l
o o m 2
|
||||||||||||||||
w
o r k i n p r
o g r e s s
|
||||||||||||||||
professor
judith donath . danah boyd . hyun-yeul lee . dan ramage
w/ support from scott golder . ming-en cho . jonathan goler |
||||||||||||||||
|
||||||||||||||||
|
||||||||||||||||
Understanding activity. |
||||||||||||||||
Given the large amount of data that we have available, it's important to figure out what the information means. This series explores the data that we have relative to "activity." Based on an average individual's quantity of posts, one's activity is the magnitude one's posts compared to the average (i.e. an "activity" of 5 means that one posts 5x the amount of the average). A PDF file, which describes the details of this work, is available here. |
||||||||||||||||
Orphans. This chart compares one's activity to their orphanage rate (or, how often no one responds to their thread-starting posts). Notice that in this aggregate graph, active participants tend to fall into two main categories: those with relatively high and those with very low orphanage rates. One possibility is that high orphan rates may be attributed to automated spam-generating ad-bots. | ||||||||||||||||
Responses per Person. This chart compare's one's activity to the average number of responses s/he receives to each post. Interestingly, nearly all active users tend to average about 1 to 2 responses per post (x asymptote), and it's only the relatively inactive posters that never elicit replies. | ||||||||||||||||
Anger. To determine anger, we quantify what it means to use "angry" language, expressions and punctuation. Again, we compare one's level of angry language to one's activity. Notice that the bounds on the distribution of Anger is less strictly defined than the bounds on the distribution of Responses per Person. (there are even some notable outliers near Activity 50 and 60). This seems to indicate that angriness is a posting trait that does not have as significant effect on the poster's activity. |
||||||||||||||||
Anger in alt.flame. As alt.flame rewards angry language, particularly in the form of "flames," we thought it would be interesting to see how one's activity correlated with one's level of anger. This graph's distribution is reminiscent of the of the overall Activity vs Anger aggregate graph. The most active posters in alt.flame are have only a slightly greater Anger value than do the most active posters overall. One interpretation of this fact is that an individual poster's level of anger is somewhat location independent - perhaps angry posters are just as angry in alt.flame as they are when posting to other newsgroups. | ||||||||||||||||