Visualizing authorship and contribution of collaborative writing in e-learning environments
summary
A system reconstructs versioned documents via a Wikipedia-style revision model, attributes tokens to authors, and combines quantitative counts with semantic FastText embeddings to detect topics and reflective writing; a WCM (weighted contribution measurement) ranks contributors. Evaluation on a small Google Docs dataset shows about 70% contribute 10% of text and yields small improvements over pure quantitative metrics. Limitations: concurrent-edit API gaps, only text analyzed, small dataset; future work validates section structure and runs user tests.
AI-written summary · about
chapters (8)
- 0:00Introduction to research
- 0:47Motivation for collaborative writing analysis
- 2:14Related work on authorship metrics
- 3:37System architecture and data processing
- 5:06Qualitative metrics and analysis methods
- 6:57Experiments and dataset overview
- 8:27Limitations and future work
- 9:53Conclusion and Q&A
related talks
31mEstimating stock price correlations using Wikipedia
30mBuilding a content-focused, scientific document authoring workflow
53mPounding Simplicity into Wiki
13mThe Company You Keep: Mapping Scientific Collaboration with Graphs
39mTemporal Analysis on Topics Utilizing Word2Vec
19mHow to look at your data
25mCombining Collaborative & Content Filtering in a Recommendation System for a Web-based DAW
36mEstimating stock price correlations using Wikipedia