Code Monkey home page Code Monkey logo

adv-ed's Introduction

Adv-ED

Source code and dataset for NAACL 2019 paper "Adversarial Training for Weakly Supervised Event Detection".

Requirements

  • python == 3.6.3
  • pytorch == 0.4.1
  • numpy == 1.15.2
  • sklearn == 0.20.0
  • pytorch-pretrained-bert == 0.2.0

Data

Due to the licence issues, we cannot share the source ACE2005 dataset or the preprocessed data.

So we specify the data format in DataFormat.md and you can preprocess the data follow the format.

Run

Put the preprocessed .npy data files in the same directory as the codes.

For the BERT models, download the Bert_base_uncase model in ../../BERT_CACHE.

Run python train.py in corresponding directory to train the model.

If you want to tune the hyper parameters, see the constant.py and change the parameters defined in the file.

Cite

If the codes help you, please cite the following paper:

Adversarial Training for Weakly Supervised Event Detection. Xiaozhi Wang, Xu Han, Zhiyuan Liu, Maosong Sun, Peng Li. NAACL-HLT 2019.

adv-ed's People

Contributors

bakser avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar

adv-ed's Issues

NA标签计算问题

作者您好,请问NA标签为什么使用平均数计算,不能当作一个新的标签计算吗?感谢您所做的工作!
image

DMCNN

DMCNN是说根据trigger的位置动态的将句子划分成两部分,并通过concat,cnn,进行trigger的分类,那trigger识别是怎么做出来的那?

Dscore_G参数写错了吧

dScores=Dscore_G(nwords,npos,nloc,nmaskL,nmaskR,nlabel,uwords,upos,uloc,umaskL,umaskR,ulabel)

这一行里面,根据Dscore_G函数的定义,输入参数应该是:

dScores = Dscore_G(nwords, nMask, nmaskL, nmaskR, nlabel,
                       uwords, uMask, umaskL, umaskR, ulabel)

源代码中的npos,nloc和upos,uloc都没有找到定义

dataprocess

为什么dimE设置为22而不是33那?ACE2005不是规定了33个事件子类型么?

How to identify the candidate trigger on the original training data ?

According to part 3.4 of the paper, authors used the small-scale labeled data to train the encoder and discriminator. But what should we do if there is no extra data? Do we need to identify the candidate triggers on dev and test set also?
By the way, since the NYT corpus is really huge, how many sentences did you use for augment data when semi-supervised training ?

关于论文中Generator的一些问题请教

您好,关于论文想请教您一个问题,希望能得到您赐教:

请问文中的Generator选择出困惑度高的实例(selecting the most confusing instances from U to cheat the discriminator)的目的是什么?是为了增强对D的训练(是否能理解为:将G选出的实例作为负例促进D的训练?), 还是理解为过滤掉noisy data减少对D的干扰呢?如果在您的整个系统架构中去掉G, 会有什么影响呢?

初入Event Detection领域,冒昧之处请见谅。期待您的回答,谢谢!

您好

您好,我最近也在做事件抽取的工作,但是苦于没有数据集,请问可以发我一份吗?我的邮箱是:[email protected]。非常感谢!

是否抽取事件论元

您好,目前我正在进行事件抽取的项目,需要对数据进行扩充。
浅读了您的论文,发现论文只涉及了事件的类型和触发词,请问能否提取事件的论元呢。

Recommend Projects

  • React photo React

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

  • Vue.js photo Vue.js

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

  • Typescript photo Typescript

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

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

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

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.