Loading...
Please wait, while we are loading the content...
Visualizing Human Concept Linking Through Wikipedia
| Content Provider | Semantic Scholar |
|---|---|
| Author | Bishara, Alex |
| Copyright Year | 2012 |
| Abstract | Wikipedia is a unique information network of all collective human knowledge together with the relationships between different entities and concepts. Articles range a variety of subjects and are usually contributed to by an active community knowledgeable on that subject. An article’s text can contain inline hyper links to other articles if an editor deems the conceptual connection strong enough to warrant explicit annotation. Previous work in [1] and [2] has made the observation that articles exhibit a sort of small world phenomenon in which one can navigate between two conceptually distant topics just by following inline page hyperlinks in relatively few hops. This work is motivated by trying to understand how the Wikipedia link graph became able to be navigated this way. By examining the Wikipedia hyperlink graph evolution, I hope to better understand the criteria used by editors for deciding if a conceptual relationship exists between two concepts worthy of addition; in effect capturing some escence of human thought processes. My early comparisons between articles created in different generations of Wikipedia’s lifetime show far heavier linking activity throughout from older articles. By examining context aware evolution, noticing which topics are linking to each other and when, seemingly intuitive hypotheses for this initial observation do not seem to be the best explanation. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://snap.stanford.edu/class/cs224w-2012/projects/cs224w-038-final.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |