It's kind of like GraphJam, except less amusing . . .
You may or may not care about my Fanlore study, but I spent approximately seven hours on one graph yesterday and it is not really useful for much of anything so I am posting it here. For posterity. Or something.
Actually, I'll post all three of my lovely graphs. Ha.
This one was pretty easy (thanks to a fortunate run-in at the lab with the other student who's studying a wiki-based community and who gave me the secret Excel formulas for extracting data from wiki history pages). For my study, I'm most interested in boundary crossing, which for Fanlore means people who edit across multiple fandoms (do they participate in multiple fandoms or do they just edit for the heck of it?).
So this graph shows a glimpse of three pages and the people who've edited them, with ties to indicate people who edit more than one page. The pages are the overview pages for D.Gray-man (D.GM), The Sentinel (TS), and due South (dS, which got kind of hidden by the lines, but it's the one in the lower left). The thicker lines indicate more edits. The three people labeled in pink are on the wiki committee. Please note my beautiful layout. You have to drag the nodes around by hand in order to get a readable layout. The professor and some classmates were impressed by my spaciousness.
The next graph focuses on one of the most active contributors. It just shows what types of pages she contributes to, with thicker lines again indicating more edits. Since it just breaks the pages into very basic category types, there's not a lot and therefore it was easy-peasy to rearrange readably.
And then there's the seven-hour, utterly useless graph from hell. Same user as above, but a more detailed and focused snapshot of her activity. It's the same information as the graph above, but I broke down the articles into fandoms they relate to and limited the timeframe to February 2009, and the thicker lines mean more edits to a page (or in this instance series of pages related to a fandom).
See, I was hoping to be able to look at where people contribute versus the fandoms they list on their profile page, to see instances of boundary crossing like I mentioned above. So the fandoms labeled in red are the ones she lists on her profile. Unfortunately, there is no magic formula to parse the dataset of wiki contributions automatically, so I interpreted 467 lines of data manually and THEN had to drag all those bloody nodes around into readable format (including sorting them vaguely by genre/category because I can't just leave them jumbled). *headdesk* There's no way I'm doing that for multiple users, so there's no way for me to make comparisons and therefore no value to this graph. Except it would be so cool if I could capture this data automatically because that is exactly what I would like to study at Fanlore. *sigh*
Now the professor wants me to make a network graph of the top ten or so contributors, which also involves a lot of manual data interpretation. Oh boy, more headaches!
Actually, I'll post all three of my lovely graphs. Ha.
This one was pretty easy (thanks to a fortunate run-in at the lab with the other student who's studying a wiki-based community and who gave me the secret Excel formulas for extracting data from wiki history pages). For my study, I'm most interested in boundary crossing, which for Fanlore means people who edit across multiple fandoms (do they participate in multiple fandoms or do they just edit for the heck of it?).
So this graph shows a glimpse of three pages and the people who've edited them, with ties to indicate people who edit more than one page. The pages are the overview pages for D.Gray-man (D.GM), The Sentinel (TS), and due South (dS, which got kind of hidden by the lines, but it's the one in the lower left). The thicker lines indicate more edits. The three people labeled in pink are on the wiki committee. Please note my beautiful layout. You have to drag the nodes around by hand in order to get a readable layout. The professor and some classmates were impressed by my spaciousness.
The next graph focuses on one of the most active contributors. It just shows what types of pages she contributes to, with thicker lines again indicating more edits. Since it just breaks the pages into very basic category types, there's not a lot and therefore it was easy-peasy to rearrange readably.
And then there's the seven-hour, utterly useless graph from hell. Same user as above, but a more detailed and focused snapshot of her activity. It's the same information as the graph above, but I broke down the articles into fandoms they relate to and limited the timeframe to February 2009, and the thicker lines mean more edits to a page (or in this instance series of pages related to a fandom).
See, I was hoping to be able to look at where people contribute versus the fandoms they list on their profile page, to see instances of boundary crossing like I mentioned above. So the fandoms labeled in red are the ones she lists on her profile. Unfortunately, there is no magic formula to parse the dataset of wiki contributions automatically, so I interpreted 467 lines of data manually and THEN had to drag all those bloody nodes around into readable format (including sorting them vaguely by genre/category because I can't just leave them jumbled). *headdesk* There's no way I'm doing that for multiple users, so there's no way for me to make comparisons and therefore no value to this graph. Except it would be so cool if I could capture this data automatically because that is exactly what I would like to study at Fanlore. *sigh*
Now the professor wants me to make a network graph of the top ten or so contributors, which also involves a lot of manual data interpretation. Oh boy, more headaches!