C5i Case Studies Gained customer experience insights from conversations using advanced text analytics and machine learning techniques

Edit This Case Study Record

	Gained customer experience insights from conversations using advanced text analytics and machine learning techniques C5i

Gained customer experience insights from conversations using advanced text analytics and machine learning techniques

C5i

Technology Category	Analytics & Modeling - Machine Learning Analytics & Modeling - Natural Language Processing (NLP)
Services	Data Science Services
Challenge	The client, a leading technology company, was capturing consumer conversations from multiple channels such as tech forums, call center agent transcripts, email, and chat. They were categorizing the customer experience (text data) manually on a minimized sample basis which did not provide a holistic view of customer experience insights. The client wanted to automate the text classification on the entire so they could gather more detailed insights about the customer experience. Read More
About Customer	The customer is a leading technology company in the Information Technology industry. They were capturing consumer conversations from multiple channels and categorizing the customer experience manually on a minimized sample basis. This method did not provide a holistic view of customer experience insights. The company wanted to automate the text classification on the entire data set to gather more detailed insights about the customer experience. Read More
Solution	Blueocean Market Intelligence simplified the complex hierarchical structure to ensure the appropriate classification as per business requirements. They used Natural Language Processing techniques such as verb replacement, lemmatization, normalization, tokenization, and stemming to ensure that the categories are captured. They converted multi-class records into single class by applying a basic as well as weightage technique. Key features were extracted by determining the importance of the terms by determining the weight of combinations. A Ridge Classifier was used to classify comments based on defined business rules. Read More Log in to view content
Contents

Technology Category

Analytics & Modeling - Machine Learning

Analytics & Modeling - Natural Language Processing (NLP)

Services

Data Science Services

Challenge

The client, a leading technology company, was capturing consumer conversations from multiple channels such as tech forums, call center agent transcripts, email, and chat. They were categorizing the customer experience (text data) manually on a minimized sample basis which did not provide a holistic view of customer experience insights. The client wanted to automate the text classification on the entire so they could gather more detailed insights about the customer experience.

About Customer

The customer is a leading technology company in the Information Technology industry. They were capturing consumer conversations from multiple channels and categorizing the customer experience manually on a minimized sample basis. This method did not provide a holistic view of customer experience insights. The company wanted to automate the text classification on the entire data set to gather more detailed insights about the customer experience.

Solution

Blueocean Market Intelligence simplified the complex hierarchical structure to ensure the appropriate classification as per business requirements. They used Natural Language Processing techniques such as verb replacement, lemmatization, normalization, tokenization, and stemming to ensure that the categories are captured. They converted multi-class records into single class by applying a basic as well as weightage technique. Key features were extracted by determining the importance of the terms by determining the weight of combinations. A Ridge Classifier was used to classify comments based on defined business rules.

Impact #1	Achieved an accuracy close to 70%
Impact #2	Provided a roadmap to enable the future classification of issues on general blogs and forums

Impact #1

Achieved an accuracy close to 70%

Impact #2

Provided a roadmap to enable the future classification of issues on general blogs and forums

Benefit #1	20% efficiency improvement
Benefit #2	400x reduced costs per coded case
Benefit #3	2% to 100% inclusion

Benefit #1

20% efficiency improvement

Benefit #2

400x reduced costs per coded case

Benefit #3

2% to 100% inclusion

Download PDF Version

Overview

Gained customer experience insights from conversations using advanced text analytics and machine learning techniques

Operational Impact

Quantitative Benefit