Abstract
Contact centers, often referred to as call centers, can deal with a large number of calls each day. Identifying commonly occurring topics of calls can help call center administrators with resource allocation and prioritization. Many existing topic analysis call center solutions utilize historical data, inference on new calls, and some degree of personalization for determining topic trends. However, a modern-day topic analytics dashboard for a call center should also facilitate a higher degree of personalization in terms of adding existing topic labels, identifying new topics, and monitoring/alerting administrators when models are stale and in need of re-training. Various techniques are proposed herein to facilitate seamless maintenance of call center models in order to achieve a higher degree of personalization of the models.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Embar, Varsha; Shrivastava, Ritvik; and Raghunathan, Karthik, "AUTOMATING UNSUPERVISED TOPIC TREND DETECTION IN CALL CENTERS", Technical Disclosure Commons, (October 22, 2024)
https://www.tdcommons.org/dpubs_series/7464