International Conference on Information Technology and Computer Science, 3rd (ITCS 2011)
80 Biological Naming Entity Recognition Basing on Conditional Random Fields
Download citation file:
- Ris (Zotero)
- Reference Manager
This paper introduces a recognition approach for identifying genes, proteins and other biological entities in biomedical literatures. This approach bases on conditional random fields and combines morphological characteristics of words and distance dependent characteristics as features to conduct the recognition. It conducts experiments on GENIAV3.02 data of JNLPBA. The testing results show that after the introduction of distance dependent, the recognition performance of system increases by 2.54% compared with the conditional random fields which adopting only single feature condition, this can obtain a better recognition performance and improve the efficiency of systematic identification.