Title
Estimating Work Engagement from Online Chat Logs
Abstract
ABSTRACT The Covid-19 pandemic has transformed our lives, and in order to aid in the prevention and spread of infection, a remote work style has rapidly proliferated. As this remote work style has proliferated, new problems have come to light. One problem is that managers cannot fully grasp the engagement level of subordinates such as in terms of absorption, dedication, and vigor due to limited in-person communications. However, as a substitute for in-person communications, online communications via text-based chat tools such as Slack and Microsoft’s Teams have become popular. Recognizing the level of work engagement in a remote work setting is difficult, so we propose a new approach that estimates this level using text-based chat tools. To evaluate the proposal, we conduct experiments using actual Slack data. The experimental results reveal that the content of the conversations do not influence the level of work engagement, but the frequency of conversations among the teams and team members does. Therefore, we develop a machine learning model that estimates the level of work engagement using only the frequency and affiliation as features. The model estimates the work engagement level using true and predicted values at a correlation coefficient of 0.72. Since the proposed model uses only the frequency and affiliation, it is valuable in actual business situations.
Year
DOI
Venue
2021
10.1145/3429360.3468184
CHI
Keywords
DocType
Citations 
work engagement, machine learning, text-based communication
Conference
0
PageRank 
References 
Authors
0.34
0
3
Name
Order
Citations
PageRank
Hiroaki Tanaka100.68
Wataru Yamada22915.22
Keiichi Ochiai344.10