This week Kansas City CBS affiliate KCTV5 used information from a Spark sampling on the H1N1 vaccine. The results, along with some video of our virtualization, were featured on last night’s 6 o’clock newscast.

In a five-day sampling of web content from last Friday to this Wednesday, Spark found that overall sentiment was 53.24% neutral. This is a big amount for such a widespread topic.

But the real red flag comes when you when narrow down the sentiment analysis to social media websites only.

On social-based sites like Twitter and Facebook, where more sentiment-based chatter usually happens, the percentage of neutral sites is even higher (a whopping 78.21%).

This suggests that social media is being used extensively to get the word out about the availability of the vaccine. Health and news organizations are successfully using social media outlets to spread the word in local markets.

The government and its agencies are doing a superb job of counteracting all of the conspiracy theorists and disinformation out there about the vaccine (and believe me, there’s a lot) by using social media tools to launch a messaging blitz.

Semantic analysis is key to learning the vernacular that people are using to talk about the H1N1 vaccine. The semantic cloud below lists the most frequently used words across the sampling.

The cloud shows that the glut of actual information from the government and health organizations is so large that it is drowning out all the sentiment-based chatter.

Value-judgment words are barely even present in the Top 100. (Except for “like,” which has many different uses.) This is an almost completely information-based semantic sampling.

As viewed in the video and in the screenshot above, a swine flu FAQ from WebMD is the most influential URL, with a link from the government-run Centers for Disease Control and Prevention that serves as a vaccine Q&A being the number-two most influential.

The New York Times topic “swine flu news” and the Wikipedia entry for “swine influenza” are third and fourth most influential URLs for this sampling, based on Spark’s proprietary algorithm.