EMNLP-2009-Generating High-Coverage Semantic Orientation Lexicons From Overtly Marked Words and a Thesaurus

[BrambleXu]

Summary:

基于lexicon的方法计算太耗时了，所以提出一个简单的方案能生成 high-coverage semantic orientation lexicon

Resource:

• pdf
• [code](
• [paper-with-code](

Paper information:

• Author:
• Dataset:
• keywords:

Notes:

4 Evaluation

We evaluated the semantic orientation lexicons both intrinsically (by comparing their entries to the General Inquirer) and extrinsically (by using them in a phrase polarity annotation task).

这里提到的是从内部和外部两种方式来检测lexicon。intrinsically的方式其实就是直接计算覆盖率的方法，而extrinsically的方式其实就是用NER的方式来进行计算。（只不过用于我的研究的话，recall的部分有重复。前者用一个百分比来对比，后者用一个task来对比）

Model Graph:

Result:：

Thoughts:

Next Reading: