ThreadMill 0.1: Social Accounting for Message Thread Collections

The Social Media Research Foundation is pleased to announce the immediate availability of ThreadMill 0.1.  ThreadMill is a free and open application that consumes message thread data and produces reports about each author, thread, forum, and board along with visualizations of the patterns of connection and activity.  ThreadMill is written in Ruby, and depends on MongoDB, SinatraRB, HAML, and Flash to collect, analyze, and report data about collections of conversation threads.

Threaded conversations are a major form of social media.  Message boards, email and email lists, twitter, blog comments, text messages, and discussion forums are all social media systems built around the message thread data structure.  As messages are exchanged through these systems, some messages are sent as a reply to a particular previous message.  As messages are sent in reply to prior messages, chains of messages form.  Message chains come in two major forms: branching and non-branching.  Branching threads are those that allow more than one message to reply to a prior message.  Non-branching threads are single chains, like a string of pearls, that allow only one message to reply to a prior message.  Many web based message boards are non-branching.  Many email systems and discussion forums are branching.

ThreadMill requires a minimal set of data elements to generate its reports.  A data table must minimally have a column of information for each message that includes the name of the message board, the forum, the thread, and the author, along with a unique identifier for each message and the date and time it was posted.  Optional data elements include the unique identifier of the message being replied to, the URL of the message, and the URL for a profile photo.

All forms of threaded message exchange can be measured.  Simple measures like the count of the number of messages or the number of authors are obvious and useful.  Other measures can be created from more sophisticated analysis.  For example, the network of connections that forms as different authors reply to one another can be extracted and analyzed using network analysis methods.  It is possible to calculate metrics from these networks of reply that describe the location of each person in the graph.

ThreadMill generates several data sets that can be used to create visualizations of the activity and structure of a message board collection.

A Treemap data set can illustrate the hierarchy of encapsulated authors within threads, threads within fora, fora within boards, and boards within collections.  Treemap visualizations of collections of threaded conversations can quickly highlight the most active or populous discussions.

An AuthorLine visualization takes the form of a double histogram, with bubbles representing each thread active in each time period sized by the volume of messages the author contributed, sorted by size.  Threads that have been initiated by the author are represented as bubbles above the center line.  Messages that the author contributes to threads started by other authors are represented as bubbles stacked below the center line.  AuthorLines quickly reveal the pattern of activity an author displays and identifies which of several types of contributors the author is.

A scatter plot visualization represents each author as a bubble in an X-Y space defined by the number of different days the author was active against the average number of messages the author contributes to the threads in which they participate.

A time series line chart reveals the days of maximum and minimum activity along with trends.

A network diagram reveals the overall structure of the discussion space and the people who occupy strategic locations within the network graph.

ThreadMill has received generous assistance from Morningside Analytics.  Bruce Woodson implemented ThreadMill.

ICWSM 2010 Liveblog, Day 3

Fourth International AAAI Conference on Weblogs and Social Media (ICWSM-10)

Michael Kearns Keynote

Experiments: Graph Coloring / Consensus / Voting

Topology of the Network vs. what was the network used for?

Voting experiments – similar to consensus, with a crucial strategic difference.

Introduce a tension between:

-Individual preferences

-Collective unity

-Color choices; challenge comes from competing incentives

Red, blue. People unaware of global network structure

Payoffs: if everyone picks same color w/in 2 minutes, experiment ends, and everyone gets some payoff. But different players have different incentives (e.g. I may get paid p if everyone converges to blue, but 2p if everyone converges to red). If there is no consensus, nobody gets a payoff

Continue reading

ICWSM 2010 Liveblog, Day 2

Fourth International AAAI Conference on Weblogs and Social Media (ICWSM-10)

***Microblogging 2***

Predicting Elections with Twitter: What 140 Characters Reveal about Political Sentiment (Tumasjan et al.)

Successful use of social media in las presidential campaign has established twitter as an integral part of political campaign toolbox

Goal: analyze on Twitter: 1. Deliberation, 2. Sentiment, 3. Prediction

Previous work:

Deliberation: Honeycutt and Herring – Twitter not only used for one-way comm, but 31% of all tweets direct a specific addressee. Kroop and Jansen – political internet discussion boards dominated by small # of heavy users

Sentiment: How accurately can Twitter inform us about the electorate’s political sentiment?

Prediction: can Twitter serve as a predictor of the election result?

Data: examined more than 100k tweets and extracted their sentiment using LIWC

Target: German federal election 2009

Results:

1. While Twitter is used as a forum for political deliberation on substantive issues, this forum is dominated by heavy users

Two widely accepted indicators of blog-based deliberation:

-The exchange of substantive issues (31% of all messages contain “@”),

-Equality of participaion: While the distribution of users across groups is almost identical with the one found on internet message boards, we find even less equality of participation for the political debate on Twitter. Additional analyses have shown users to exhibit a party-bias in the volume and sentiment of messages.

2. The online sentiment in tweets reflects nuanced offline differences between the politicians in our sample.

LIWC profiles:

-Leading candidates: Very similar profile for all leading candidates, only polarizing political characters, such as liberal leader and socialist, deviate in line with their roles as opposition leaders. Messages mentioning Steinmeir (coalition leader) are most tentative

3. Similarity of profiles is a plausible reflection of the political proximity between the parties

Key findings: high convergence of leading candidates, more divergence among politicians of governin grand coalition than among those of a potential right wing coalition

4. Activity on Twitter prior to election seems to validly reflect the election outcome (MAE 1.65%), and joint party mentions accurately reflect the political ties between parties.

From Tweets to Polls: Linking Text Sentiment to Public Opinion Time Series (Brendan O’Connor)

Continue reading

ICWSM 2010 Liveblog, Day 1

Fourth International AAAI Conference on Weblogs and Social Media (ICWSM-10)

We will be liveblogging (when possible) from ICWSM 2010, going on now!

Keynote: Bob Kraut, CMU

ICWSM 2010 - Bob Kraut
implications for community design
-offline theories of socialization helpful, not definitive
-online communities can build in good socialization practice
-e.g. WP welcoming committee
Two Types of Commitments to Groups
-identity based groups
-bond based groups
Added Identity & Bond Features to MovieLens
Introduced Subgroups into MovieLens
Identity features that focus on subgroups
Individual profiles
bond-based design:+11% logins
identity-based design:+44% logins