Call Center Text Sentiment Analysis Solution

• Call center data has been extracted as document
• Wrong spelling mistakes have been fixed for Turkish and English
• Models that has been created with DL Algorithms (Word2Vec, Doc2Vec, FastText), have been trained 80 percent of call center data and tested 20 percent of data
• Sentimental analysis have been realized on the customer’s conversation data

Input(s):

• Sentiment Analysis Tool

Main feature(s)

• Each part is implemented in Python programming language using the algorithms that are coming from machine and deep learning text analytics literature. The Big Data environments that include MLLib Libraries, ocumentbased NoSQL databases like MongoDB, Real Time Stream Analytic tools like Spark Streaming and frastructure like Spark cluster used as a platform

Output(s):

• Intent classification Module
• Customer satisfaction measurement Module
• Agent performance measurement Module
• Entity Recognition Module
• Spelling Correction Module

• Integration pre-study is needed since different call centers can use different infrastructure
• Data can be trained due to the needs of the different call-centers

Call Centers

Licencing