Automated Document Grading using Principal Component Analysis

  • Srujana Inturi, Madhuri Vennu, Rachana Kavukuntla


This study researches the general adequacy of the utilization of n-grams separated terms, the conglomeration of such expressions, and a blend of capacity extraction systems in building a mechanized exposition kind reviewing (AETG) contraption. The paper fixated on the difference in the primary angle investigation (PCA) through incorporating n-grams states as go into the PCA set of rules. Printed versions of inspectors' stamping plans and softcopies of understudies' responses for 2 subjects, data mining  and Internet Of Things, provided on the branch of computer science and engineering from CBIT, in 2019 II semester have been utilize like casing research. The textual contented about marking methods have been transcript interested in virtual documents the use of identical report format as the student response. The files had been preprocessed intended for stop words removal and every key-phrase stemmed to cope with morphological differences. N-gram phrases (N=2, 3) have been extract for the duration of all students’ solution scripts. The files had been represented within the vector location version as a report time period Matrix. Fundamental factor evaluation (PCA) set of rules is customized via incorporating n-gram terms as contribution to present day PCA to infer changed head perspective examination (MPCA) calculation. The MPCA changed into used to decrease the meager condition of the lattice. Record likeness develop as estimated the utilization of cosine comparability recognition which in correlation each understudy's answer content report vector with the stamping plan report vector. The MPCA based AETG machine beat the PCA proportionate having a high colossal relationship and lessening suggest total slip-ups while the human marker appraisals are contrasted with the ones of the framework. We expect to investigate different processes on the way to capable of imprison non-textual substance within our further planning.

How to Cite
Rachana Kavukuntla, S. I. M. V. (2020). Automated Document Grading using Principal Component Analysis. International Journal of Advanced Science and Technology, 35 - 42. Retrieved from