Ontology Design and Use for Subjective Evaluation of Computer Science Examinations

Chapter

Abstract

Ontology is the study of groups or classes of things. It is a classification of concepts in a domain. Domain Ontology represents the concepts which belong to part of the world. Ontology along with set of individual instances of classes constitutes a knowledge base. In this paper, Domain Ontology of Computer Graphics is prepared using subject–predicate–object representation, where subject is a class or instance, predicate is property of class or instance and object is values of properties. This Ontology is then used for subjective evaluation of answers submitted by students in examinations. The statistical techniques, namely Latent Semantic Analysis (LSA), BiLingual Evaluation Understudy (BLEU), Generalized Latent Semantic Analysis (GLSA), Maximum Entropy, and Hybrid Technique based on LSA and BLEU are used for performing evaluation of student answer along with Ontology. When the students’ answers are to be evaluated, the details related to the concept are fetched from the Ontology. The point of access is provided by human examiner. When short questions are to be answered, then directly related information is fetched. When longer questions are to be answered, then more details are fetched using Ontology property, instances, and subclasses. After performing preprocessing, the students’ answers are given as input along with subject-specific Ontology to statistical techniques. The sentences in each student answer are classified as belonging to an Ontology concept using statistical techniques. Then, total number of concepts found in each answer is divided by total number of concepts in concept map to generate the similarity score between concept map and student answer. The development and implementation of Ontology-based evaluation involve: development of Ontology, extraction of Ontology from RDF file, and implementation of statistical techniques in Java Programming Language. The Ontology is implemented with the help of Protégé tool. The developed application is tested using 10 questions of Computer Graphics. The techniques are tested both with and without Ontology. The OHYBRID technique shows the best correlation as compared to all other techniques. Maximum Entropy has correlation varying from 0.61 to 0.96. There is a lot of improvement in Maximum Entropy with Ontology. GLSA Ngram 1 and 2 do not show much improvement with use of Ontology as correlation varies from 0.44 to 0.87 for NGRAM1 and 0.41 to 0.84 from NGRAM2. The performance of the Ontology-based evaluation is compared with evaluation without Ontology. It is found that with use of Ontology the performance is more streamlined as individual feedback can be given to students. The feedback includes the concepts that are missing in student answer which is more near to human-like evaluation. It is concluded that use of Ontology makes the evaluation more thorough and near to accurate.

References

  1. 1.
    Noy, N. F., & McGuinness, D. L. (2001). What is an ontology and why we need it.Google Scholar
  2. 2.
    Ozcan, R., & Aslangdogan, Y. A. (2004). Concept based information access using ontologies and latent semantic analysis. Department of Computer Science and Engineering, 8, 2004.Google Scholar
  3. 3.
    Dridi, O. (2008). Ontology-based information retrieval: Overview and new proposition. In Second International Conference on Research Challenges in Information Science, 2008. RCIS 2008, pp. 421–426.Google Scholar
  4. 4.
    Oleiwi, S. S., & Yasin, A. (2013). Classify the scientific paper to multi categories using ontology.Google Scholar
  5. 5.
    Wu, S. -H., Tsai, T. -H., & Hsu, W. -L. (2003). Text categorization using automatically acquired domain ontology. In Proceedings of the sixth international workshop on Information retrieval with Asian languages-Volume 11 (pp. 138–145).Google Scholar
  6. 6.
    Bulakh, P. (2013). Automation of subjective answer evaluation, 4(1), 1328–1332.Google Scholar
  7. 7.
    Youn, S., Arora, A., Chandrasekhar, P., Jayanty, P., Mestry, A., & Sethi, S. (2009). Survey about ontology development tools for ontology-based knowledge management (pp. 1–26). University of Southern California.Google Scholar
  8. 8.
    Mylopoulos, J., Buitelaar, P., Olejnik, D., Sintek, M., Della Valle, E., Castagna, P., & Brioschi, M. (2003). OntoLT: A protégé plug-in for ontology extraction from text (pp. 31–44).Google Scholar
  9. 9.
    D’Avanzo, E., Lieto, A., & Kuflik, T. (2008). Manually vs semiautomatic domain specific ontology building (Doctoral dissertation, Thesis in Information and Commercial Electronics).Google Scholar
  10. 10.
    Sarda, N. L. (2007). Ontology-enabled database management systems. In Ontologies (pp. 563–584). Springer, Boston, MA.Google Scholar
  11. 11.
    Zouaq, A., & Nkambou, R. (2008). Building domain ontologies from text for educational purposes. IEEE Transactions on Learning Technologies, 1(1), 49–62.CrossRefGoogle Scholar
  12. 12.
    Foltz, P. W., Kintsch, W., & Landauer, T. K. (1998). The measurement of textual coherence with latent semantic analysis. Discourse Processes, 25(2–3), 285–307.CrossRefGoogle Scholar
  13. 13.
    Pérez, D., Pérez, D., Gliozzo, A., Gliozzo, A., Strapparava, C., Strapparava, C., Alfonseca, E., Alfonseca, E., Rodriguez, P., Rodriguez, P., Magnini, B., & Magnini, B. (2005). Automatic assessment of students’ free-text answers underpinned by the combination of a BLEU-inspired algorithm and latent semantic analysis. In Proceedings of the Eighteenth International Florida Artificial Intelligence Research Society Conference, FLAIRS 2005—Recent Advances in Artificial Intelligence (pp. 358–363).Google Scholar
  14. 14.
    Islam, M. (2010). Automated essay scoring using generalized. In Proceedings of 13th International Conference on Computer and Information Technology (ICCIT 2010).Google Scholar
  15. 15.
    Sukkarieh, J. Z., & Blackmore, J. (2009). C-rater: Automatic content scoring for short constructed responses. FLAIRS Conference (pp. 290–295).Google Scholar
  16. 16.
    Devi, M. S., & Mittal, H. (2016). Machine learning techniques with ontology for subjective answer evaluation. International Journal on Natural Language Computing, 5(2), 01–11.CrossRefGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  1. 1.GGDSD CollegeChandigarhIndia

Personalised recommendations