Product-Name-NER-model
| Property | Value |
|---|---|
| Parameter Count | 38.5M |
| License | Apache 2.0 |
| Language | Traditional Chinese |
| Framework | BERT-based |
What is Product-Name-NER-model?
Product-Name-NER-model is a specialized Named Entity Recognition (NER) model designed for Traditional Chinese product descriptions. This model excels at extracting 16 different attributes from product names, ranging from basic information like brand names and colors to detailed specifications and packaging details.
Implementation Details
Built on BERT architecture with 38.5M parameters, this model demonstrates robust performance with a macro F1-score of 0.7807. It particularly excels in identifying capacity (0.9220 F1), weight (0.9105 F1), and color attributes (0.8898 F1).
- Optimized for Traditional Chinese text processing
- Supports 16 distinct attribute categories
- Uses token classification approach for precise entity recognition
Core Capabilities
- Brand identification (F1: 0.8770)
- Product series and name extraction (F1: 0.5941)
- Physical attributes detection (color, size, weight)
- Target audience classification
- Package and specification analysis
Frequently Asked Questions
Q: What makes this model unique?
This model's uniqueness lies in its comprehensive approach to product attribute extraction, covering 16 different aspects of product descriptions in Traditional Chinese. Its particular strength in identifying technical specifications and physical attributes makes it valuable for e-commerce applications.
Q: What are the recommended use cases?
The model is ideal for e-commerce platforms, product cataloging systems, and inventory management solutions that deal with Chinese product descriptions. It's particularly effective for automated product classification and attribute extraction in large-scale product databases.
