BrambleXu/knowledge-graph-learning

RepEval(WS)-2016-Evaluating Word Embeddings Using a Representative Suite of Practical Tasks #301

BrambleXu posted onGitHub

Summary:

评价word embedding的方法只依靠word similarity benchmark不好,所以本文提出了vivo,通过一系列的downstream tasks来评价word embeddings。还有online的版本,提交后就能自动进行评价。

Resource:

  • pdf
  • [code](
  • [paper-with-code](

Paper information:

  • Author: Manning
  • Dataset:
  • keywords:

Notes:

关于评价的话,其实可以多用几个数据集,来证明一下,基于character的语言的coverage对于NER的评判 (让评价更方便的系统,也可以写一篇论文)

Model Graph:

Result:

image

image

Thoughts:

Next Reading:


Fund this Issue

$0.00
Funded
Only logged in users can fund an issue

Pull requests