E-commerce
Sentiment analysis and annotation best practices helped to increase efficiency 5 times.

Training data for AI and machine learning in e-commerce

About case:
About case:
Industry and use case
Retail and e-commerce,
sentiment analysis
DATA
200,000 customer reviews
Project duration
30 days
Industry and use case
Retail and e-commerce,
sentiment analysis
DATA
200,000 customer reviews
Project duration
30 days
Challenge:
Challenge:
The client wanted to develop a neural network that could help determine customer reviews' topics and perform sentiment analysis. They contacted Training Data to annotate 200,000 customer reviews.
Solution:
Solution:
The 40-person team from Training Data classified 10+ topics and 5+ sentiments. The personal manager was in touch with the customer 24/7, making it possible to quickly discuss and change the original project scope and labelling requirements. In addition, the team helped to redesign the work of the customer service team to improve efficiency.
Outcomes:
Outcomes:
High-quality data accelerated customer support processes 5 times
Automatic sentiment analysis and review classification
Real-time identification of negative reviews and instant follow-ups
Feedback
The annotators and project manager not only quickly labelled a large quantity of data but also redesigned a process that improved the entire work of our department. Very efficient and professional team!
Mark T. ML Team Lead in a Bank