Research:AI-Assisted Wikipedia Updating
This page documents a research project in progress.
Information may be incomplete and change as the project progresses.
Please contact the project lead before formally citing or reusing results from this page.
The Blender Lab at University of Illinois Urbana-Champaign (UIUC) is developing a research tool to help Wikipedia editors identify and add information across articles.
We are recruiting ~20 experienced Wikipedia editors to test our tool. Each participant will evaluate suggested edits on approximately 50 articles for accuracy and provide feedback on the system’s functionality, user interface, and overall effectiveness.
- - - Are you interested? Let us know here! - - -
We will provide automatically generated suggestions to help you edit Wikipedia Articles (more details in the How Does It Work? section). Currently, this is a research prototype; however, we aim to refine this free tool so it may help and support the broader Wikipedia community!
About Blender Lab at UIUC
[edit]The Blender Lab is a research lab at the University of Illinois Urbana-Champaign. Its previous research projects include: SmartBook and Claim-Radar You can learn more at Blender Lab - Publications.
Project Overview
[edit]Why AI-Assisted Wikipedia Updating?
[edit]Wikipedia, one of the world's largest repositories of human knowledge, serves as the primary reference source for millions globally. Despite its scale and success, its community-driven model faces challenges in keeping all content current and comprehensive. Therefore, we aim to help Wikipedia human editors to make (1) timely, (2) efficient, and (3) objective edits to articles.
Take a look at our sample system suggestions for (1) Taylor Swift and (2) Joe Biden.
How Does It Work?
[edit]Knowledge Extraction from Wikipedia: Our system analyzes Wikipedia edit histories to identify meaningful knowledge updates. Using advanced classification algorithms, we distinguish substantive content additions from routine maintenance edits like style changes or reverted content. The system creates a structured knowledge base where each entry contains the target entity (person, organization, or event), new information added, and the precise timestamp of the update.
Real-time News Monitoring and Update Generation: The system continuously monitors news coverage related to entities in our knowledge base. When it detects new, Wikipedia-relevant information, it identifies gaps between current Wikipedia content and new developments, generates properly formatted edit suggestions, includes edit summaries and citations to reliable sources, and submits proposed changes for human editor review.
Review and Approval: All suggested edits undergo mandatory review by Wikipedia editors, who maintain full control over accepting or rejecting changes. This human oversight ensures all updates meet Wikipedia's quality standards and editorial guidelines.
Timeline
[edit]We are planning to work with editors within the month of February.
- - - Are you interested? Let us know here! - - -
Have any questions?
[edit]Feel free to reach out to us by emailing revanth3 [at] illinois [dot] edu.