Update README.md
Browse files
README.md
CHANGED
|
@@ -17,13 +17,34 @@ tags:
|
|
| 17 |
- BERT
|
| 18 |
---
|
| 19 |
|
| 20 |
-
# Banking
|
| 21 |
|
| 22 |
-
|
| 23 |
|
| 24 |
-
|
| 25 |
-
|
| 26 |
-
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
- BERT
|
| 18 |
---
|
| 19 |
|
| 20 |
+
# Banking Customer Service Intent Classifier
|
| 21 |
|
| 22 |
+
This model is designed to classify customer service queries into different intents, based on the type of inquiry made by the customer. It was fine-tuned on a **synthetic** dataset of realistic banking customer service interactions and can classify the following intents:
|
| 23 |
|
| 24 |
+
- `transaction_query`
|
| 25 |
+
- `password_reset`
|
| 26 |
+
- `loan_inquiry`
|
| 27 |
+
- `fraud_report`
|
| 28 |
+
- `credit_card_application`
|
| 29 |
+
- `balance_inquiry`
|
| 30 |
+
|
| 31 |
+
## Model Overview
|
| 32 |
+
|
| 33 |
+
The model is a fine-tuned BERT-based architecture that classifies text inputs into one of the six specified intents. It leverages the **transformers** library by Hugging Face for tokenization and model loading.
|
| 34 |
+
|
| 35 |
+
## Intended Use
|
| 36 |
+
|
| 37 |
+
This model is suitable for deployment in applications that require automatic classification of customer service queries, such as:
|
| 38 |
+
|
| 39 |
+
- Chatbots
|
| 40 |
+
- Virtual assistants
|
| 41 |
+
- Automatic re-routing incoming emails,calls, and texts
|
| 42 |
+
|
| 43 |
+
It can be used to classify various types of banking queries, such as requests for account balance, loan inquiries, or fraud reports.
|
| 44 |
+
|
| 45 |
+
## Installation
|
| 46 |
+
|
| 47 |
+
To install the necessary dependencies, use the following:
|
| 48 |
+
|
| 49 |
+
```bash
|
| 50 |
+
pip install transformers torch
|