Fun with Semantic Networks

I admit I’m kind of a nerd.

Being a nerd has its advantages, like making a killing at pop culture Quizzo and the ability to spout random movie trivia. (For example, did you know that Steven Spielberg named the “Jaws” mechanical shark after his lawyer, Bruce?)

Another advantage to being a nerd is when it applies to my professional career.

Image: healingdream / FreeDigitalPhotos.net

For instance, I get really geeked out about semantic networks, which have applications for public relations, marketing, social media, and lots more.

Put simply, a semantic network is a visualization of the relationships between words. The analysis is how researchers (or us communicators) interpret those relationships.

Semantic network analysis lends itself to public relations and social media research because it can help communicators visualize content and identify opportunities or potential issues.

There are several and paid tools out there that can help you chart semantic networks. Here are a few I have bookmarked and find helpful in my communications research:

1. Wordle - this is a quick, easy way to generate a semantic visualization. Wordle asks users to simply take a text (be sure to convert Word docs into Plain Text) and it will generate a word cloud where the most prevalent words appear larger than others to emphasize dominance in the text.

Word clouds are helpful to get a quick-hitting analysis of common threads or topics in Tweets, blog comments, or news articles.

Here’s a Wordle of all my blog posts, showing that I write predominantly about media, PR, and “social,” which could be linked to media:

Image courtesy of Wordle

2. MentionMap – hat tip to Adam Vincenzini for bringing this tool to my attention (and for sucking up an entire afternoon once I got into it!)

MentionMap pulls public information from a Twitter account (Tweets, hashtags, @replies) and creates an incredible, almost organic network visualization to demonstrate the connections between the content and the origin.

This visualization network is helpful because it can identify strong discussion streams among the Twitter account and others, and also identifies dominant hashtag discussions.

Here’s a MentionMap of my Twitter Account, showing I contribute to the #PR and #socpharm discussions frequently:

Image courtesy of MentionMap

3. Crawdad – this is a paid software I became familiar with through a graduate class on social and semantic networks.

Crawdad is more advanced and technical than the other tools mentioned. Like Wordle, it asks for text that can be converted into a semantic network chart where prominent words and conversations are connected.

The researcher can code certain words, remove unnecessary ones, and create a comprehensive network to identify dominant themes and connections. Crawdad also allows the researcher to compare texts, for instance press release messaging vs. news articles, to add another layer of information.

Here’s an example of a semantic network I created through Crawdad a few years ago, using news articles from The Philadelphia Inquirer about Purdue Pharma. It looks kind of crazy, but makes sense to see the prominent words that appeared in the news articles over a period of time:

So, while my random knowledge of classic films may not do much for my career, using semantic networks in communications research certainly does. It’s fun to learn something new and to see data visualized in a colorful, engaging way.

Semantic network analysis may not be the end all, be all of communications research. While I like to believe that data can’t lie, the analysis part that is connected to the human mind can be fallible.

So what gets you geeked out that has been helpful in your professional role? Do you have any other semantic network research tools to share?

About these ads