Using Augmented Cognition to Examine Differences in Online Handwriting Recognition for Native and Non-native Writers

The handwritten recognition (HWR) is a complex task with variety of challenges associated with natural language, variety in the styles of writing, variety and nuances of alphabets etc. The core research in handwritten recognition focuses around Latin alphabet and corresponding languages. However, differences between the languages using Latin as their main script are still major: from changed letter frequencies to additional letters. Additionally, handwriting practices and styles are not developed consistently within the same language; for example - cursive vs print calligraphy. As a result of globalization estimated 50% of world's population speaks second language [1]. Researching characteristics of non-native handwriting has been done by various educational and second language research purposes but remains largely unaddressed in the context of augmented cognition using online handwritten recognition. We researched differences and similarities of online handwriting between native and non-native speakers of English, Georgian, Chinese and Korean speakers. We have also examined related research for Arabic, Italian and Malay handwritings. As a result, we have identified key characteristics of non-native speakers' distinguishing from the native ones. In addition, we have identified differences based on writers' individual maturity with the second language.