TY - GEN
T1 - Automatic detection of everyday social behaviours and environments from verbatim transcripts of daily conversations
AU - Yordanova, Kristina Y.
AU - Demiray, Burcu
AU - Mehl, Matthias R.
AU - Martin, Mike
N1 - Funding Information:
K. Yordanova is funded by the German Research Foundation, grant number YO 226/1-1; at the time this research was conducted, K. Yordanova and M. R. Mehl were funded by the University of Zurich’s Digital Society Initiative in the context of the DSI Fellowships and Collegium Helveticum.
Publisher Copyright:
© 2019 IEEE.
PY - 2019/3
Y1 - 2019/3
N2 - Coding in social sciences is a process that involves the categorisation of qualitative or quantitative data in order to facilitate further analysis. Coding is usually a manual process that involves a lot of effort and time to produce codes with high validity and interrater reliability. Although automated methods for quantitative data analysis are largely used in social sciences, there are only a few attempts at automatically or semi-automatically coding the data collected in qualitative studies. To address this problem, in this work we propose an approach for automated coding of social behaviours and environments based on verbatim transcriptions of everyday conversations. To evaluate the approach, we analysed the transcripts from three datasets containing recordings of everyday conversations from: (1) young healthy adults (German transcriptions), (2) elderly healthy adults (German transcriptions), and (3) young healthy adults (English transcriptions). The results show that it is possible to automatically code the social behaviours and environments based on verbatim transcripts of the recorded conversations. This could reduce the time and effort researchers need to assign accurate codes to transcribed conversations.
AB - Coding in social sciences is a process that involves the categorisation of qualitative or quantitative data in order to facilitate further analysis. Coding is usually a manual process that involves a lot of effort and time to produce codes with high validity and interrater reliability. Although automated methods for quantitative data analysis are largely used in social sciences, there are only a few attempts at automatically or semi-automatically coding the data collected in qualitative studies. To address this problem, in this work we propose an approach for automated coding of social behaviours and environments based on verbatim transcriptions of everyday conversations. To evaluate the approach, we analysed the transcripts from three datasets containing recordings of everyday conversations from: (1) young healthy adults (German transcriptions), (2) elderly healthy adults (German transcriptions), and (3) young healthy adults (English transcriptions). The results show that it is possible to automatically code the social behaviours and environments based on verbatim transcripts of the recorded conversations. This could reduce the time and effort researchers need to assign accurate codes to transcribed conversations.
KW - Automated coding
KW - Natural language processing
KW - Social behaviour analysis
UR - http://www.scopus.com/inward/record.url?scp=85070195343&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85070195343&partnerID=8YFLogxK
U2 - 10.1109/PERCOM.2019.8767403
DO - 10.1109/PERCOM.2019.8767403
M3 - Conference contribution
AN - SCOPUS:85070195343
T3 - 2019 IEEE International Conference on Pervasive Computing and Communications, PerCom 2019
BT - 2019 IEEE International Conference on Pervasive Computing and Communications, PerCom 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2019 IEEE International Conference on Pervasive Computing and Communications, PerCom 2019
Y2 - 12 March 2019 through 14 March 2019
ER -