Telco Customer Churn with Realistic Customer Feedback
Dataset combining telecom customer data and realistic customer feedback generated by GPT-3.5, useful for classification, sentimental analysis and churn prediction.
Approximately 13 MB, 4 CSV and PKL files, 76 columns, textual and tabular data
CC BY 4.0
Description
The dataset Telco Customer Churn with Realistic Customer Feedback extends a classic dataset on telecom customer loyalty by adding realistic customer reviews generated by GPT-3.5. Each entry includes customer data and textual feedback, making NLP and churn prediction tasks easier.
What is this dataset for?
- Train classification and sentiment analysis models
- Develop comprehensive churn prediction pipelines combining tabular and text data
- Testing NLP models with realistic synthetic data
Can it be enriched or improved?
Yes, it's possible to add real customer feedback, integrate additional customer behavior data, or refine text generation with newer models. Annotating feelings manually can also improve quality.
🔎 In summary
🧠 Recommended for
- Data scientists
- NLP developers
- Churn analysts
🔧 Compatible tools
- Pandas
- Scikit-learn
- Hugging Face Transformers
- Spark
💡 Tip
Test different training strategies by combining tabular and textual data to improve prediction.
Frequently Asked Questions
Does this dataset include real customer data or only generated feedback?
The customer data is real, but the textual feedback was generated automatically by GPT-3.5.
Can this dataset be used for commercial use?
Yes, the CC BY 4.0 license allows commercial use under attribution.
What file formats are provided?
Four files: three CSVs for tabular data and one PKL file for the saved model.
