Bots, fake news, and disinformation–oh my! It seems that not a day goes by without new reports of social media bots being used by political propagandists both foreign and domestic. At times, the situation seems hopeless. But it is not. My experience over the last year with incorporating basic social media intelligence (SOCMINT) activities into my courses gives me hope that we can be smarter about social media than we are now.
For the last two semesters, I have incorporated a series of assignments related to “social media intelligence” (SOCMINT) into my course on IT and Global Conflict at the University of Utah. So far, these assignments have been a success. Students have enjoyed the assignments and have also done consistently well on them. I figured it was time to write up a post about the objectives of the assignments, the tools students have been using, and the results so far.My experience incorporating basic #SOCMINT activities into my courses gives me hope that we can be smarter about #socialmedia than we are now. #OSINT Click To Tweet
Why Teach Social Media Intelligence?
The internet and social media have emerged as important sources of data used by journalists, academics, businesses, and governments to follow and understand a range of issues and actors online. These include monitoring for brand mentions, as well as monitoring narratives and participants involved in various global conflicts. Of course, with reports of Russian use of Facebook and Twitter during the 2016 U.S. presidential election, we know that there is a race to use social media as tools to project influence on a global scale.
This provided the context for incorporating these assignments into my course. In fall 2016, we were in the midst of the election season. But there were already accusations of Russian interference through a combination of hacking, leaking, and armies of online trolls and bots. Students were well aware of this situation and, even in conservative Utah, very much concerned.
At the same time, many commentators pointed to a lack of media and internet literacies among the American public as a reason for Russia’s success. Social media was causing “filter bubbles” and served to reinforce peoples’ pre-existing beliefs. Social media allowed fake news to go viral, reaching unsuspecting internet users who were ill-prepared to suss out real reporting from conspiracy theory.
Enter social media intelligence…
What is Social Media Intelligence?
Social media intelligence, or SOCMINT for short, is often described as a sub-discipline of open source intelligence (OSINT). In turn, these “INTs” are among a number of intelligence disciplines with their roots in the world of national security–e.g. militaries and spies. These include disciplines of human intelligence (HUMINT), imagery intelligence (IMINT), signals intelligence (SIGINT), and more. SOCMINT then is an increasingly important source of openly available information on the internet that can be collected and exploited to develop intelligence related to important national security or law enforcement matters.
Beyond this, however, the phrase “social media intelligence” is also used frequently among marketing professionals. In this case, the term has a similar meaning but is more related to monitoring brand mentions, customer satisfaction, etc. on social media in an effort improve understanding of market conditions, effectiveness of advertising campaigns, etc.
From these two definitions alone, I thought it important for students to get some hands-on experience with how tools for monitoring, collecting, analyzing, and disseminating information on social media work. If militaries and spies are using the tools and techniques of social media marketing and monitoring to wage a new kind of war–or so the theory goes–then students should get a taste of those tools and techniques.
But my use of “social media intelligence” also had a second meaning that reflected a second objective of the assignments. That was to be intelligent about social media, what it does and doesn’t do, what we can and cannot learn, how it can be manipulated, etc. Thus, the two main learning objectives of these assignments was for students to
- Use tools of social media monitoring, collection, analysis, and dissemination to gain insight on a topic or issue of concern to them, but also
- To be more intelligent consumers of social media content as a result of this experience.
Social Media Intelligence Activities
In the series of assignments I have used so far, students are introduced to a collection of online tools that can be used to follow an issue of interest to them that is also relevant to the topic of the course. This use of social media is sometimes called “social media intelligence” (SOCMINT) and is a variant of open source intelligence (OSINT).
The first assignment in the series is to decide on an issue to follow and a set of search terms that will be appropriate for effectively following that issue.
Over the course of the semester, students are introduced to, and complete assignments using, the following tools and search terms to following their chosen issue:
1) Google News & RSS Aggregator
Students select a set of search terms and conduct a search of Google News. They subscribe to the RSS feed for these search results using an online RSS aggregator. In the first semester, we used Feedly. But we switched to Inoreader this last fall as Feedly no longer supports Google News RSS feeds in its free offering. Students report on their findings and on the usefulness of these tools. This includes providing a list of sources reporting on their issue during the exercise period (e.g. New York Times, Politico, etc.), a description of the current debate about their chosen issue based on the news coverage during the exercise period, and their personal reflection on the tool.
Next, students use the same or closely related search terms used in the Google News & RSS Aggregator exercise to search Twitter using the TwXplorer tool provided by the Knight Lab. During the exercise period, students run their search once per day and save a snapshot of the results. As before, students report on their findings at the end of the exercise period. Reports include sharing links to daily snapshots, reflection on the usefulness of the tool, the most common hashtags used, the most shared news stories, and how those news stories compare to the ongoing results of their Google News search.
The third assignment in the series involves using the Twitter Archiving Google Spreadsheet (TAGS) tool. They use TAGS to pull and archive Twitter search results over the exercise period. Again, they report on the findings of their research, including the “Top Tweeters,” volume of tweets over time, and most retweeted tweets, as reported by TAGS. Once again, students reflect on how useful this tool is and what they learned about their topic by using this tool in conjunction with ongoing use of Google News, RSS, and TwXplorer.
Based on their previous research, students create a Twitter list of the “Top Tweeters” who are posting about their topic. That list is added as a column in the Twitter application, TweetDeck. They monitor the list during the exercise period and report on what they learned about the issue they are tracking using this tool during the exercise period, as well as their assessment of the value of TweetDeck for following their chosen issue.
Next, students use the service, TweetedTimes, to create a “newspaper” based on their list of “Top Tweeters,” their search terms, or both. They will take note of the news items posted each day by people in their Twitter list and how those compare to the most shared news items from their ongoing monitoring of TwXplorer and the news items aggregated from Google News in their RSS readers. They report on similarities and differences in news items and, once again, reflect on the usefulness of this tool.
6) Simple Twitter Bots
Finally, in the last assignment, we turn our attention from monitoring and collecting social media content to posting it. Students use the Iterative Twitter Bot, which works based on a Google Spreadsheet, to build simple Twitter bots that automatically post content to Twitter related to the issue they are tracking. Once again, they reflect on their use of the tool, in particular how their views about “bots” and their dangers has been impacted (if at all) by making a bot of their own.
SOCMINT in the Classroom: The Results So Far
A number of things have stuck out to me so far based on student reports after each assignment.
The first is that these assignments have been wildly popular. Students enjoy learning these tools and being able to follow an issue of concern to them. Most students enter the class with issues related to the class that are of interest and concern to them. Sometimes we may think that students do not pay attention to the news, especially international affairs. What this assignment has shown me is that this is not the case for most of my students. Of course, they are a self-selected group. Nonetheless, it has been rare that there is a student who can name no issue of concern to them that they would like to use as the topic of their SOCMINT assignments.
Second, I think we also tend to think of students as tech savvy, especially when it comes to social media. This is a remnant of the over-used “digital natives” rhetoric. Similar to my experience teaching web design over the years, I do not find that students are particularly savvy when it comes to social media use. There are several ways I find this to be the case:
- Students have often not used the platforms we focus on in our assignments (Google News and Twitter) in more than a casual way. For example, few are aware of how to effectively search for content. Most have never used advanced search queries or Boolean logic. Almost none have used or even heard of RSS before.
- Many students initially see Twitter as an alternative source of news to the news stories they are finding in Google News. Each semester, we have had a discussion in class about what we are really trying to learn from social media. That is, we are trying to learn about peoples’ perceptions, what is catching their attention, what is trending, etc. We are not necessarily using social media (Twitter in this case) as a news source, but rather, as a tool to gauge public discourse around an issue of concern to us. Who is talking? What are they saying? What are they sharing? Etc. This is why I include the Google News and RSS assignment. Though not really “social media,” these tools provide a point of comparison. Students come to realize that the news stories catching the attention of, or being shared by, at least some of the people they follow on Twitter differ from what they are seeing in their Google News results.
- The Simple Twitter Bot assignment was newly added this last semester. The goal was to show how easy it is to set up a bot. In turn, I hoped that this would help demystify “bots” to some degree. However, “simple” as it may have been, this assignment proved the most challenging to students, especially those without any kind of technical background (which was most). Thus, I worry that the assignment did not really have its intended effect, or may have even reinforced to some that all bots are sophisticated and frightening creations of “l337” Russian hackers.
Many students initially see Twitter as an alternative source of news. But we are not using Twitter as a news source, but rather, as a tool to understand public discourse. #SOCMINT #OSINT Click To TweetWhat’s Next?
So far, assignments have focused on Twitter. As social media platforms go, Twitter is fairly easy to search, collect, and analyze. So, it is a good intro for students. Facebook, Instagram, and others can provide more of a challenge. However, I do think that future iterations on this series of assignments should include these other platforms.
I am considering adding or replacing one or two of the current assignments with ones that combine gathering Facebook shares data using SharedCount and CrowdTangle. This would allow students to take a series of news stories from their Google News results and see how those are being shared and by whom on Facebook, for example.
Another option could include finding and collecting multiple forms of data surrounding a particular issue or event from multiple platforms. For example, I may have students find and collect tweets, images, and videos on an issue or event using Twitter, Instagram, and YouTube.
However, all of this points to one final observation and challenge: There just isn’t enough time for it all! The primary focus of the class in which I have included these assignments is not SOCMINT, and certainly not OSINT or digital research methods more broadly.
In the long term, I think there could be great value in developing and offering such courses at the university level. The goal would not be to turn out digital stalkers or spies. Rather, the goal would be to use such activities and tools to improve students’ “intelligence” about news and social media, including the benefits and limitations of each. Of course, such a course might also just improve students’ ability to conduct online research and make effective use of the vast amount of information they have at their fingertips.