News Article

UC Berkeley Calendar Network:
A Campuswide Event Calendar Project

by Allison Bloodworth, eBerkeley;
Jeffery Kahn, Public Affairs;
Jon Conhaim, eBerkeley

February 14 , 2005

Although UC Berkeley boasts an incredibly rich array of public events taking place on any given day — from a Management of Technology lecture at the Haas School of Business to a film series at the Pacific Film Archive, from a physics colloquium to a dance performance at Cal Performances — it can be difficult for people interested in these events to find out about them. These potential event attendees may be UC Berkeley faculty, staff, or students, or the general public.

Currently, the campus has some 70 online event calendars. The Public Affairs department attempts to aggregate and publicize as much public event information as possible in the current campuswide calendar, UC Berkeley Events, but it is not comprehensive. Imagine a potential student visiting campus for a day and wanting to find information on all the events he or she might want to attend that afternoon. If they wanted to do a really thorough search, they would need to look at more than 70 different calendars!

In another scenario, imagine a student or professor interested in biosensor research — for instance, a sensor that can detect anthrax. That person might find seminars of interest published in the calendars of the Physics Department, the School of Public Health, the College of Engineering, and in any of the life sciences academic units on campus. It is clearly difficult for a potential event attendee to get a comprehensive view of all the events happening here at UC Berkeley, or to find a particular event of interest to them.

Early in 2003, Adjunct Professor Bob Glushko of the School of Information Management & Systems (SIMS) approached the UC Berkeley Public Affairs office, which administers the campuswide calendar, suggesting that there must be a better way to allow the many online public event calendars on the UC Berkeley campus to share event information. Dr. Glushko followed up with a visit to a Webnet meeting to make his case. Many calendar owners were receptive and with this a wide-ranging collaboration, the UC Berkeley Calendar Network project, was born.

Background

Today, owners of campus calendars use many different techniques and methods to publish events on their own web-based calendars, from creating static HTML pages to developing more complex, database-driven event management systems. However, because there is currently no automated way to send event information from one calendar's system to the main campus website calendar, these calendar administrators must take the extra step of manually entering their event information into a separate online form in order to publish their events there. After entering event information on their own website, many calendar administrators don't feel compelled to take that extra step to enter it again at another website, so much event information never makes it to the gateway calendar.

Likewise, colleges and department have no easy way of passing along events of interest to other colleges. For instance, a Haas School of Business event on intellectual property law would certainly be of interest to individuals at the Boalt Hall School of Law. However, currently the calendars of these two graduate schools do not talk to one another. In fact, no calendar at UC Berkeley is capable of readily sharing information with another. On a campus that stresses interdisciplinary interaction, this is a problem that cannot be ignored.

First Steps

To remedy this Balkanization, eBerkeley, Public Affairs, and IST formed a partnership with SIMS. Initially, a team consisting of students, faculty, and staff from SIMS, IST—Workstation Support Services, eBerkeley, and the Lawrence Hall of Science began work on a standard data model for a public event. This group began by analyzing 24 different campus calendars, which were representative of the different types found on campus, and investigating existing Internet-based calendar standards, such as iCalendar (RFC 2445). The team then used Dr. Glushko's Document Engineering methodology to create a standard data model of an event, which captured key information about public events ranging from date and time, to sponsors, to speakers/performers, to admission charges.

Sara Leavitt, webmaster of the Lawrence Hall of Science, who collaborated on the development of the event model, describes her reason for participating in the project this way:

Despite what people say about the campus being decentralized, I have often found that collaborating cross-departmentally produces a better service (e.g., TeleBEARS, Online Schedule of Classes, etc). I joined the calendar project because I believed that this was a case where more could be accomplished as a group than in our separate departments. For a campuswide calendar to be a success, we needed to harmonize the different requirements early in the process. The calendar network will enable us to more widely publicize events at the Lawrence Hall of Science, as well as make it possible to subscribe to other calendars' events that might be of interest to our visitors.

Over the past year, our group has conducted interviews, user tests, and focus groups with faculty, students, and staff from some two dozen colleges, departments, and organizations on campus. This user-centered design process has allowed us to determine the collective and often unique needs of these diverse organizations for a next-generation calendar, and is integral to creating applications that are flexible enough to promote collaboration on campus. It has also allowed potential campus customers of the application to play an active role in its development and become excited about using the new system. Scot Hacker, webmaster of the Graduate School of Journalism, had created a similar application himself for his department and was asked to be a participant early in the process. Lessons learned from his experience were invaluable to the project team. He explains:

As webmaster for the Graduate School of Journalism, I spent weeks developing a departmental event calendar in PHP/MySQL. What started as a seemingly simple project quickly grew innumerable tentacles as we realized how difficult it was to stuff real-world events into a comprehensible interface. Events sometimes span multiple locations, multiple sponsors, multiple days; have different ticketing and reservation requirements for students/faculty/staff/general public; may or may not be associated with live or archived webcasts; need to be summarized or detailed in different ways on different parts of the site; require that different sorts of email notifications be sent out; need to generate always-current RSS feeds; and on and on. Although we have a workable system in place today, there are still event circumstances that are tricky to fit into our current model. And we still lack integration with other online calendars sponsored by other departments. So I know from experience what a massive undertaking this distributed campuswide events system is.

The Application

The result of the initial efforts to understand this complex domain was a prototype and design for an enterprise application that will allow campus web-based event calendars to share event information: the UC Berkeley Calendar Network (UCBCN). This system was initially presented in May 2005 as the award-winning Master's thesis project of four SIMS students: Allison Bloodworth, Myra Liu, Nadine Fiebrich, and Zhanna Shamis. The UCBCN eliminates the need for every school, college, department, or other campus organization to design and deploy their event calendar in isolation. Instead, the UCBCN allows each calendar to use a web-based calendar management tool to both create and customize a dynamic, web-based calendar and manage their events in a central event repository based on the standard data model of an event.

The central event repository will store all campus events for calendars participating in the system, but each calendar administrator will be able to use a web-based calendar management tool to decide on an event-by-event basis which of their events they would like to make public and available for publication in other campus calendars. Administrators of each calendar may then decide which of the "shared" events from other calendars they want to publish in their calendar. To avoid the daily task of sorting through the many shared events taking place across the campus, the application includes a feature wherein each calendar can subscribe — by keywords, sponsors, and other criteria — to automatically pull events of interest into their calendar queue. From this pending queue, administrators then can decide which events to include in their own calendar.

The calendar management tool will also allow calendar administrators to use a range of different methods based on their skill level to customize the look and feel of their calendar. An administrator who doesn't know HTML can readily change the color, font, header, and footer of the default calendar. Alternatively, a skilled web programmer can use CSS and XSLT to completely customize the look and feel of a calendar. An additional benefit of the UCBCN is that it will allow many organizations that previously did not have the resources to create a web-based calendar to do so easily and efficiently. Finally, because the event model has been encoded as a W3C XML Schema document, departments with specialized web development or data storage needs that cannot use the calendar management tool and central event repository to create their calendar will still be able to send events to and pull events from the repository via XML data feeds.

Conclusion

Last fall, Public Affairs submitted an eBerkeley Innovation Project proposal to secure the funds necessary to bring this project to fruition. They were awarded an Innovation grant, and Allison Bloodworth was hired to work with Jeffery Kahn of Public Affairs, and Jon Conhaim of the eBerkeley Program Office, to develop the functional and technical requirements and to complete the interface design.

In November, Ms. Bloodworth presented a paper she coauthored with Dr. Glushko called "Model-driven Application Design for a Campus Calendar Network" at the XML 2004 conference. This paper discusses the way Document Engineering and User-Centered Design methodologies were applied in the development of the UCBCN. The presentation was very well received, and Ms. Bloodworth was selected for one of two "Best Presentation" awards from more than 100 presenters. The paper and presentation also promoted the event model to a large group of potential users. Because the model is general enough to apply to any public event, we intend to make it available in various XML Schema repositories, and hope its use will become widespread, potentially developing into a standard.

Dr. Glushko is very encouraged by the success of this project so far. He explains:

The Event Calendar Network effort is a great example of the way Document Engineering and User-Centered Design methods can be followed to create innovative applications that solve real problems in information-intensive domains.

Development of the final specification for the UCBCN is currently in the works, and eBerkeley and Public Affairs hope to be able to offer this public event calendar system to all schools, departments, and campus organizations within the next year. We also plan to make this application open-source and publicly available for any organization that would like to use it. We believe this project is a fine example of the way User-Centered Design and Document Engineering methodologies can be applied to create applications that are flexible enough to promote collaboration on campus. eBerkeley plans to continue to partner with SIMS as well as other campus schools, departments, and organizations to make this collaborative approach the blueprint for future campus development projects.

Additional Information

More information on the project, including the "Model-driven Application Design for a Campus Calendar Network" paper, can be found on the UC Berkeley Calendar Network website or by contacting Allison Bloodworth.

More information on the Document Engineering methodology can be found on Dr. Glushko's Infosys 243 : Document Engineering course webpage.