The starting point in each case is a Twitter list in a particular topic area. The list I’m going to use in this example is @jobrodie’s scicomm list, a list of science communicators on Twitter.

The first recipe-in-progress tries to identify other lists in the area… It works as follows:

- for each person on the original list (candidate members), find the other lists they are mentioned on to give a set of candidate lists;
- count the number of candidate members (members of the original list) on each candidate list.

We can now display lists containing N or more people on the original list:

/JoBrodie/scicomm Candidate members: 61 Subscribers: 9
/ebsuckling/science Candidate members: 23 Subscribers: 14
/lisushi/science-world Candidate members: 22 Subscribers: 6
/Frazerbw/science Candidate members: 22 Subscribers: 2
/efficientva/science Candidate members: 22 Subscribers: 7
/cptyler/science Candidate members: 20 Subscribers: 15
/LouWoodley/london-sci-comm Candidate members: 20 Subscribers: 7
/ousefulAPI/solo10 Candidate members: 18 Subscribers: 1
/mocost/london Candidate members: 17 Subscribers: 25
/girlinterruptin/science-science-policy Candidate members: 16 Subscribers: 2
/TwistedBacteria/science-comm-edu Candidate members: 16 Subscribers: 35
/utne_altwire/science2 Candidate members: 16 Subscribers: 3
/physicus/scicom Candidate members: 15 Subscribers: 4
/TheUrbanOctopus/science Candidate members: 15 Subscribers: 7

Or we can order the lists with more than P subscribers by subscriber count:

/sciencebase/scientwitters 11 434
/sciencebase/sci-comms 14 142
/tlists/science-995 9 77
/teobesta/science-part-1 14 38
/TwistedBacteria/science-comm-edu 16 35
/mjrobbins/women-science-tweeps-3 9 28
/mocost/london 17 25
/emble64/science-chat 13 25
/NicolasLoubet/science-tech-world 9 23
/nemski/science 9 22
/LouWoodley/solo10-attendees 12 20
/kristinalford/sciencecomms 12 20
/OpheliaBottom/journos-and-scientists 9 20
/pfanderson/science-research 12 20
/sc_k/bloggers-writers-journali 11 19
/AndreiaASoares/sci-comm 10 19
/soloconf/solo11-attendees 10 18
/cptyler/science 20 15

Or we could order lists according to some function of both subscriber count and number of list members who are also on the original list…

The second recipe attempts to identify notable individuals in a topic area based on a set of lists in the topic area. Given a list of lists (candidate lists), such as either of the above, we can then find the members of each list, and count the number of candidate lists they appear on…

['alicebell', 17, 734, 5353, 11534]
['edyong209', 16, 735, 10504, 27207]
['skyponderer', 15, 1359, 2544, 7664]
['JoBrodie', 14, 969, 1794, 25889]
['ProfBrianCox', 14, 94, 390506, 5345]
['SciencePunk', 14, 148, 5066, 8728]
['imascientist', 14, 806, 3396, 12833]
['alokjha', 14, 880, 5123, 3774]
['sciencebase', 14, 1776, 11431, 9092]
['RogerHighfield', 14, 432, 5674, 10770]

Here, then, we have view over people who appear to be heavily represented in lists in a particular subject area, where those lists are identified by means of similarity with an initial list.

Another approach to discovering notable individuals in an area would be to take all the members of a set of candidate lists and plot their social connections, both within members of set of set of candidate lists (i.e. look to see how the members of a set of candidate lists follow each other), as well as across their whole friends lists (i.e. plot the connections between members of the candidate lists and all their friends, whether or not they appear on one or more of the candidiate lists).

PS Rather than seeding the recipe with an initially prepared list, we might also seed it with a search around a particular search term. So for example, it’s possible to generate an initial set of candidate members from a simple Twitter search. A search on a given keyword identifies a set if users who have tweeted using that keyword, which we can then use to identify a set of lists in the area and a set of users who appear across those lists. Searching on the keyword “education” turns up the following people spread across lists that mentioned people who had recently tweeted the keyword “education”:

'JobsDirectUSA', 'coolcatteacher', 'AmericaHires','educationweek', 'edutopia', 'DellEDU', 'web20classroom', 'elearnnet', 'C4LPT', 'UPSjobs'

(I suspect the original search turned up users who were represented on job related lists!)

It’s not brilliant, but it’s a start…

Let’s take @jiscCetis as an example; this account follows no-one, is followed by a few, hasnlt much of a tweet history and is listed by a handful of others.

Here’s the follower network, based on how the followers of @jiscetis follow each other:

There are three (maybe four) clusters there, plus all the folk who don’t follow any of the @jisccetis’ followers…: do these follower clusters make any sort of sense I wonder? (How would we label them…?)

The next thing I thought to do was look at the people who were on the same lists as @jisccetis, and get an overview of the territory that @jisccetis inhabits by virtue of shared list membership.

Here’s a quick view over the folk on lists that @jisccetis is a member of. The nodes are users named on the lists that @jisccetis is named on, the edges are undirected and join indivduals who are on the same list.

Plotting “co-membership” edges is hugely expensive in terms of upping the edge count that has to be rendered, but we can use a directed bipartite graph to render the same information (and arguably even more information); here, there are two sorts of nodes: lists, and the memvers of lists. Edges go from members to listnames (I should swap this direction really to make more sense of authority/hub metrics…?)

Another thing I thought I’d explore is the structure of the co-list membership community. That is, for all the people on the lists that @jisccetis is a member of, how do those users follow each other?

It may be interesting to explore in a formal way the extent to which the community groups that appear to arise from the friending relationships are reflected (or not) by the make up of the lists?

It would probably also be worth trying to label the follower group – are there “meaningful” (to @jisccetis? to the @jisccetis community?) clusters in there? How would you label the different colour groupings? (Let me know in the comments…;-)