![](bga.png)
6 CONCLUDING REMARKS
We proposed a method for automatically evaluating
statements and a feedback system for the purpose of
improving the discussion skills of participants at
meetings. For automatically evaluating statements,
we set five evaluation indicators based on acoustic
features: voice size, voice intonation, speech speed,
fluency, and tempo. We also set three evaluation
indicators based on linguistic features: conciseness,
relevance to topic, and consistency of context.
We also argued that participants’ heart rate (HR)
data should be taken advantage of to effectively
evaluate the answer-quality of Q&A segments in
discussions. We developed a system for acquiring
heart rates on the basis of a discussion mining (DM)
system with the help of a non-invasive device, i.e.,
Apple Watch, worn by participants. To verify our
argument, we generated 3 binary classification
models for evaluation, logistic regression, support
vector machine, and random forest, and selected the
7 most meaningful features out of all 18 HR and HR
variability features.
Next, we analyzed the result of automatically
evaluating discussion skills and proposed a
mechanism for generating review and advice text
using sentences and graphs on the basis of the values
of indicators of discussion ability. In the analysis on
automatic evaluation, temporal situation, relevance to
other evaluations, and comparison with past results
were considered. Also, to encourage participants to
improve their discussion skills, sentences in
review/advice text were categorized into three types:
factual sentences, advice sentences, and
encouragement sentences. We also collected the
sentence elements of these sentences, and the
review/advice generation mechanism set weights to
them in consideration of the relationships between the
evaluation indicators and the sentence elements and
the past presentation situation. The mechanism
generates sentences so as to maximize the weight of
the elements. The generated sentences and graphs are
optimum for improving discussion skills. We
confirmed that the review/advice text can express the
evaluation results appropriately and is effective for
improving the discussion skills of participants.
Future tasks include long-term participant-based
experiments on evaluating discussion skills and on
training and on extending the training process to
motivate students to continue training on the basis of
gamification techniques (Ohira et al. 2014).
REFERENCES
Nagao, K., Kaji, K., Yamamoto, D., Tomobe, H., 2004.
Discussion Mining: Annotation-Based Knowledge
Discovery from Real World Activities, In Proceedings
of the Fifth Pacific-Rim Conference on Multimedia, pp.
522–531.
Camm, A. J., Malik, M., Bigger, J., Breithardt, G., Cerutti,
S., Cohen, R., Coumel, P., Fallen, E., Kennedy, H.,
Kleiger, R., 1996. Heart Rate Variability: Standards of
Measurement, Physiological Interpretation and Clinical
Use, European Heart Journal, Vol. 17, No. 3, pp. 354–
381.
Luque-Casado, A., Zabala, M., Morales, E., Mateo-March,
M., Sanabria, D., 2013. Cognitive Performance and
Heart Rate Variability: The Influence of Fitness Level,
PLoS ONE,
https://doi.org/10.1371/journal.pone.0056935.
Pereira, T., Almeida, P. R., Cunha, J. P., Aguiar, A., 2017.
Heart Rate Variability Metrics for Fine-Grained Stress
Level Assessment, Computer Methods and Programs
in Biomedicine, Vol. 148, pp. 71–80.
Peng S., Nagao, K., 2018. Automatic Evaluation of
Presenters’ Discussion Performance Based on their
Heart Rate, In Proceedings of the 10th International
Conference on Computer Supported Education
(CSEDU 2018).
Kurihara, K., Goto, M., Ogata, J., Matsusaka, Y., Igarashi,
T., 2007. Presentation Sensei: A Presentation Training
System using Speech and Image Processing, In
Proceedings of the 9th International Conference on
Multimodal Interfaces, pp. 358–365.
Barzilay R., Lapata, M., 2008. Modeling Local Coherence:
An Entity-based Approach, Computational Linguistics,
Vol. 34, No. 1, pp. 1–34.
Iakovakis, D., Hadjileontiadis, L., 2016. Standing
Hypotension Prediction Based on Smartwatch Heart
Rate Variability Data: A Novel Approach, In
Proceedings of the 18th International Conference on
Human-Computer Interaction with Mobile Devices and
Services, pp. 1109–1112.
Wang, X., Ding, X., Su, S., Li, Z., Riese, H., Thayer, J. F.,
Treiber, F., Snieder, H., 2009. Genetic Influences on
Heart Rate Variability at Rest and During Stress,
Psychophysiology, Vol. 46, No. 3, pp. 458–465, 2009.
Nagao, K., Morita, N., Ohira, S., 2017. Evidence-Based
Education: Case Study of Educational Data Acquisition
and Reuse, Journal on Systemics, Cybernetics and
Informatics: JSCI, Vol. 15, No. 7, pp. 77–84, ISSN:
1690-4524 (Online).
Ohira, S., Kawanishi, K., Nagao, K., 2014. Assessing
Motivation and Capacity to Argue in a Gamified
Seminar Setting, In Proceedings of the Second
International Conference on Technological Ecosystems
for Enhancing Multiculturality. TEEM ’14, pp. 197–
204.
KDIR 2019 - 11th International Conference on Knowledge Discovery and Information Retrieval
418