This data set has 312912 rows and 14 columns. for our Prediction we worked with just two columns, which are the issue columns later renamed to Consumer_complaint_narrative as the text features and the product column as the target(label), and our generating predictive model will be MultinomialNB under naives bayes classifier
Column Position | Atrribute Name | Definition | Data Type |
---|---|---|---|
2 | PRODUCT | COMPLIANT CATEGORY | Qualitative |
3 | ISSUES | Consumer complaint narrative | Qualitative |