The National Science Foundation’s National Science Digital Library (NSDL) awarded Howard Turtle, professor at the Syracuse University School of Information Studies (iSchool) $97,151 to update and improve the computer-assisted translation (CAT) tools used to tag uploaded educational lesson plans with federal and state education standards. The project is subcontract from the National Science Foundation grant to Cornell University, the current host of the NSDL web site.
“CAT is a tool we developed several years ago that uses natural language processing to evaluate standard documents and lesson plans,” said Turtle, the director of the Center for Natural Language Processing (CNLP) at the iSchool. “It’s getting old and needed to be upgraded.”
The goal of the NSDL site is to have a comprehensive digital library of science, technology, engineering and mathematic resources for teachers and educational institutions to share with each other. The CAT tool Turtle is improving automatically scans lesson plans developed and uploaded by teachers, librarians and institutions to the NSDL site and then suggests the educational standards that best describe the lesson plan. The user can then pick the best standard that fits the document so other users can easily find plans that fit the education standard they need.
For example, a teacher in New York State could upload a math lesson plan used in a third grade class. The CAT tool would scan the lesson plan and suggest the state or federal standards that apply to the plan, with allowances that the standards differ from each other from federal to state and state to state. The teacher would choose the best of the suggestions. If none fit, a standard could be manually entered.
“The preferred method is to use a human cataloguer who chooses the best standard for the document,” Turtle said. “Human selection of the right standard also provides feedback to the system.”
Turtle is working with CNLP programmers Steve Rowe and Ramin Gharakhanzadeh and iSchool students Wenbo Zhang G’11 and Yatish Hedge G’11 to incorporate federal and state educational standards into the CAT tool. The team also seeks to enhance the display for presenting and maintaining the system, improve the underlying algorithm, speed up the tool, and mount the service on a CNLP server from where it was previously hosted on Cornell University’s servers.
Howard Turtle is a nationally known scholar in search engine technologies, having developed a formal retrieval model based on Bayesian Inference Networks that formed the basis of the University of Massachusetts’ INQUERY Retrieval System and of West Publishing’s natural language search product. He was the chief scientist at both West Publishing and Online Computer Library Center. He holds a Ph.D. in Computer Science from the University of Massachusetts, Amherst, and an M.S. in Computer Science and B.A. in English and Mathematics from the University of Wisconsin.