|
|
Test |
|
|
\ No newline at end of file |
|
|
# Context
|
|
|
Today, the worldwide needs in terms of education are growing. Naturally, depending on the language, the culture, and the domain, the modalities of providing this education varies a lot. In the context of the European project [X5GON](https://www.x5gon.org/), we are interested in using tools from artificial intelligence in order to cross the aforementioned differences, and provide open educational resources to any learner. To this aim, we can work with educational data, coming from various repositories. The versatility of such data requires to develop adaptive algorithms, able to cope with multiple modalities. One important objective is to be able to represent the educational content in a way users may easily catch. One way of doing so is to provide summaries of the content.
|
|
|
|
|
|
Using recent advances in sequence to sequence models, we are now able to develop interesting text summarization models, that can be used in real applications. Though the users can directly assess the goodness of summarization if they are already aware of the content, it may be difficult to automatically quantify the quality of the results for a large number of ressources. Interestingly The quality of summarization can also be evaluated through the use of standard metrics, such as Recall-Oriented Understudy for Gisting Evaluation (ROUGE), that include a vast family of variants [2].
|
|
|
|
|
|
# This repository
|
|
|
This repository contains source codes for the ideas presented in our [report](). It is a platform that unifies state-of-the-art text summarisation tools, slightly modified to adapt to our specific needs.
|
|
|
|
|
|
This wiki will guide you through the setup and usage of the tools.
|
|
|
1. [Setup](https://gitlab.univ-nantes.fr/E193437H/text-summarisation-yale/-/wikis/Setup)
|
|
|
2. [Pointer-generator](https://gitlab.univ-nantes.fr/E193437H/text-summarisation-yale/-/wikis/Pointer-Generator-(pg))
|
|
|
3. [Unified-summarization](https://gitlab.univ-nantes.fr/E193437H/text-summarisation-yale/-/wikis/Unified-summarisation)
|
|
|
4. [Evaluation]() |
|
|
\ No newline at end of file |