Page MenuHomePhabricator

RFC: WikiVideos - Auto generation of videos from Wikimedia projects
Open, Needs TriagePublic

Description

Introduction:

The purpose of this RFC is to propose the development of a tool named "WikiVideos," a one-tap solution that automatically creates videos from Wikipedia and Wikimedia Commons content. This tool aims to cater to the growing trend of short video-based learning, providing users with a visually engaging way to consume information.

Vision:
Envision a tool where, at the tap of a word, WikiVideos seamlessly transforms textual information into dynamic visual content. By intelligently integrating data from Wikipedia, structured insights from Wikidata, and compelling multimedia from Wikimedia Commons, WikiVideos aims to offer users an accessible and engaging way to explore and learn about a wide range of topics, providing a rich and personalized learning experience.

Background:
During the Indic Wikimedia Hackathon event, the idea of WikiVideos was discussed, receiving positive feedback from the community. The tool will leverage Open AI technologies to extract information from Wikipedia, gather images and media files from Wikimedia Commons, and seamlessly stitch them together to create informative videos.

Objectives:

  • Automated Video Creation: Develop a tool that automates the process of creating videos by extracting information from Wikipedia articles.
  • Integration with Wikimedia Commons: Incorporate images and media files from Wikimedia Commons to enhance the visual appeal and richness of the videos.
  • User-Friendly Interface: Design an intuitive one-tap solution to make WikiVideos accessible and user-friendly for a broad audience.
  • Open AI Technologies: Leverage Open AI technologies to enhance the efficiency and effectiveness of content extraction and video creation.

Benefits:
Accessibility: WikiVideos will provide a more accessible and engaging way for users to consume information.
Community Collaboration: Encourage collaboration within the Wikimedia community by providing a new avenue for content creation and consumption.

Implementation
The implementation of WikiVideos will involve collaboration between developers, designers, and the Wikimedia community. The tool will be designed to respect Wikipedia's content guidelines and licensing requirements.

Feedback and Discussion
We would like to hear from the Wikimedia community to provide feedback, suggestions, and concerns regarding the WikiVideos project. Please share your thoughts on the proposed objectives, benefits, and any potential challenges that should be considered.

Conclusion
WikiVideos has the potential to revolutionize the way users engage with Wikimedia content, making it more accessible and appealing. Your input is crucial in shaping the future of this project.

Event Timeline

Gopavasanth renamed this task from Request for Comments (RFC) - WikiVideos to RFC: WikiVideos - Auto generation of videos from Wikimedia projects.Dec 17 2023, 10:12 AM

Given Video extension and general Video I'd recommend a name which includes an additional verb :)

Can this tool also include audio in it? For example on arywiki we have many articles that have audio recording, so we could generate a video with the recorded article, and some visual content related to it.

Can this tool also include audio in it? For example on arywiki we have many articles that have audio recording, so we could generate a video with the recorded article, and some visual content related to it.

This sounds like good Idea. But the problem is the video generated will be a summery of article. So audio might not align with video. Currently there is plan to use a speech to text service. Still we can try this thing also.

Can this tool also include audio in it? For example on arywiki we have many articles that have audio recording, so we could generate a video with the recorded article, and some visual content related to it.

This sounds like good Idea. But the problem is the video generated will be a summery of article. So audio might not align with video. Currently there is plan to use a speech to text service. Still we can try this thing also.

My idea is to allow the use of pre-recorded audio for the article, if available, not just automatically generated audio. On arywiki, we have more than 470 articles recorded by editors, and the audio is added at the top of the article (here's an example). Text to speech is unfortunately not available for ary (Moroccan Darija), because it's not a standardized language (no standard spelling of words).