VISUAL SEARCH INTERFACE
Introduction
In the beginning of 2015, the Industrial Design Intensive class developed a project together with the Umeå University Library. In order to To facilitate research and education at Umea University, by redesigning library website. This ended winning an annual digital innovation call from a company in Umea, competing with other projects from tech, business, and design students.
Result
We redesigned the Library website overall most-used feature: the search engine! We used Venn diagram approach to expand modularity and ways users might want to build their own search. It creates a more personal, interactive , easy and informative experience for students and researchers.
Details
In collaboration Victor Muñoz and Artur Becker
Duration: 3 weeks
Projects
User groups
We sought to understand how the library mission was perceived by its users, dividing these into smaller groups: students and researches, as part of the academic community; and "teenage bookworms", that are not (yet) inside the university but might search books in its large catalog. After interviews with users from each group, we created personas to represent them.
Ideation: Accounting for different needs
The research phase brought the key insight that users interact with the library website in different ways, and to appeal to them, the library website had to manage to accommodate the different possible mindsets. Three alternatives of redesigns for the library website were proposed:
1) Log-in encouraging interface
A more subtle rework of the website aims to encourage the user to log in to the library network - an action, that according to the librarians interviewed, greatly enhanced the user experience.
2) Branches website structure
An overhaul of the website focusing on making the navigation of its internal categories easier to understand, through a tree-like interface.
3) Venn diagram exploratory search
An alternative search interface for the catalog, using Venn diagrams to intersect interest areas and explore new books.
User testing
With the new concept under development, Lo-fi mockups of the interface were printed and presented to potential users, to simulate the searching experience.
Co- designe workshop
Partnering with the librarians, a workshop was arranged with primary school children seeking to find inspiration on how they would catalog and classify information (represented by different objects laid on the floor).
Visual search interface
Despite the expenses of providing access to all the library resources, library users are not aware of this depth of resources the library has to offer. Search refinement options are tucked away in a sidebar, where those returning least results are hidden. Refining the search can only be done selecting one new filter at a time. This requires loading a new page for every step or to return to a previous state.
Outcome of ideation
From these alternatives, we found that the most impactful on accomplishing the library mission would be the redesign of its overall most-used feature: the search engine. The Venn diagram approach, however, could not be used due to it's restriction to three intersecting sets, using perfect circles. Therefore, we retinkered the idea based on the tree-like navigation design of alternative
.jpg)
.jpg)
.jpg)
Reserach methods and tools
Our research process consist of different methods such as Interviews,creating persona, co design workshop with kids and user testing.
Paradigm translation
Our final concept bestows the thinking behind modular synthesizers and sound editing software: linking of a source into sequential refinement steps through inputs and outputs.
Lets the user visualize the entirety of resource categories.
Provides visual feedback for the search filtering.
It is constantly refreshing the search results in the same page.
Relies only on the existing database and boolean operations.
The Visual Search Interface
Whether by selecting filters from the word cloud or writing it and linking the balloons, the user can logically link multiple search parameters.
Search refinement doesn’t have to follow only one path either, since the tool permits switching between filter categories with a simple mouse scroll.
The color on the balloons serves as a feedback of how many results are being returned at the end nodes. The closer to the background color, the less results.
Whether by selecting filters from the word cloud or writing it and linking the balloons, the user can logically link multiple search parameters.
The concept enables the user to do searches of scalable complexity without needing to be familiar with boolean algebra, giving he or she freedom to concentrate solely in the semantics of the search.
The depth of complexity that can be achieved by combining boolean operations, as well as the possibility to bypass and focus steps of the filtering process greatly reduces navigation time in comparison to a one-page-per-step model.