The Problem With Web Calendars On Campus
Due to the decentralized nature of computing on the UC Berkeley campus, different schools, departments, and campus organizations often create applications on an ad-hoc basis. The lack of campus-wide guidelines and standards for designing and building applications make it difficult for developers to design for interoperability and reuse. Consequently, the Berkeley campus is inundated with applications serving a similar purpose and repurposing similar content but built with different technologies and based on different, and often incompatible, data models. Although this is the case for many types of applications on campus, our group has decided to focus our master's project on the calendaring application.
As you can imagine there are many web-based calendars on the Berkeley campus containing events of interest to the entire campus community and events intended for a particular audience. Many of these events are cross-listed in different calendars. Usually the process of cross-listing an event requires manually copying the event information from one calendar and pasting it into the next. This is problematic in many ways. First the process of re-entering event information wastes time and is inherently error prone. Second, replicating event information increases data storage costs, can compromise the general integrity and timeliness of event information, and can increase overall complexity. Finally, incompatible data models limit the amount and type of event information that can be repurposed. These issues hinder the creation and consumption of web-based event information on the Berkeley campus.
What We're Doing About It
Our master's project has several goals. In the fall of 2003 members of our project team undertook a collaborative effort with colleagues and advisors from e-Berkeley, The Center for Document Engineering, The Berkeley Lawrence Hall of Science, and CalAgenda, the online calendar service for faculty and staff at UC Berkeley, to develop a standard data model flexible and scalable enough to accommodate the requirements of most calendars on the Berkeley campus. In the spring of 2004 the results of our modeling effort will be encoded in xml-based event schemas used to validate instances containing event information. We will further refine our event schemas through a comprehensive needs assessment in which we will interview calendar owners and users across the Berkeley campus. This phase will also serve as the starting point for identifying the potential features and functionality of a calendar management tool for calendar authors and the potential features of a calendar interface that could easily integrate with existing campus websites. An essential component of our project is building a scaled down prototype of a central repository that would house all campus-wide events, ensuring interoperability and reducing redundancy.
By the end of this semester we plan to have a functional prototype demonstrating the process of submitting events to a central event repository and subscribing to xml-based event feeds from the same repository. Our prototype will also demonstrate the process of transforming the event feed information into a visually compelling calendar displayable within an existing campus website.