big real-world networks like the Web, as there is no full record of their evolution. Wikipedia now allows us to witness, and validate, preferential attachment at work on its graph.

conclusion

The usefulness of an online encyclopedia depends on multiple factors, including breadth and depth of coverage, organization and retrieval interface, and trustworthiness of content. In Wikipedia more depth eventually translates into breadth, because the Wikipedia style guidelines recommend the splitting of overly long articles. The evolution of articles and links in Wikipedia allows us to model the system’s growth. Our finding that the ratio of incomplete vs. complete articles remains constant yields a picture of sustainable coverage and growth. An increasing ratio would result in thinner coverage and diminishing utility and a decreasing ratio of incomplete vs. complete articles to eventual stagnation of Wikipedia growth.

The idea of growth triggered by undefined references is supported by our second finding—that most new articles are created shortly after a corresponding reference to them is entered into the system. We also found that new articles are typically written by different authors from the ones behind the references to them. Therefore, the scalability of the endeavor is limited not by the capacity of individual contributors but by the total size of the contributor pool.

Wikipedia’s incremental-growth model, apart from providing an in-vivo validation of Barabási’s scale-free network-development theory, suggests that the processes we have discovered may continue to shape Wikipedia in the future. Wikipedia growth could be limited by invisible subjective boundaries related to the interests of its contributors. Our growth model suggests how these boundaries might be bridged. Consider that references to nonexistent entries prompt creation of these entries and assume that all human knowledge forms a fully connected network. Wikipedia’s coverage will broaden through a breadth-first graph traversal or flood-filling process, albeit over an uneven time progression.

it turns out
that the reality
of Wikipedia’s
development is
located comfortably
between the
two extremes
of nonexistent
link inflation and
deflation.

How far might the Wikipedia process carry us? In Jorge Luis Borges’s 1946 short story “On Exactitude in Science,” the wise men of the empire undertake to create a complete map of the empire; upon finishing, they realize the map was so big it coincided with the empire itself.

 

References

1. Barabási, A.-L. and Albert, R. Emergence of scaling in random networks. Science 286, 5439 (Oct. 15 1999), 509–512.

2. Bryant, S., Forte, A., and Bruckman, A. Becoming Wikipedian: Transformation of participation in a collaborative online encyclopedia. In Proceedings of the 2005 International ACM SIGGROUP Conference on Supporting Group Work (Sanibel Island, FL, Nov. 6–9). ACM Press, New York, 2005, 1–10.

3. Capocci, A. Servedio, V., Colaiori, F., Buriol, L., Donato, D., Leonardi, S., and Caldarelli, G. Preferential attachment in the growth of social networks: The case of Wikipedia. Physical Review E, 74, 036116 (2006).

4. Cunningham, W. and Leuf, B. The Wiki Way: Quick Collaboration on the Web. Addison-Wesley, Boston, MA, 2001.

5. Denning, P., Horning, J., Parnas, D., and Weinstein, L. Wikipedia risks. Commun. ACM 48, 12 (Dec. 2005), 152–152.

6. Garfield, E. Citation Indexing: Its Theory and Application in Science, Technology, and Humanities. John Wiley & Sons, Inc., New York, 1979.

7. Giles, J. Internet encyclopaedias go head to head. nature 438, 7070 (Dec. 15, 2005), 900–901.

8. Mehler, A. Text linkage in the wiki medium: A comparative study. In Proceedings of the EACL 2006 Workshop on new Text: Wikis and Blogs and Other Dynamic Text Sources (Trento, Italy, Apr. 6, 2006), 1–8.

9. Mitzenmacher, M. A brief history of generative models for power law and lognormal distributions. Internet Mathematics 1, 2 (2003), 226–251.

10. Remy, M. Wikipedia: The free encyclopedia. Online Information Review 26, 6 (2002), 434.

11. Stvilia, B., Twidale, M., Smith, L., and Gasser, L. Assessing information quality of a community–based encyclopedia. In Proceedings of the International Conference on Information Quality (Cambridge, MA, Nov. 4–6, 2005), 442–454.

12. Viégas, F., Wattenberg, M., and Dave, K. Studying cooperation and conflict between authors with history flow visualizations. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (Vienna, Austria, Apr. 24–29). ACM Press, New York, 2004, 575–582.

13. Voß, J. Measuring Wikipedia. In Proceedings of the 10^th International Conference of the International Society for Scientometrics and Informetrics (Stockholm, July 24–28, 2005), 221–231.

This work is partially funded by the European Commission’s Sixth Framework Programme under contract IST-2005-033331 “Software Quality Observatory for Open Source Software (SQO-OSS).”

Diomidis Spinellis ( [e-mail]) is an associate professor of information system technologies in the Department of Management Science and Technology at the Athens University of Economics and Business, Athens, Greece.

Panagiotis Louridas ( [e-mail]) is a software engineer in the Greek Research and Technology Network and a researcher at the Department of Management Science and Technology, Athens University of Economics and Business, Athens, Greece.