TY - GEN
T1 - ICS-assist
T2 - 18th International Conference on Service-Oriented Computing, ICSOC 2020
AU - Fu, Min
AU - Guan, Jiwei
AU - Zheng, Xi
AU - Zhou, Jie
AU - Lu, Jianchao
AU - Zhang, Tianyi
AU - Zhuo, Shoujie
AU - Zhan, Lijun
AU - Yang, Jian
PY - 2020
Y1 - 2020
N2 - Efficient and appropriate online customer service is essential to large e-commerce businesses. Existing solution recommendation methods for online customer service are unable to determine the best solutions at runtime, leading to poor satisfaction of end customers. This paper proposes a novel intelligent framework, called ICS-Assist, to recommend suitable customer service solutions for service staff at runtime. Specifically, we develop a generalizable two-stage machine learning model to identify customer service scenarios and determine customer service solutions based on a scenario-solution mapping table. A novel knowledge distillation network called “Panel-Student” is proposed to derive a small yet efficient distilled learning model. We implement ICS-Assist and evaluate it using an over 6-month field study with Alibaba Group. In our experiment, over 12,000 customer service staff use ICS-Assist to serve for over 230,000 cases per day on average. The experimental results show that ICS-Assist significantly outperforms the traditional manual method, and improves the solution acceptance rate, the solution coverage rate, the average service time, the customer satisfaction rate, and the business domain catering rate by up to 16%, 25%, 6%, 14% and 17% respectively, compared to the state-of-the-art methods.
AB - Efficient and appropriate online customer service is essential to large e-commerce businesses. Existing solution recommendation methods for online customer service are unable to determine the best solutions at runtime, leading to poor satisfaction of end customers. This paper proposes a novel intelligent framework, called ICS-Assist, to recommend suitable customer service solutions for service staff at runtime. Specifically, we develop a generalizable two-stage machine learning model to identify customer service scenarios and determine customer service solutions based on a scenario-solution mapping table. A novel knowledge distillation network called “Panel-Student” is proposed to derive a small yet efficient distilled learning model. We implement ICS-Assist and evaluate it using an over 6-month field study with Alibaba Group. In our experiment, over 12,000 customer service staff use ICS-Assist to serve for over 230,000 cases per day on average. The experimental results show that ICS-Assist significantly outperforms the traditional manual method, and improves the solution acceptance rate, the solution coverage rate, the average service time, the customer satisfaction rate, and the business domain catering rate by up to 16%, 25%, 6%, 14% and 17% respectively, compared to the state-of-the-art methods.
KW - Intelligent customer service
KW - Natural language processing
KW - Deep learning
KW - Distilled learning
UR - http://www.scopus.com/inward/record.url?scp=85098251582&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-65310-1_26
DO - 10.1007/978-3-030-65310-1_26
M3 - Conference proceeding contribution
AN - SCOPUS:85098251582
SN - 9783030653095
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 370
EP - 385
BT - Service-Oriented Computing
A2 - Kafeza, Eleanna
A2 - Benatallah, Boualem
A2 - Martinelli, Fabio
A2 - Hacid, Hakim
A2 - Bouguettaya, Athman
A2 - Motahari, Hamid
PB - Springer, Springer Nature
CY - Cham, Switzerland
Y2 - 14 December 2020 through 17 December 2020
ER -