Turnitin Releases Scoring Engine for Texts
Turnitin has released the
Turnitin Scoring Engine, a service designed to
provide automated scoring of essays and short answer texts.
The new tool is designed to provide instant scoring of written assignments for students in online environments and to assess writing at
the district, state or institutional level.
Based on natural language processing research launched at Carnegie Mellon University and developed by LightSide Labs, a recent acquisition
of Turnitin, the scoring engine "analyzes the lexical, syntactic, and stylistic features of writing, such as word choice and genre
conventions, unlike other automated essay scoring programs that rely on simple metrics like word count," according to a news release. "Turnitin
Scoring Engine discovers the unique, content-based patterns in an existing set of hand-scored essays and learns to replicate that judgment as
an accurate, reliable, on-demand assessment of student writing."
The Turnitin Scoring Engine begins with existing rubrics and anonymized essays with existing scores in an effort to teach the system to
value the same things an institution's instructors do. The system also tests itself automatically against instructor scores to double-check
itself before full integration.
"The current process for mass assessment of student writing is appalling," said Elijah Mayfield, vice president of new technologies at
Turnitin, in a prepared statement. "Turnitin Scoring Engine changes the way we think about assessing student writing, while maintaining the
high levels of accuracy and reliability that institutions demand. We want institutions to spend less time and energy on testing and placement.
This will encourage schools to let instructors focus more on helping students learn to write and focus less on preparing for tests and grading
test essays."
About the Author
Joshua Bolkan is contributing editor for Campus Technology, THE Journal and STEAM Universe. He can be reached at [email protected].