Authors :
Eshita Gangwar, Rutvi Sutaria.
Volume/Issue :
Volume 3 - 2018, Issue 12 - December
Google Scholar :
https://goo.gl/DF9R4u
Scribd :
https://goo.gl/Gugs8m
Thomson Reuters ResearcherID :
https://goo.gl/KTXLC3
Abstract :
Customer service centers are most crucial part of any company. They represent the company and communicate with the customers on its behalf. These centers also impart valuable information through customer feedback. As they form the bridge between the company and its customers, it is important they convey information timely and effectively. Emails are the most popular means of business communication. In order to efficiently and effectively utilize time, the customer service centers need to extract the relevant information from these emails. Then they need organizing it and promptly respond to the customers accordingly. In this paper, we propose categorizing emails based on the customer reviews. The polarity (positive or negative) of the customer emails is determined along with its probability and also determine the topic of the email. After preliminary data pre-processing, we use the NaiveBayes algorithm based approach to classify the emails and then the topic modeling is performed. This way, the project helps to classify emails and determine their subjects to save valuable time for the employees.
Keywords :
Emails, Customer Care, Categorization, Topic Modeling.
Customer service centers are most crucial part of any company. They represent the company and communicate with the customers on its behalf. These centers also impart valuable information through customer feedback. As they form the bridge between the company and its customers, it is important they convey information timely and effectively. Emails are the most popular means of business communication. In order to efficiently and effectively utilize time, the customer service centers need to extract the relevant information from these emails. Then they need organizing it and promptly respond to the customers accordingly. In this paper, we propose categorizing emails based on the customer reviews. The polarity (positive or negative) of the customer emails is determined along with its probability and also determine the topic of the email. After preliminary data pre-processing, we use the NaiveBayes algorithm based approach to classify the emails and then the topic modeling is performed. This way, the project helps to classify emails and determine their subjects to save valuable time for the employees.
Keywords :
Emails, Customer Care, Categorization, Topic Modeling.
