Identification of essential attributes in conclusions of student reports

Abstract

Abstract This work seeks to help students in improving their first research reports, based on natural language processing techniques. We present a Conclusion model that includes three schemes: Goal Connectedness, Judgment and Speculation. These subsystems try to account for the main expected attributes in conclusions, specifically the Connectedness with the general objective of the research, the evidence of value Judgments, and the presence of Future work as a result of the student reflection after the inquiry. The article details the schemes, a validation of the approach in an annotated corpus, and a pilot test with undergraduate students. Results of a prior validation indicate that student writings indeed adhere to such attributes, especially at graduate level. Statistical results of the pilot test showed that undergraduate students in an experimental group achieved improved conclusion content when compared with the control group.

Keywords

• Natural language processing
• Automated text evaluation
• Conclusion formulation
• Goal connectedness
• Reports assessment