Paper short abstract:

JALEA (JApanese LEArning) is a new e-Learning system inspired by the hypermedia dictionary of Japanese grammar BunpoHyDict (Mariotti 2008). I will discuss some characteristic features of JALEA and demonstrate how JALEA can use authentic interconnected video and images for language learning.

Paper long abstract:

JALEA (JApanese LEArning) is a new e-Learning system inspired by the hypermedia dictionary of Japanese grammar BunpoHyDict (Mariotti 2008). In Japanese language pedagogy, BunpoHyDict approach is still very innovative for its focus on personal navigation and discovery as well as the integration of authentic, interconnected video and images. However, it raises some fundamental questions:

1. How can data related to learner navigation be gathered to improve the learner experience?

2. How can data be usable through different devices (e.g., pc, tablets, smartphones, etc.)

3. How can a project be maintainable and sustainable as it grows exponentially due to the constant updates of authentic materials?

JALEA project addresses the above three questions as follows:

In JALEA:

#1: Student data is gathered and analyzed by separating program code from data with a database oriented approach.

#2: To allow the usage of different devices, JALEA has been designed with a responsive technology, which allows content size and element position to change in accordance with the window or screen size.

#3: The JALEA application follows the principle of SOC (Separation of Concern) to allow maintainability and sustainability: it offers a layer accessible to the learner (frontend), and a layer accessible to the maintainers to add content and media (backend). Moreover, to allow for easier content management, several components have been automatized. For example:

• Automatic hyperlinks: the maintainer wraps the content with grammar-like tags that are automatically converted into hyperlinks.

• Automatic text parsing and kana->romaji auto conversion: the 'Mecab' morphological analyzer has been implemented side by side with a custom dictionary that permits users to control and correct the parsing process when needed.

• Text-to-voice automatic and hybrid conversion: the text is converted into voice but the operator can set parameters like character voice, intonation and pitch, and can native speech file uploaded.