Named entity recognition based on Verbs Associated with Human Activities (VAHA)
Motivated by a hypothesis such that there always exists verb specifically describes human being conduct, this thesis propose a NER system which aims to identify NEs that perform human activity based on verb analysis in an autonomous manner. In other words, this research attempt to identify NEs by o...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2013
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | Motivated by a hypothesis such that there always exists verb specifically describes human being conduct, this thesis propose a NER system which aims to identify NEs that perform human activity based on verb analysis in an autonomous manner. In other words, this research attempt to identify NEs by observing the existence of Verbs Associated with Human Activities (VAHA). VAHA architecture exhibits five significant characteristics where (i) it requires no training data for NER, thus eliminating human intervention (manually labeling data set or creating gazetteer), (ii) it is applicable to small text size and even on a single text, (iii) it is adaptable across different domains (fiction and non-fiction), (iv) it requires no gazetteer (person name list) during NER and (v) no anaphora resolution is required prior to NER. |
|---|