Before entering the MLIS program at the Information School, I worked for the University of Washington Program for Educational Transformation Through Technology (PETTT) as a web applications developer and educational technology researcher. When I went to graduate school, my job with PETTT became a graduate assistantship. During my first year at the iSchool, I assisted PETTT in the development of a non-functional demo for the Washington State Bar Association (WSBA) Council on Public Legal Education (CPLE) Legal Gateway Project, now known as LawForWa.org. The goal of the Legal Gateway Project was to create a public portal to the myriad web sites containing information for the general public on legal issues in Washington State. PETTT and several people at the iSchool were involved in this project due to its relevance to educational technology and content management.

After the demo, I saw an opportunity in the Legal Gateway Project to apply some of the knowledge gained in my MLIS coursework in a real-life setting. During a project meeting, I suggested hiring interns from the Information School to assist in the collection of resources and creation of a taxonomy for the web site. They asked me if I would be willing to work as an intern on the methods and technology they would use to collect and categorize the resources. I consented, and I immediately started the process for setting up a Directed Fieldwork through Lynnea Erickson, the Information School's fieldwork coordinator.

In the summer of 2002, I went to work for the WSBA. My mission was twofold:

• Guide the creation of a taxonomy that would provide structure for the CPLE's Legal Gateway web site
• Create a database for storing records classified within that taxonomy

Process

I worked with a group of other interns from the Information School, UW Law School, and Seattle University Law School who were collecting resources and suggesting terms for the taxonomy. Marc Lampson, a PhD student at the Information School with a background in law, was the supervisor for the interns. The other iSchool interns, since they had some experience with taxonomies, teamed up with the law students to help them choose terms. My role was to draft the structure of the taxonomy, create a database to hold the terms and records, and load the collected resources into the database.

The Taxonomy

My coursework in the MLIS program had greatly influenced my view of taxonomic structures. I believed that faceted classification -- the multi-dimensional, poly-hierarchical categorization of information, was the best method (see the section on Classification and Information Architecture for my argument). After taking classes on Organization of Information and Resources (LIS 530) and Classification Theory (LIS 535), I felt confident that I understood the concepts involved in constructing a faceted classification scheme. However, conveying this understanding to a group of law students and to the Legal Gateway Project team proved to be quite challenging.

I consulted with Judge Marlin Appelwick (CPLE Chair), Pam Inglesby (WSBA Education Manager) and Marc Lampson on the facets we should use for the taxonomy. We came up with four a priori dimensions: Purpose, Type, Topic, and Question. Isolates in each facet were to be arranged hierarchically. For instance, "Employment -- Discrimination -- Age" was an isolate for the Topic facet. I then created a blank MS Excel spreadsheet for collecting resources.

In one of the first meetings with the interns, I explained the concept of faceted classification and introduced the spreadsheet. I laid out guidelines for selecting terms and notation for creating categories. My presentation was met with nods from the iSchool interns and perplexed expressions from the Law School interns. It was then I realized that faceted classification might be too complicated a concept to convey to non-information professionals. So, taking a lesson from the Photoshop Workshops, I suggested to Marc that we team iSchool interns with Law School interns and assign each team of 2 to a different area of the law.

He did so, and the interns went about collecting and categorizing resources in their spreadsheets. We continued to meet as a group and individually with Marc Lampson once a week to discuss our progress. By the end of Summer quarter, we had a fledgling (though imperfect) taxonomy and a modest collection of over 1000 resources.

The Database

Meanwhile, I worked with Eric Maddox, PETTT web and database developer, to create an Entity Relationship Diagram (ERD) and logical schema for the database that would store the records. From this, we decided on a system architecture that would best accommodate the data design.

Eric began building the data storage mechanism while I created an interface for uploading, editing and viewing the documents. We connected the database and the interface together and began testing the basic functionality.

Bringing it together

Before loading the records collected by the interns into the database, I had to clean up the spreadsheets. I combined the separate files into one spreadsheet and started the arduous process of normalizing the data. After a week of this, I began to see problems with the approach we had taken. Very few of the interns had truly understood the purpose of the classification process, mostly due to a lack of collaboration with their teams. The terms some chose were full of legal jargon, while others used entry terms to describe categories. Different people had different conventions for representing the terms, and few had followed the guidelines I had given them at the beginning of the quarter. In short, it was a mess.

I did my best to clean up the data and load it into the database, but the technology was having problems as well. Theoretically, everything should have worked, but we were stymied by the idiosyncrasies of the platform & approach we had chosen. In the end, we opted for a simpler, more scalable system.

Reflection

As with many projects involving technology, we ran into some frustrating obstacles, such as scalability and performance problems. I also encountered firsthand some of the difficulties of interdisciplinary teamwork. Our team suffered from a lack of collaboration and clear communication, possibly due to the differences in language between the law students and the librarianship students.

In spite of these challenges, the project was finished to the satisfaction of the stakeholders, and we came away with a greater appreciation for the intricacies of taxonomy construction.

What's Next

I have continued to work with Eric and other PETTT team members to guide the development of the Legal Gateway web site, now called LawForWa.org. They are nearing a public release of the site, but much work remains behind the scenes on the content management system, scalability and speed of the technology, and, of course, the taxonomy. The Topic facet was too vague and became too bloated with terms, so I have suggested breaking it out into three sub-facets: Problem, Context, and Strategy. The Question facet was also loaded with a great deal of duplicates and vagaries, and I have recommended arranging it hierarchically as well, according to the need addressed in each question. I will be working with the LawForWa team to refine their taxonomy and test its utility with the potential users of the site.

Artifacts

• Link to LawForWa site
• Fieldwork learning objectives
• Fieldwork progress report
• Fieldwork final report