Collaboration detection that preserves privacy of students’ speech

Sree Aurovindh Viswanathan, Kurt VanLehn

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Collaboration is a 21st Century skill as well as an effective method for learning, so detection of collaboration is important for both assessment and instruction. Speech-based collaboration detection can be quite accurate but collecting the speech of students in classrooms can raise privacy issues. An alternative is to send only whether or not the student is speaking. That is, the speech signal is processed at the microphone by a voice activity detector before being transmitted to the collaboration detector. Because the transmitted signal is binary (1 = speaking, 0 = silence), this method mitigates privacy issues. However, it may harm the accuracy of collaboration detection. To find out how much harm is done, this study compared the relative effectiveness of collaboration detectors based either on the binary signal or high-quality audio. Pairs of students were asked to work together on solving complex math problems. Three qualitative levels of interactivity was distinguished: Interaction, Cooperation and Other. Human coders used richer data (several audio and video streams) to choose the code for each episode. Machine learning was used to induce a detector to assign a code for every episode based on the features. The binary-based collaboration detectors delivered only slightly less accuracy than collaboration detectors based on the high quality audio signal.

Original languageEnglish (US)
Title of host publicationArtificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings
EditorsSeiji Isotani, Peter Hastings, Amy Ogan, Bruce McLaren, Eva Millán, Rose Luckin
PublisherSpringer Verlag
Pages507-517
Number of pages11
ISBN (Print)9783030232030
DOIs
StatePublished - Jan 1 2019
Event20th International Conference on Artificial Intelligence in Education, AIED 2019 - Chicago, United States
Duration: Jun 25 2019Jun 29 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11625 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference20th International Conference on Artificial Intelligence in Education, AIED 2019
CountryUnited States
CityChicago
Period6/25/196/29/19

Fingerprint

Privacy
Students
Detectors
Detector
Binary
Microphones
Interactivity
Learning systems
Speech
Collaboration
Speech Signal
Assign
Machine Learning
Choose
Alternatives
Interaction

Keywords

  • Collaborative learning
  • Learning analytics
  • Machine learning

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Viswanathan, S. A., & VanLehn, K. (2019). Collaboration detection that preserves privacy of students’ speech. In S. Isotani, P. Hastings, A. Ogan, B. McLaren, E. Millán, & R. Luckin (Eds.), Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings (pp. 507-517). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11625 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-030-23204-7_42

Collaboration detection that preserves privacy of students’ speech. / Viswanathan, Sree Aurovindh; VanLehn, Kurt.

Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings. ed. / Seiji Isotani; Peter Hastings; Amy Ogan; Bruce McLaren; Eva Millán; Rose Luckin. Springer Verlag, 2019. p. 507-517 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11625 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Viswanathan, SA & VanLehn, K 2019, Collaboration detection that preserves privacy of students’ speech. in S Isotani, P Hastings, A Ogan, B McLaren, E Millán & R Luckin (eds), Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11625 LNAI, Springer Verlag, pp. 507-517, 20th International Conference on Artificial Intelligence in Education, AIED 2019, Chicago, United States, 6/25/19. https://doi.org/10.1007/978-3-030-23204-7_42
Viswanathan SA, VanLehn K. Collaboration detection that preserves privacy of students’ speech. In Isotani S, Hastings P, Ogan A, McLaren B, Millán E, Luckin R, editors, Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings. Springer Verlag. 2019. p. 507-517. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-23204-7_42
Viswanathan, Sree Aurovindh ; VanLehn, Kurt. / Collaboration detection that preserves privacy of students’ speech. Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings. editor / Seiji Isotani ; Peter Hastings ; Amy Ogan ; Bruce McLaren ; Eva Millán ; Rose Luckin. Springer Verlag, 2019. pp. 507-517 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{09d475c9983242e6bb090fdc74d2e3e6,
title = "Collaboration detection that preserves privacy of students’ speech",
abstract = "Collaboration is a 21st{\^A} Century skill as well as an effective method for learning, so detection of collaboration is important for both assessment and instruction. Speech-based collaboration detection can be quite accurate but collecting the speech of students in classrooms can raise privacy issues. An alternative is to send only whether or not the student is speaking. That is, the speech signal is processed at the microphone by a voice activity detector before being transmitted to the collaboration detector. Because the transmitted signal is binary (1{\^A} ={\^A} speaking, 0{\^A} ={\^A} silence), this method mitigates privacy issues. However, it may harm the accuracy of collaboration detection. To find out how much harm is done, this study compared the relative effectiveness of collaboration detectors based either on the binary signal or high-quality audio. Pairs of students were asked to work together on solving complex math problems. Three qualitative levels of interactivity was distinguished: Interaction, Cooperation and Other. Human coders used richer data (several audio and video streams) to choose the code for each episode. Machine learning was used to induce a detector to assign a code for every episode based on the features. The binary-based collaboration detectors delivered only slightly less accuracy than collaboration detectors based on the high quality audio signal.",
keywords = "Collaborative learning, Learning analytics, Machine learning",
author = "Viswanathan, {Sree Aurovindh} and Kurt VanLehn",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-23204-7_42",
language = "English (US)",
isbn = "9783030232030",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "507--517",
editor = "Seiji Isotani and Peter Hastings and Amy Ogan and Bruce McLaren and Eva Mill{\'a}n and Rose Luckin",
booktitle = "Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings",

}

TY - GEN

T1 - Collaboration detection that preserves privacy of students’ speech

AU - Viswanathan, Sree Aurovindh

AU - VanLehn, Kurt

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Collaboration is a 21st Century skill as well as an effective method for learning, so detection of collaboration is important for both assessment and instruction. Speech-based collaboration detection can be quite accurate but collecting the speech of students in classrooms can raise privacy issues. An alternative is to send only whether or not the student is speaking. That is, the speech signal is processed at the microphone by a voice activity detector before being transmitted to the collaboration detector. Because the transmitted signal is binary (1 = speaking, 0 = silence), this method mitigates privacy issues. However, it may harm the accuracy of collaboration detection. To find out how much harm is done, this study compared the relative effectiveness of collaboration detectors based either on the binary signal or high-quality audio. Pairs of students were asked to work together on solving complex math problems. Three qualitative levels of interactivity was distinguished: Interaction, Cooperation and Other. Human coders used richer data (several audio and video streams) to choose the code for each episode. Machine learning was used to induce a detector to assign a code for every episode based on the features. The binary-based collaboration detectors delivered only slightly less accuracy than collaboration detectors based on the high quality audio signal.

AB - Collaboration is a 21st Century skill as well as an effective method for learning, so detection of collaboration is important for both assessment and instruction. Speech-based collaboration detection can be quite accurate but collecting the speech of students in classrooms can raise privacy issues. An alternative is to send only whether or not the student is speaking. That is, the speech signal is processed at the microphone by a voice activity detector before being transmitted to the collaboration detector. Because the transmitted signal is binary (1 = speaking, 0 = silence), this method mitigates privacy issues. However, it may harm the accuracy of collaboration detection. To find out how much harm is done, this study compared the relative effectiveness of collaboration detectors based either on the binary signal or high-quality audio. Pairs of students were asked to work together on solving complex math problems. Three qualitative levels of interactivity was distinguished: Interaction, Cooperation and Other. Human coders used richer data (several audio and video streams) to choose the code for each episode. Machine learning was used to induce a detector to assign a code for every episode based on the features. The binary-based collaboration detectors delivered only slightly less accuracy than collaboration detectors based on the high quality audio signal.

KW - Collaborative learning

KW - Learning analytics

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=85068336671&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068336671&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-23204-7_42

DO - 10.1007/978-3-030-23204-7_42

M3 - Conference contribution

SN - 9783030232030

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 507

EP - 517

BT - Artificial Intelligence in Education - 20th International Conference, AIED 2019, Proceedings

A2 - Isotani, Seiji

A2 - Hastings, Peter

A2 - Ogan, Amy

A2 - McLaren, Bruce

A2 - Millán, Eva

A2 - Luckin, Rose

PB - Springer Verlag

ER -