Automated program repair is in the spotlight in academia and the field to reduce the time and cost of maintenance for developers. Recently, APR has continued to study based on deep learning models to understand and learn how to fix software bugs. Text-to-Text Transfer Transformer(T5), which scored state- of-the-art in NLP benchmarks, also showed promising result on program repair in recent studies. In deep learning based program repair, abstraction techniques are commonly used to avoid vocabulary problems and learn fine code transformation for generating bug fixing patches. However, there is not enough systematic analysis of code abstraction according to each feature of the bug in deep-learning based program repair. We leverage TFix, Text-to-Text transformer based neural program repair, to evaluate how code abstraction technique affect neural program repair. Following our results, defect specific code abstraction gets a higher average BLEU score than the formal code abstraction technique in both of T5 and multilingual-T5(mT5) model based TFix results. Also, mT5 model based TFix, which is applied defect specific code abstraction gets a higher BLEU score in 31 error types of 52 ESLint error types than TFix.