London Business School Knowledge Capture

Fizzing with ideas for knowledge capture and sharing

A web application to help compile, publish and annotate academic reading lists and other collections of links.

From a central interface, administrators can set up course or client-specific reading lists and collections of links. Users then annotate, comment on and tag the links to build their own curated collections.

The challenge

We were approached by London Business School to think about how their academic staff presented course reading lists, and how students could best consume and share their new-found knowledge.

What followed was a research and development prototyping exercise that explored the various different ways that a corpus of links can be curated, imported, shared, annotated and classified. A rapid development-feedback cycle meant we could explore a good number of ideas despite constraints on time and budget.

The solution

Using well-established frameworks for backend and front-end development, we were able to explore a number of functional ideas in a short space of time. The abstraction from the database layers that Django provides meant we could quickly make changes to the data model withou extensive re-coding. The Bootstrap framework gave us a palette of interaction components that meant we could assemble an interface with the minimum of fuss.

We considered a number of different ideas to assist administrators and faculty members in the setting up, curation, importing, grouping and addition of links to the system:

Curated reading lists can be added by faculty members, with the links annotated and ordered.
Reading lists can be exported in RSS or HTMl format, with historical snapshots embedded in external systems.
A satellite site is a branded collection of links from the corpus, including reading lists. These can be set up quickly for a targeted course or client.
A browser toolbar button allows faculty members to easily add links to any of their reading lists or satellite sites.
Links can be imported from spreadsheets or CSV files that are uploaded or available remotely. A remote spreadsheet will be regularly checked for additional links, which can then be auotmatically imported into the system.
All links added to the system can be scanned for certain criteria (e.g. sources, tags, authors, etc.) and automatically added to any reading list or satellite site.
Links can be added using Twitter

From the student/user perspective, we explored a number of different features to try to help users learn from, embellish and recall knowledge:

All links can be tagged by all users - allowing personal collections to be easily built
Fully-threaded comments for every link means students can discuss items on the reading list
A system of micro-interactions, or "bumps", allows users to quickyl favourite, recommend or warn about a particular link.
Where a link's body text is available, users can highligh and annotate particular passages
We have recently been using the D3 data visualisation library to build visual models of the data and assist user interaction

Technologies used

Python, Django, MySQL, Bootstrap, HTML, CSS, Javascript, D3 data visualisation, Amazon Elastic Beanstalk, Twitter API.