Light

platanus-hy / sememes_codriven_text_matching Goto Github PK

View Code? Open in Web Editor NEW

4.0 2.0 0.0 10.05 MB

Co-Driven Recognition of Semantic Consistency via the Fusion of Transformer and HowNet Sememes Knowledge

Python 100.00%

chinese-nlp hownet knowledge-representation semantic-analysis transformer-encoder knowledge-data-codriven semantic-consistency sememes-knowledge

sememes_codriven_text_matching's Introduction

Co-Driven Recognition of Semantic Consistency via the Fusion of Transformer and HowNet Sememes Knowledge

This is the code of our paper underreview for ESWC-2023, there are some explanations for it

For non-pretraining models, you need to run Pre-processing.py to generate data before running the models. For pretaining models, please download BERT models before you run hownet_bert.py.
We just take BERT model and BQ dataset for example, it is easy to expand to other text semantic matching datasets or replace with other pretraining models.
Our experimental results on the BQ, AFQMC and PAWSX-zh datasets are as follows:

BQ corpus

models	pretaining model	Acc	F1
DSSM	×	77.12	76.47
MwAN	×	73.99	73.29
DRCN	×	74.65	76.02
Ours	×	78.81	76.62
Improvement	×	+2.19%	+1.96%
BERT-wwm-ext	√	84.71	83.94
BERT	√	84.50	84.00
ERNIE	√	84.67	84.20
Ours-BERT	√	84.82	84.33
Improvement	√	+0.177%	+0.464%

AFQMC dataset

models	pretaining model	Acc	F1
DSSM	×	57.02	30.75
MwAN	×	65.43	28.63
DRCN	×	66.05	40.60
Ours	×	66.62	42.93
Improvement	×	+0.86%	+5.7%
BERT-wwm-ext	√	81.76	80.62
BERT	√	81.43	79.77
ERNIE	√	81.54	80.81
Ours-BERT	√	81.84	81.93
Improvement	√	+0.097%	+1.38%

PAWSX-zh dataset

models	pretaining model	Acc	F1
DSSM	×	42.64	59.43
MwAN	×	52.70	52.65
DRCN	×	61.24	56.52
Ours	×	62.55	59.72
Improvement	×	+2.13%	+0.48%
BERT-wwm-ext	√	77.23	76.52
BERT	√	77.06	77.16
ERNIE	√	78.02	77.59
Ours-BERT	√	78.33	77.96
Improvement	√	+0.397%	+0.476%

sememes_codriven_text_matching's People

Contributors

Stargazers

Watchers

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.