Inferring social nature of conversations from words: Experiments on a corpus of everyday telephone conversations

Anthony Stark, Izhak Shafran, Jeffrey Kaye

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Language is being increasingly harnessed to not only create natural human-machine interfaces but also to infer social behaviors and interactions. In the same vein, we investigate a novel spoken language task, of inferring social relationships in two-party conversations: whether the two parties are related as family, strangers or are involved in business transactions. For our study, we created a corpus of all incoming and outgoing calls from a few homes over the span of a year. On this unique naturalistic corpus of everyday telephone conversations, which is unlike Switchboard or any other public domain corpora, we demonstrate that standard natural language processing techniques can achieve accuracies of about 88%, 82%, 74% and 80% in differentiating business from personal calls, family from non-family calls, familiar from unfamiliar calls and family from other personal calls respectively. Through a series of experiments with our classifiers, we characterize the properties of telephone conversations and find: (a) that 30 words of openings (beginnings) are sufficient to predict business from personal calls, which could potentially be exploited in designing context sensitive interfaces in smart phones; (b) our corpus-based analysis does not support Schegloff and Sack's manual analysis of exemplars in which they conclude that pre-closings differ significantly between business and personal calls - closing fared no better than a random segment; and (c) the distribution of different types of calls are stable over durations as short as 1-2 months. In summary, our results show that social relationships can be inferred automatically in two-party conversations with sufficient accuracy to support practical applications.

Original languageEnglish (US)
Pages (from-to)224-239
Number of pages16
JournalComputer Speech and Language
Volume28
Issue number1
DOIs
StatePublished - 2014

Fingerprint

Telephone
Telephone switchboards
Experiment
Industry
Experiments
Sufficient
Human-machine Interface
Social Behavior
Social Interaction
Veins
Natural Language
Transactions
Classifiers
Classifier
Predict
Series
Business
Corpus
Processing
Demonstrate

Keywords

  • Conversation telephone speech
  • Social networks
  • Social relationships

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Theoretical Computer Science

Cite this

Inferring social nature of conversations from words : Experiments on a corpus of everyday telephone conversations. / Stark, Anthony; Shafran, Izhak; Kaye, Jeffrey.

In: Computer Speech and Language, Vol. 28, No. 1, 2014, p. 224-239.

Research output: Contribution to journalArticle

@article{1cd005c61ec9469db84abb8dea9a8cc1,
title = "Inferring social nature of conversations from words: Experiments on a corpus of everyday telephone conversations",
abstract = "Language is being increasingly harnessed to not only create natural human-machine interfaces but also to infer social behaviors and interactions. In the same vein, we investigate a novel spoken language task, of inferring social relationships in two-party conversations: whether the two parties are related as family, strangers or are involved in business transactions. For our study, we created a corpus of all incoming and outgoing calls from a few homes over the span of a year. On this unique naturalistic corpus of everyday telephone conversations, which is unlike Switchboard or any other public domain corpora, we demonstrate that standard natural language processing techniques can achieve accuracies of about 88{\%}, 82{\%}, 74{\%} and 80{\%} in differentiating business from personal calls, family from non-family calls, familiar from unfamiliar calls and family from other personal calls respectively. Through a series of experiments with our classifiers, we characterize the properties of telephone conversations and find: (a) that 30 words of openings (beginnings) are sufficient to predict business from personal calls, which could potentially be exploited in designing context sensitive interfaces in smart phones; (b) our corpus-based analysis does not support Schegloff and Sack's manual analysis of exemplars in which they conclude that pre-closings differ significantly between business and personal calls - closing fared no better than a random segment; and (c) the distribution of different types of calls are stable over durations as short as 1-2 months. In summary, our results show that social relationships can be inferred automatically in two-party conversations with sufficient accuracy to support practical applications.",
keywords = "Conversation telephone speech, Social networks, Social relationships",
author = "Anthony Stark and Izhak Shafran and Jeffrey Kaye",
year = "2014",
doi = "10.1016/j.csl.2013.06.003",
language = "English (US)",
volume = "28",
pages = "224--239",
journal = "Computer Speech and Language",
issn = "0885-2308",
publisher = "Academic Press Inc.",
number = "1",

}

TY - JOUR

T1 - Inferring social nature of conversations from words

T2 - Experiments on a corpus of everyday telephone conversations

AU - Stark, Anthony

AU - Shafran, Izhak

AU - Kaye, Jeffrey

PY - 2014

Y1 - 2014

N2 - Language is being increasingly harnessed to not only create natural human-machine interfaces but also to infer social behaviors and interactions. In the same vein, we investigate a novel spoken language task, of inferring social relationships in two-party conversations: whether the two parties are related as family, strangers or are involved in business transactions. For our study, we created a corpus of all incoming and outgoing calls from a few homes over the span of a year. On this unique naturalistic corpus of everyday telephone conversations, which is unlike Switchboard or any other public domain corpora, we demonstrate that standard natural language processing techniques can achieve accuracies of about 88%, 82%, 74% and 80% in differentiating business from personal calls, family from non-family calls, familiar from unfamiliar calls and family from other personal calls respectively. Through a series of experiments with our classifiers, we characterize the properties of telephone conversations and find: (a) that 30 words of openings (beginnings) are sufficient to predict business from personal calls, which could potentially be exploited in designing context sensitive interfaces in smart phones; (b) our corpus-based analysis does not support Schegloff and Sack's manual analysis of exemplars in which they conclude that pre-closings differ significantly between business and personal calls - closing fared no better than a random segment; and (c) the distribution of different types of calls are stable over durations as short as 1-2 months. In summary, our results show that social relationships can be inferred automatically in two-party conversations with sufficient accuracy to support practical applications.

AB - Language is being increasingly harnessed to not only create natural human-machine interfaces but also to infer social behaviors and interactions. In the same vein, we investigate a novel spoken language task, of inferring social relationships in two-party conversations: whether the two parties are related as family, strangers or are involved in business transactions. For our study, we created a corpus of all incoming and outgoing calls from a few homes over the span of a year. On this unique naturalistic corpus of everyday telephone conversations, which is unlike Switchboard or any other public domain corpora, we demonstrate that standard natural language processing techniques can achieve accuracies of about 88%, 82%, 74% and 80% in differentiating business from personal calls, family from non-family calls, familiar from unfamiliar calls and family from other personal calls respectively. Through a series of experiments with our classifiers, we characterize the properties of telephone conversations and find: (a) that 30 words of openings (beginnings) are sufficient to predict business from personal calls, which could potentially be exploited in designing context sensitive interfaces in smart phones; (b) our corpus-based analysis does not support Schegloff and Sack's manual analysis of exemplars in which they conclude that pre-closings differ significantly between business and personal calls - closing fared no better than a random segment; and (c) the distribution of different types of calls are stable over durations as short as 1-2 months. In summary, our results show that social relationships can be inferred automatically in two-party conversations with sufficient accuracy to support practical applications.

KW - Conversation telephone speech

KW - Social networks

KW - Social relationships

UR - http://www.scopus.com/inward/record.url?scp=84885421426&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84885421426&partnerID=8YFLogxK

U2 - 10.1016/j.csl.2013.06.003

DO - 10.1016/j.csl.2013.06.003

M3 - Article

AN - SCOPUS:84885421426

VL - 28

SP - 224

EP - 239

JO - Computer Speech and Language

JF - Computer Speech and Language

SN - 0885-2308

IS - 1

ER -