RecommendationAssistant is an open source Nextcloud app that is not finished yet. The app is the result of my Master’s Thesis (read more here).
The result of my thesis was an app consisting of two components: Collaborative Filtering and Content-Based Recommendation. Due to some limitations and ineffectiveness of libraries for parsing file content, I decided to remove Content-Based Recommendation from the app. There are different recommendation techniques that can be used instead. But I have nothing decided yet.
Features:
- File recommendations based on Collaborative Filtering
- Displaying recommendations on the Nextcloud main page
What does RecommendationAssistant mean for me?
I am of the opinion that RecommendationAssistant is my best software product so far. Therefore I will work on the app whenever I have the opportunity.
Translations:
The app is available in english language and can be translated to other languages.
Limitations:
There are very few possibilities to test the app. I need a lot of data (files) from users to make a trustworthy statement about the results of the algorithms used.
Since I don’t have this possibility at the moment, I test with the MovieLens data. You can learn more about MovieLens here.
Support:
I am happy to help you in case of any problems. Just send me an email or open a GitHub issue.
License:
RecommendationAssistant is licensed under AGPLv3.
Contribution:
You can contribute to the app by forking the project on GitHub and creating a pull request. Send me an email if you do not know how!
Installation:
Search for “RecommendationAssistant” on the Nextcloud Appstore or install the app manually. Download the app from GitHub and unzip the folder. Move the app in the “apps” folder of your Nextcloud. Make sure that you have read and write permissions on the folder. The app is now available at your “Apps” page in your Nextcloud’s admin section. Enable the app and enjoy 🙂
Documentation:
The theory behind the app is documented in my Master’s Thesis. Feel free to contact me if you need further information.
Want to donate?
Simply use the following button:
