Using Closed-Set Speaker Identification Score Confidence to Enhance Audio-Based Collaborative Filtering for Multiple Users

Research output: Contribution to journalJournal articleResearchpeer-review

1 Citation (Scopus)
167 Downloads (Pure)

Abstract

In this paper, we utilize a closed-set speaker-identification approach to convey the ratings needed for collaborative filtering-based recommendation. Instead of explicitly providing a rating for a given program, users use a speech interface to dictate the desired rating after watching a movie. Due to the inaccuracies that may be imposed by a state-of-the-art speaker identification system, it is possible to mistake a user for another user in the household, especially when the users exhibit similar or identical age and gender demographics. This leads to the undesirable effect of injecting unwanted ratings into the collaborative rating matrix, and when the users have different tastes, can result in the recommendation of undesirable items. We therefore propose a simple confidence-based heuristic that utilizes the log-likelihood scores from the speaker identification front-end. The algorithm limits the degree to which unwanted ratings negatively affect the integrity of the ratings information. Using real-speaker utterances over a range of age and gender demographics, we compare our approach against upper and lower-bound (nonspeaker-identification-based) baseline systems. Results show that by taking the confidence into account of users that we were able to improve upon the lower-bound that unconditionally accepts ratings by a relative 6.9%.
Original languageEnglish
JournalIEEE Transactions on Consumer Electronics
Volume64
Issue number1
Pages (from-to)11-18
Number of pages8
ISSN0098-3063
DOIs
Publication statusPublished - 2018

Fingerprint

Collaborative filtering
Identification (control systems)

Cite this

@article{793bbba249de463ca66751a5a150a493,
title = "Using Closed-Set Speaker Identification Score Confidence to Enhance Audio-Based Collaborative Filtering for Multiple Users",
abstract = "In this paper, we utilize a closed-set speaker-identification approach to convey the ratings needed for collaborative filtering-based recommendation. Instead of explicitly providing a rating for a given program, users use a speech interface to dictate the desired rating after watching a movie. Due to the inaccuracies that may be imposed by a state-of-the-art speaker identification system, it is possible to mistake a user for another user in the household, especially when the users exhibit similar or identical age and gender demographics. This leads to the undesirable effect of injecting unwanted ratings into the collaborative rating matrix, and when the users have different tastes, can result in the recommendation of undesirable items. We therefore propose a simple confidence-based heuristic that utilizes the log-likelihood scores from the speaker identification front-end. The algorithm limits the degree to which unwanted ratings negatively affect the integrity of the ratings information. Using real-speaker utterances over a range of age and gender demographics, we compare our approach against upper and lower-bound (nonspeaker-identification-based) baseline systems. Results show that by taking the confidence into account of users that we were able to improve upon the lower-bound that unconditionally accepts ratings by a relative 6.9{\%}.",
author = "Shepstone, {Sven Ewan} and Zheng-Hua Tan and Kristoffersen, {Miklas Str{\o}m}",
year = "2018",
doi = "10.1109/TCE.2018.2811250",
language = "English",
volume = "64",
pages = "11--18",
journal = "I E E E Transactions on Consumer Electronics",
issn = "0098-3063",
publisher = "IEEE",
number = "1",

}

Using Closed-Set Speaker Identification Score Confidence to Enhance Audio-Based Collaborative Filtering for Multiple Users. / Shepstone, Sven Ewan; Tan, Zheng-Hua; Kristoffersen, Miklas Strøm.

In: IEEE Transactions on Consumer Electronics, Vol. 64, No. 1, 2018, p. 11-18.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Using Closed-Set Speaker Identification Score Confidence to Enhance Audio-Based Collaborative Filtering for Multiple Users

AU - Shepstone, Sven Ewan

AU - Tan, Zheng-Hua

AU - Kristoffersen, Miklas Strøm

PY - 2018

Y1 - 2018

N2 - In this paper, we utilize a closed-set speaker-identification approach to convey the ratings needed for collaborative filtering-based recommendation. Instead of explicitly providing a rating for a given program, users use a speech interface to dictate the desired rating after watching a movie. Due to the inaccuracies that may be imposed by a state-of-the-art speaker identification system, it is possible to mistake a user for another user in the household, especially when the users exhibit similar or identical age and gender demographics. This leads to the undesirable effect of injecting unwanted ratings into the collaborative rating matrix, and when the users have different tastes, can result in the recommendation of undesirable items. We therefore propose a simple confidence-based heuristic that utilizes the log-likelihood scores from the speaker identification front-end. The algorithm limits the degree to which unwanted ratings negatively affect the integrity of the ratings information. Using real-speaker utterances over a range of age and gender demographics, we compare our approach against upper and lower-bound (nonspeaker-identification-based) baseline systems. Results show that by taking the confidence into account of users that we were able to improve upon the lower-bound that unconditionally accepts ratings by a relative 6.9%.

AB - In this paper, we utilize a closed-set speaker-identification approach to convey the ratings needed for collaborative filtering-based recommendation. Instead of explicitly providing a rating for a given program, users use a speech interface to dictate the desired rating after watching a movie. Due to the inaccuracies that may be imposed by a state-of-the-art speaker identification system, it is possible to mistake a user for another user in the household, especially when the users exhibit similar or identical age and gender demographics. This leads to the undesirable effect of injecting unwanted ratings into the collaborative rating matrix, and when the users have different tastes, can result in the recommendation of undesirable items. We therefore propose a simple confidence-based heuristic that utilizes the log-likelihood scores from the speaker identification front-end. The algorithm limits the degree to which unwanted ratings negatively affect the integrity of the ratings information. Using real-speaker utterances over a range of age and gender demographics, we compare our approach against upper and lower-bound (nonspeaker-identification-based) baseline systems. Results show that by taking the confidence into account of users that we were able to improve upon the lower-bound that unconditionally accepts ratings by a relative 6.9%.

U2 - 10.1109/TCE.2018.2811250

DO - 10.1109/TCE.2018.2811250

M3 - Journal article

VL - 64

SP - 11

EP - 18

JO - I E E E Transactions on Consumer Electronics

JF - I E E E Transactions on Consumer Electronics

SN - 0098-3063

IS - 1

ER -