New Project Makes Wikipedia's Data More Accessible to AI

An innovative Wikimedia project makes Wikidata more accessible to artificial intelligence, encouraging open source use.

New Project Makes Wikipedia's Data More Accessible to AI
Illustration of the Wikidata Embedding project for open access to AI data

An Initiative to Simplify Access to Open Source Data for Artificial Intelligence

Wikimedia Deutschland has recently launched a new project aimed at making Wikidata, Wikimedia's vast open knowledge platform, easier to access for AI applications. The project seeks to democratize the use of high-quality data, allowing even small initiatives to use this resource without the financial or human resources of large companies.

Goals and Technologies Involved

The core of the initiative is turning Wikidata into semantic vectors, enabling more efficient searches and integration into advanced AI systems such as those based on retrieval-augmented generation (RAG). DataStax, a vector database provider, and Jina AI, which contributes open source models for vectorizing text data, are collaborating on the project. This method helps reduce AI errors, improve the accuracy of answers and make it easier to detect vandalism or harmful edits.

Impact and Values of the Initiative

The goal is to use Wikidata as a reliable base of shared knowledge, countering the spread of false information generated by AI systems. The collaboration is based on the principles of transparency, free access and participatory community, respecting privacy policies and human rights. Data security and quality are priorities, fostering a fairer and more reliable AI ecosystem, including for open source projects.

Future Implications and Community Involvement

Beta tests are planned for 2025, along with public presentations to promote adoption of this infrastructure. The initiative reinforces the idea that AI can be an open and collaborative alternative, contributing to a more respectful and sustainable information ecosystem with a positive impact on society and on democratic access to knowledge.

Conclusions

  • The project simplifies access to Wikidata through vectorization, improving search and reliability.
  • It supports the development of APIs and AI systems that are more precise and transparent.
  • It supports a commitment to an open and inclusive digital future.