File size: 1,131 Bytes
f402c48
 
a8fd0c9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f402c48
48d434f
 
 
dc3930b
48d434f
8e306c8
0225b00
aa4459a
 
48d434f
 
18e9d83
48d434f
 
 
9d13433
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
---
license: apache-2.0
language:
- ca
- da
- de
- en
- es
- fr
- nl
- el
- is
- it
- 'no'
- pt
- sv
pipeline_tag: text-classification
---
# Occupational CANINE: HISCO Classification Model

## Overview
OccCANINE is a version of [CANINE](https://huggingface.co/google/canine-s) which has been finetuned to automatically convert occupational descriptions into standardized HISCO codes using a CANINE model. This tool facilitates historical occupational data analysis with over 90% accuracy across 13 languages.

See more on: [GitHub.com/christianvedels/OccCANINE](https://github.com/christianvedels/OccCANINE)

Read the paper on arXiv: [https://arxiv.org/abs/2402.13604](https://arxiv.org/abs/2402.13604)

## Key Features
- **High Accuracy**: Over 90% accuracy, recall, and precision.
- **Multilingual Support**: Trained on 14 million description-HISCO code pairs across 13 languages.
- **Efficiency**: Rapidly processes descriptions into HISCO codes.

## Contribution and Support
Developed at the University of Southern Denmark by Christian Møller Dahl, Torben Johansen and Christian Vedel with contributions from various sources.