The semantic web versus Wikipedia?

From the day of birth of the Wikipedia project, the online encyclopedia has been a highly controversial case. The debates range from information accuracy (including the inquiry of the information-knowledge relation once again) to the pyramidic usermodel (often undermined as too free or too open) or elsewhere as a fierce debate on the contributions coming from businesses, or just from innocent bots.

Less is spoken of the site’s hyperlink culture (or hypertext race), in which in the ultimate goal is to ‘sew together’ articles. A process of more and more overlap occur regarding the different subjects. In the ambition of finding the overlaps, there´s not always the same kind of accuracy in semantically connecting the pages from a users point of view. As a study of the MIT shows, people browse more effectively when the linked pages share a logical semantic structure. The weaker hypertexts work as a obstacle for effective in this case.

Meanwhile, in the more mainstream tendencies of Wikipedia, hacks and mods are slowly taking place. Although the site’s priority has always been the outkeeping of hackers, the platform of ‘play’ is obviously a small island in the seas of information for numbers of reasons. Therefore, finding a mod of the encyclopedia came as a suprise. In the Wikipedia Game visitors instantly participate in a browsing-contest. The concept is simple: get to a given entry starting from a given page. Players for example are given the startpoint of the entry about Long Island, where the goal is the entry about Arkansas. The competition-page puts the Wikipedia-page in a frameset, and counts everyone’s score. The websurf competition is up for 200 seconds before the page finally refreshes with new keywords. The participant who uses the least clicks to get the ‘trail’ right scores. Competitors get to see the hypertextual route as each match ends.

As modest as the netgame presents itself, it also projects characteristics on the side of the information portal as well on the media usage. One could argue the user, while competing, gets more aware of its searching (in)capabilities or realizes the temptations of ‘hypertext distraction’.

In the same way, a highly similar game was developed by the MIT as a cognition study. In Wikispeedia people were confronted with the same game setup as in the Wikipedia Game. Only here the results would be measured to get insight in the (supposedly increasing) browsing capabilities of the individual. The project was an alternative take on the MIT’s Open Mind Common Sense Project.

The researchers of Wikispeedia state in their research paper, called Wikispeedia: An Online Game for Inferring Semantic Distances between Concepts:

“The initial getting-away and the final homing-in are much more predictable after seeing game data than before, and the idea is to use the information gain to guess where the homing-in phase, and thus the relevant part of a single game path, starts.” (West et al., 2009)

This quantative study would provide data to optimize the web in a more semantic and therefore meaningful way. (In respect to the MIT paper: the project goals were never to develop semantic frameworks for the Wikipedia system, but to get a greater understanding of human cognition in a hypertextual environment.)

Although the gathering of statistical user information could possibly contribute to a greater understanding of the common sense aspect of webbrowsing, the case of Wikipedia could as well be a heavily controversial case to build upon.

In the sense of getting from A to B, users could be assisted with a certain context-specific algoritm. However, maybe the results of the Wikispeedia project would only become of practical use if someone governs this hypertext fettishism. Therefore understanding of the cognition of webbrowsing is only useful when the system and its users are literate enough to give up the interest of information completeness to usability. As the information is structured as it is it’s maybe more fair to reverse the questions into: do Wikispeedia and The Wikipedia Game contribute to user literacy over Wikipedia? And in a broader perspective: what would be the consequences be if a information network like Wikipedia becomes focused mainly on the semantic research? Would information get ‘hidden’ as it’s only presented when its algoritms predict so? Most likely, the fields of semantic web-browsing with emphasizing the user-experience would clash with the stubborn dogma’s within of the (or any) web-based encyclopedia. The tensions in the editing have already exposed in the hypertextual warfront as we know it (in which completeness obviously wins over usability).

The final research paper would elaborate more on this semantic web in relation to the ideals of a web encyclopedia. In this comparison of interests, also a reflection upon the current situation of Wikipedia could be extracted. (I any case, this is not a disqualification of Wikispeedia project, rather it is a extrapolation of the projects ambitions upon the Wikipedia context.)