Advertisement

Semantic Matching of Open Texts to Pre-scripted Answers in Dialogue-Based Learning

Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11626)

Abstract

Gamification is frequently employed in learning environments to enhance learner interactions and engagement. However, most games use pre-scripted dialogues and interactions with players, which limit their immersion and cognition. Our aim is to develop a semantic matching tool that enables users to introduce open text answers which are automatically associated with the most similar pre-scripted answer. A structured scenario written in Dutch was developed by experts for this communication experiment as a sequence of possible interactions within the environment. Semantic similarity scores computed with the SpaCy library were combined with string kernels, WordNet-based distances, and used as features in a neural network. Our experiments show that string kernels are the most predictive feature for determining the most probable pre-scripted answer, whereas neural networks obtain similar performance by combining multiple semantic similarity measures.

Keywords

Answer matching Semantic similarity Natural Language Processing Neural network 

Notes

Acknowledgments

This activity has received funding from the European Institute of Innovation and Technology (EIT). This body of the European Union receives support from the European Union’s Horizon 2020 research and innovation programme. This research was also partially supported by the 644187 EC H2020 RAGE project.

References

  1. 1.
    Jeuring, J., et al.: Communicate! — a serious game for communication skills —. In: Conole, G., Klobučar, T., Rensing, C., Konert, J., Lavoué, É. (eds.) EC-TEL 2015. LNCS, vol. 9307, pp. 513–517. Springer, Cham (2015).  https://doi.org/10.1007/978-3-319-24258-3_49CrossRefGoogle Scholar
  2. 2.
    Holtgraves, T., Han, T.L.: A procedure for studying online conversational processing using a chat bot. Behav. Res. Methods 39(1), 156–163 (2007)CrossRefGoogle Scholar
  3. 3.
    Rajpurkar, P., Zhang, J., Lopyrev, K., Liang, P.: Squad: 100,000+ questions for machine comprehension of text. arXiv preprint arXiv:1606.05250 (2016)
  4. 4.
    Richardson, M., Burges, C.J.C., Renshaw, E.: MCTest: a challenge dataset for the open-domain machine comprehension of text. In: Conference on Empirical Methods in Natural Language Processing (EMNLP 2013), Seattle, Washington, USA, pp. 193–203. ACL (2013)Google Scholar
  5. 5.
    Feng, M., Xiang, B., Glass, M.R., Wang, L., Zhou, B.: Applying deep learning to answer selection: a study and an open task. In: 2015 IEEE Workshop on Automatic Speech Recognition and Understanding (ASRU), pp. 813–820. IEEE (2015)Google Scholar
  6. 6.
    Postma, M., van Miltenburg, E., Segers, R., Schoen, A., Vossen, P.: Open Dutch WordNet. In: Global WordNet Conference, p. 300, January 2016Google Scholar
  7. 7.
    Mihalcea, R., Corley, C., Strapparava, C.: Corpus-based and knowledge-based measures of text semantic similarity. In: 21st International Conference on AAAI, Boston, Massachusetts, vol. 1, pp. 775–780. AAAI Press (2006)Google Scholar
  8. 8.
    Budanitsky, A., Hirst, G.: Evaluating WordNet-based measures of lexical semantic relatedness. Comput. Linguist. 32(1), 13–47 (2006)CrossRefGoogle Scholar
  9. 9.
    Lodhi, H., Saunders, C., Shawe-Taylor, J., Cristianini, N., Watkins, C.: Text classification using string kernels. J. Mach. Learn. Res. 2, 419–444 (2002)zbMATHGoogle Scholar
  10. 10.
    Ionescu, R.T., Popescu, M., Cahill, A.: Can characters reveal your native language? A language-independent approach to native language identification. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1363–1373 (2014)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.Computer Science DepartmentUniversity Politehnica of BucharestBucharestRomania
  2. 2.Computer Science DepartmentUtrecht UniversityUtrechtThe Netherlands

Personalised recommendations