Abstract:
Abstract
A named entity (NE) is a word or a phrase that clearly identifies one item from a set
of other items that have similar features. The term named entity refers to organization,
person, and location names in a text. Named entity recognition (NER) is the process of
locating and classifying named entities in text into predefined entity categories.
The proposed approach to Gamo language entity recognition involves four major layered
processes. These are corpus collection and dataset development, distributed representation, context encoder, and tag decoder layer, all the layered processes are interconnected
and arranged in a way that an output of a current process is an input to the next process. For distributed representation character and word level embedding used, context
encoding BiLSTM was used, and decoding CRF was used.
NER for GL with deep learning based on features that extracted from word-level embedding alone achieved performance of accuracy 93.94%, precision 73.96%, recall 68.62% and
F1-Score 71.19%, and word and character level embedding together achieved performance
accuracy 94.59%, precision 76.15%, recall 72.49%, F1-Score 74.28%. So the second model
improves precision 2.19%, recall 3.87% and F1-Score 3.09 respectively.
NER for GL with deep learning based on features that extracted with word-level embedding in conjunction with character level embedding outperforms deep learning based on
feature extracted with word-level embedding.
Keyword:GLNER,Encoder, Decoder
