRrelevance, and missing metadata (e.g., supervisor) limits decision-making.
7) Date is usually selected as a range.
Users expect to select a date range, but the current design only supports choosing single years.
Suggested Changes
New Information Architecture
High Fidelity Mock
