https://github.com/lonepatient/bert-attribute-value-extract

A Pytorch implementation of "Scaling Up Open Tagging from Tens to Thousands: Comprehension Empowered Attribute Value Extraction from Product Title" (ACL 2019).
https://github.com/lonepatient/bert-attribute-value-extract

attention attribute-value-extract bert crf lstm pytorch

Last synced: 11 days ago
JSON representation

A Pytorch implementation of "Scaling Up Open Tagging from Tens to Thousands: Comprehension Empowered Attribute Value Extraction from Product Title" (ACL 2019).

• Host: GitHub
• URL: https://github.com/lonepatient/bert-attribute-value-extract
• Owner: lonePatient
• Created: 2020-01-21T12:28:25.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2020-02-18T15:01:33.000Z (over 4 years ago)
• Last Synced: 2024-05-14T00:15:06.150Z (6 months ago)
• Topics: attention, attribute-value-extract, bert, crf, lstm, pytorch
• Language: Python
• Size: 422 KB
• Stars: 58
• Watchers: 4
• Forks: 14
• Open Issues: 4
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
## BERT FOR Attribute Value Extraction

A Pytorch implementation of "Scaling Up Open Tagging from Tens to Thousands: Comprehension Empowered Attribute Value Extraction from Product Title" (ACL 2019).

paper: https://www.aclweb.org/anthology/P19-1514.pdf

### Structure of the model

The author explicitly model the semantic representations for attribute and title, and develop an attention mechanism to capture the interactive semantic relations in-between to enforce our framework to be attribute comprehensive.

![](./outputs/model.png)

Architecture of the proposed attribute-comprehension open tagging model

### requirement

1. pytorch=1.3.0

2. cuda=9.0

### data

```text

{"id": 19, "title": "热风2019年春季新款潮流时尚男士休闲皮鞋透气低跟豆豆鞋h40m9107", "attribute": "款式", "value": "豆豆鞋"}

```

download link: [BaiDuYun](https://pan.baidu.com/s/1pChOPH0bShN5elcH2rSbcw)

### How to use the code

1. Modify the configuration information in `run_attr_crf.py` or `run_attr_crf.sh` .

2. `sh run_attr_crf.sh`