Text communication using smartphone is ubiquitous, but small screens and dense keyboard structures frequently cause typing errors that hinder user experience. Previous research has primarily focused on overall error rates, with limited investigation into microscopic error patterns or the specific combinational characteristics of the Korean language. This study utilizes a dataset of smartphone QWERTY keyboard inputs to address this gap. Data was preprocessed by decomposing Korean characters into individual graphemes—consonants and vowels—allowing for a granular comparison of error occurrences. Typing errors were quantitatively analyzed and classified into spatial categories, comprising horizontal and vertical errors, and temporal categories, including ordering and omission errors. The analysis revealed that spatial errors were significantly more prevalent than temporal errors, with horizontal errors resulting from adjacent key interference being the most frequent type. Furthermore, distinct error patterns emerged based on whether the input was a consonant or a vowel, as well as its specific row location on the keyboard. These results provide critical baseline data for optimizing ergonomic keyboard layouts for mobile devices.
목차
Abstract 1. 서론 2. 이론적 배경 3. 연구 방법 3.1 데이터셋 설명 3.2 오류 추출 및 전처리 방법 3.3 오류 유형 분류기준 3.4 키보드 매핑 및 오류율 산출 방식 4. 분석 결과 4.1 오류 유형별 전체 분포 4.2 자음-모음 키에 따른 오류 4.3 키열에 따른 오류율 분석 4.4. 키행에 따른 오류율 분석 4.5. 키별 분석 5. 연구 결과의 논의 6. 결론 [References]