Monday, November 16, 2009

My Public Comment on Race to the Top -- Technology & Innovation in Assessment

My name is Tom Hoffman, from Providence, Rhode Island. I am a technology consultant, specializing in student information and assessment systems. I am project manager of SchoolTool, an open source administrative platform for schools. I also work with the CanDo project, which is an open source competency tracking application used by Career and Technical Centers in Virginia. I am a former English teacher in the Providence Public Schools with a Masters in Teaching English from Brown University.

I would like to recommend some specific facets of the technology platform for assessment, particularly in reference to Race to the Top Criteria:

B.(C)(2) Accessing and using State data: decision-makers in the continuous improvement of efforts in such areas as policy, instruction, operations, etc...

B.(C)(3)(iii) Making the data from instructional improvement systems, together with statewide longitudinal data system data, available and accessible to researchers...

These requirements suggest a high degree of data portability, interoperability, and integration, with aspirations for complex data warehousing, business intelligence and inferencing expert systems.

One of the technical foundations of this type of platform is the development of ontologies, defined as "a formal representation of a set of concepts within a domain and the relationships between those concepts." (1) Dr. Baker introduced this concept earlier.

The potential role of ontologies in educational research and throughout the implementation of educational technologies and data systems parallels to their growing role in biomedical research.

I would specifically propose funding the creation of a National Center for Educational Ontology, modelled on the National Center for Biomedical Ontology, which is funded by the National Institutes of Health (NIH).

"The goal of the Center is to support biomedical researchers in their knowledge-intensive work, by providing online tools and a Web portal enabling them to access, review, and integrate disparate ontological resources in all aspects of biomedical investigation and clinical practice." (2)

The Center is funded by the NIH Roadmap for Biomedical Research's Bioinformatics and Computational Biology initiative. The Roadmap "was launched in September, 2004, to address roadblocks to research and to transform the way biomedical research is conducted by overcoming specific hurdles or filling defined knowledge gaps... These are programs that might not otherwise be supported by the NIH ICs because of their scope or because they are inherently risky." (3)

With a consistent, ongoing commitment to the development and use of ontologies, the National Institute of Health's Recovery Act fund is already supporting 61 current research projects using or contributing to biomedical ontologies. (4)

By comparison in education, despite contributions from a disparate set of actors including the National Center for Research on Evaluation, Standards, & Student Testing (CRESST) at UCLA and Jes and Co., a 501c3 education research organization, there is no central hub for research, development and use of ontologies, individual projects tend to emerge and disappear, and in particular there is no commitment to the kind of open and collaborative environment that now typifies biomedical ontology.

For example, the National Forum on Educational Statistics at the Department of Education has created a National Educational Data Model. It is similar to an ontology, but the data model is more constrained and potentially much less rich and powerful than an ontological approach. However, it would be an obvious foundation for development of a subsequent set of educational ontologies.

CRESST has developed several detailed domain ontologies for specific subjects such as Algebra as part of their research, however, unlike their peers in biomedical research, publishing, collaborating and promoting those ontologies does not seem to be a priority, which limits their influence and impact.

Similarly, I can see from their presentations that CRESST have developed a tool called CRESST Knowledge Mapper that looks quite useful, but does not seem to be publicly available, either commercially or for free, and thus does not contribute to or promote further development of domain ontologies in education. In contrast, the National Center for Biomedical Ontology's Protege editor is an active and prosperous open source software project that has become an industry standard application.

As was the case in the biomedical field, an investment in educational ontology is relatively high risk and does not fit obviously into existing programs. If we don't start the process while we have this unique stimulus windfall, I don't know when we will.

Be assured, however, that this is essential foundational research. Given the vast ambition for educational data systems, ontologies will become as integral to educational research as they have become in the biomedical field, and sooner or later the value of our solutions will be bottlenecked by the quality of our ontologies.





No comments: