Towards Context-Aware Automated Writing Evaluation Systems

Writing is a crucial skill in our society, which is regularly exerted by students across all disciplines. Automated essay scoring and automatic writing evaluation systems can support professors in the evaluation of written texts and, conversely, help students improving their writing. However, most of those systems fail to consider the context of the writing, such as the targeted audience and the genre. In this paper, we depict our vision towards new-generation AES systems that could evaluate written products while considering their specific context. In education, such tools could support students not only in adapting their written product to their particular context, but also in identifying points for improvement and situational settings where their writing is less proficient.