Getting a Grip on Data in Motion to Understand How We Collaborate

Recently I had a great discussion with my friends from Cisco about how to use dynamic social network analysis - initially of e-mail archives, but now extended to all types of communication archives - to improve collaboration in organizations. In particular, we use the virtual mirroring process, which I described in previous blog posts.

Analyzing the Communication Network of the 2012 COINs Seminar

In this seminar about 50 students from the five universities MIT/SCAD/Aalto University/ University of Cologne/ University of Bamberg worked together for five months in multinational virtual project teams as COINs (Collaborative Innovation Networks). They formed ten teams ranging in size from three to six students from at least two locations, working on a project related to social media and social network analysis (2012 course Web site). Students were asked to cc all their project-related e-mail communication to a dummy mailbox. This allowed us to construct a virtual mirror of ongoing communication within and between teams. At the end of the course, each team presented their results to their classmates in a virtual meeting. Each of the ten presentations was ranked by the students in the three categories “presentation quality, content quality, and creativity”. Comparing the virtual mirror of communication with the peer and instructor ratings permitted us to identify the communication patterns leading to the most highly-ranked work output.

Qualitative Analysis
Looking at the group network below, the different teams can be clearly recognized. The communication of each team is shown in a different color, usually team members are clustered together as a COIN, with external collaborators and other students being in more peripheral positions.

Analyzing the contribution index between team members shows that members of the same team tend to show similar behavior regarding the ratio of e-mails sent to e-mails received. Clusters of dots of the same color are members of the same group, overall we find that higher-performing teams tend to communicate more actively, with more similar send/receive ratios.
The temporal social surface indicates creativity, as there is a relatively large group of high-betweenness class members which is constantly changing over time, in earlier work we found this to be a reliable predictor of creativity.

The 6 snapshots of the 10 teams’ communication networks over the 5 months show Tuckman’s four phases in the life of a team: forming, storming, norming, and performing. We see how the main instructor in the uppermost picture at right is most centrally, but how then teams start connecting in the middle row, and how they then huddle together team-by-team to focus on their work in the bottom-most pictures.

The group betweenness centrality curve as well as the absolute number of messages sent and received shown below illustrate the higher traffic in the forming, storming, and norming phase, followed by the lower traffic in the second performing phase.

The sentiment curve illustrates the same phenomenon, with higher emotionality (defined as the sum of positivity and negativity) in the forming and storming phase in the first half of the course. X-axis is always days in these pictures.


Quantitative analysis
In the second half of this analysis we investigate what communication patterns will be indicative of high-quality work. The first pattern is “oscillations in betweenness centrality curves”.

The above picture illustrates the team ranked most creative by the instructors (blue shaded lines, each line titled “Series X” is one actor's betweenness over 115 days), and the team ranked least creative (red shaded lines). As can be easily seen, the centrality of most actors in the low-ranked team hovers around the zero-lines: these actors will be peripheral in the social e-mail network shown in the first picture of this document.

As the correlations below illustrate, the instructor rating of creativity (each of the instructors at each of the five participating locations ranked the 10 presentations) correlates highly (0.83**) with oscillation in betweenness centrality. As the picture above shows, the team rated the most creative had 80 oscillations, i.e. handovers in leadership, compared to the lowest rated team with less than 40 oscillations.

A similar correlation was identified for speed of response. The faster a team’s members communicated with the lead instructor, the higher the team’s work output was rated by the other students.

The correlation between the peer rating of a team’s content with its communication balance with the main instructor (Peter) is -0.719**, i.e. the higher the communication balance, the more communication with Peter, the better is the team’s content.

The more emotional the language of a team in the e-mails they exchange, the higher is the content rated by its peers.

The positivity in the e-mails sent to the main instructor is also highly predictive of high ratings in all criteria that have been rated (Presentation, Content, Creativity).


The conclusions for high-functioning teams are therefore:
1. Pass the baton frequently: the more leadership rotates among team members, the more creative the output will be
2. Communicate rapidly with the instructor, and among team members
3. Use emotional language: praise when praise is due, but also say when something is not ok.

Swarm Creativity to Explore Outer Space

Yesterday I learned from Rod Dunne about his Science Fiction novel "Terra Swarm" which extrapolates how swarms of scouting spacecraft could search the cosmos in search of habitable planets, forming hives as they go. COIN's form the backdrop for the concepts for a story about deep space explorers and their feelings of liberty within such a swarm.

A great read...

Cool COIN Movie from Chile!

When I was in Chile in August 2012 at the invitation of good friend Cristobal Garcia, "Do Future Andes 2012 " a great movie about the future of Swarmcreativity and COINs in Chile was produced. Besides the discussion of the benefits of creative swarms I also find the Chilean nature stunningly beautiful.

Last Minute Election Analysis in Ohio and Florida

As everybody knows, today the  next President of the United States gets elected. Common wisdom says that it's the swing states such as Florida and Ohio that will decide the outcome of the race. Rohan Kulkarni, who has already done a really nice job analyzing the outcome of the Vice Presidential debate between Ryan and Biden, has looked more closely at what voters in Ohio and Florida have to say about Obama and Romney. Here is his blog post.

Joe Biden and Paul Ryan: Pre and Post Debate

I would like to start by thanking Peter for inviting me to write this post on the Swarm Creativity blog. I have been in Peter’s COINs class for over a month now and I am thoroughly fascinated by what social network analysis can teach us.

For the COINs class’ midterm exam that involves gathering, analyzing and inferring from social media data, I was assigned the topic “Joe Biden against Paul Ryan”. It was a great topic for me since I was planning on doing some coolhunting on this topic after the Vice Presidential debate that evening. Conducting the analysis before the debate made the results even more interesting. I have used Condor’s Twitter Collector to collect Twitter feeds and Condor Web Collector’s Google Blog Search to collect website data. Here are the results, pre and post-debate.

Pre-Debate Analysis

Twitter

 
Following image shows the network of users talking about Paul Ryan and Joe Biden on Twitter. The blue dots are users talking about Joe Biden, the green are talking about Paul Ryan and the red are talking about both of them.

We can see that there are more people talking about Paul Ryan compared to those talking about Joe Biden. The following image shows that the betweenness centrality (or popularity) of Paul Ryan is much greater than that of Joe Biden.

Now let us see what people are actually talking about. Following image shows the terms widely used by users while talking about Joe Biden.

We can see that “Obama”, “Ryan” and “tonight” are quite important (or have higher “betweenness centralities” in SNA terminology). This is expected. However, the interesting terms that appear with a decent amount of betweenness centrality are “win” and “good”.

 Now let us see what terms are being used by people talking about Paul Ryan.

Here also, we can see that “Joe”, “Biden” and “tonight” have high betweenness centralities. Interestingly, although the term “good” appears with high betweenness centrality, just as it did in Joe Biden’s network, the term “win” appears with a very low betweenness. Could this mean that although fewer people are talking about Joe Biden, they are talking more optimistically about him?

To go a little deeper, I used Condor’s Taxo tool to further analyze the twitter feeds. I extracted terms from 3 wikipedia categories
1) Emotion
2) Economics
3) Employment

  
Following image shows the terms from these taxonomies that appear in tweets about Joe Biden.

The interesting terms that appear with high frequency are “hope”, “shame” and “economics”.

 
Following image shows the terms from these taxonomies that appear in tweets about Paul Ryan

The interesting terms that appear with high frequency are “economics”, “employment”, “hope”, “motivation” and “victory”.

Although Joe Biden’s taxonomy view doesn’t look too conclusive, Paul Ryan’s view seems very positive. Moreover his betweeness centrality is also high on twitter. Thus I can say that Paul Ryan is clearly more popular on twitter and positively so.

Google Blog Search

Now let us look at the google blog search results.

 
Following image shows the network of websites talking about Joe Biden and Paul Ryan. The blue dots are websites talking about Biden, the green are talking about Ryan and red are talking about both.

Following image shows the betweenness centralities for Paul Ryan and Joe Biden. 

Clearly Joe Biden is more central in blogs. This is different from twitter, where Ryan was more central. 

Now let us see which website is more influential. I have filtered “Joe Biden” and “Paul Ryan” from the websites. Following image shows the betweenness centralities of the top websites.

We can see that huffingtonpost.com is the most influential website in the blog search, followed by foxnewsinsider.com.

Now let us look at the sentiment analysis of the content on the websites. The dots above the zero represent positive sentiment and the ones below the zero represent negative sentiment.


Joe Biden’s sentiment view:

We see a lot of positive sentiment for Joe Biden.

Paul Ryan’s sentiment view:

Ryan’s sentiment view looks quite balanced and is not very skewed towards positive, unlike Biden’s, which is clearly positive.

Conclusions:

1) Twitter analysis shows Paul Ryan to be clearly more popular and positive
2) Blog analysis shows that Joe Biden is more popular and positive
3) Since twitter is more dynamic than blogs, I would conclude that Joe Biden has a historical positive sentiment going for him, but Ryan is the new and popular kid on the block.
4) We need to keep an eye on huffingtonpost.com and foxnewsinsider.com to see what they are saying about the 2 candidates. What they say is clearly very influential

Post-Debate Analysis

Twitter

The twitter feeds after the debate were collected every hour from the time the debate ended at about 10:30PM EDT till the next morning. This was done so as to be able to see the change in the betweenness centralities of the two candidates over the few hours after the debate.

 

The following image shows the network of users talking about Joe Biden and Paul Ryan after the debate.

 

The blue dots are users talking about Joe Biden, the green are users talking about Paul Ryan and the red dots are users talking about both the candidates. Unlike the pre-debate view, we can see a much larger fraction of people is talking about Joe Biden after the debate.

 

The following image shows the betweenness centralities of the 2 candidates.

Joe Biden is in fact more central post-debate than he was before the debate. This does not tell the complete story though. Let’s analyze the content of the tweets to figure out what people are actually saying about the 2 candidates.

The following image shows the terms that appeared frequently in tweets about Joe Biden.

In addition to the obvious “tonight”, “Obama” and “vice” we see the terms “rude” and “laughing” being used a lot. These terms do indicate a bit of negativity in the tweets about Biden. After the debate, Joe Biden was criticized quite a bit for his condescending and rude demeanor during the debate. 

The following image shows the terms that appeared frequently in tweets about Paul Ryan.

There is nothing that really stands out that may be attributed to a positive or negative tone in the tweets about Paul Ryan.

Now let us look at how the betweenness centrality of the two candidates changed over the few hours after the debate. The blue line represents the betweenness centrality of Joe Biden and the green line represents the betweenness centrality of Paul Ryan over time.

We can see a surge in Joe Biden’s centrality for some time after the debate. However, Paul Ryan does catch up and finally surpass Joe Biden in centrality in due course of time. This indicates that more people did talk about Biden for some time after the debate, but went back to talking about Ryan soon. Moreover, the analysis of terms shows that quite a bit of the talk about Biden might have been negative.

Google Blog Search

The following image shows the network of websites talking about Joe Biden and Paul Ryan after the debate.

We see that Joe Biden and Paul Ryan have comparable betweenness centralities, unlike the view before the debate. Paul Ryan definitely seems to have gained some ground in the blog space. Now let us look at the sentiment analysis of the content in the websites.

Joe Biden’s sentiment view:

Biden’s sentiment view is still quite positive but it does seem to have lost some positive sentiment.

Paul Ryan's Sentiment View:

 

Paul Ryan’s sentiment view is clearly positive now, unlike the one before the debate.

Let us now look at the websites that are most influential. Once again I removed the search terms “Joe Biden” and “Paul Ryan” from the websites network. Here is the picture showing the most influential websites.

huffingtonpost.com is once again the most influential website. However, foxnewsinsider.com seems to have disappeared after the debate.

Conclusions:

Here are some conclusions based on the analysis after the debate 

1) People spoke about Joe Biden a lot more on twitter after the debate than they did before the debate
2) The content of tweets about Joe Biden might have had some negative tone.
3) Websites spoke more about Paul Ryan after the debate than they did before the debate
4) The sentiment analysis of the content on blogs shows an increase in positive sentiment for Paul Ryan and a slight decrease in positive sentiment for Joe Biden
5) The website to keep an eye on for news on these two candidates in huffingtonpost.com
6) Paul Ryan certainly seems to be the winner in terms of likability of the only Vice Presidential debate in this race to the White House. However, what effect the debate has had on their electability and consequently that of the presidential candidates is a coolhunting task in itself

I am now looking forward to the next coolhunting task that I plan to undertake. The analysis of the second Presidential debate.