Towards Facilitating the Accessibility of Web 2.0 Texts through Text Normalisation

The Web 2.0, through its different platforms, such as blogs, social networks, microblogs, or forums allows users to freely write content on the Internet, with the purpose to provide, share and use information. However, the non-standard features of the language used in Web 2.0 publications can make social media content less accessible than traditional texts. For this reason we propose TENOR, a multilingual lexical approach for normalising Web 2.0 texts. Given a noisy sentence either in Spanish or English, our aim is to transform it into its canonical form, so that it can be easily understood by any person or text simplification tools. Our experimental results show that TENOR is an adequate tool for this task, facilitating text simplification with current NLP tools when required and also making Web 2.0 texts more accessible to people unfamiliar with these text types.