Analyze and Visualize Public Opinion Data From the Web
For years, students sat through statistics courses with no real direction. There wasn’t a clear-cut benefit to the data analysis.
That’s no longer the case.
Statistical thinking helps businesses derive insights and make data-based decisions. It allows departments to predict future trends and mitigate risk. And along with our technology partner Rosoka, our team at BrightPlanet uses modern statistical thinking to analyze and visualize public opinion data we harvest from the Web.
Harvesting and Enriching Online Data
The Internet is infinite. Instead of being overwhelmed by the endless stream of knowledge, we think it’s important to realize what this means for your business: an unlimited amount of information waiting for you to harvest and use.
It starts with data harvesting.
When BrightPlanet harvests data from a website, we collect all of the text-based content. This means we’re left with an original record of the page, regardless of whether it changes or is moved.
The next step is enrichment, which is done through entity extraction. In simple terms, this means we’re identifying the words within the data that hold value to you. It can be people, places, brand names — whatever information you want highlighted.
The enrichment step is done through software, but it’s similar to someone reading word-by-word and highlighting the key terms. Keep that in mind when you’re considering the value of data harvesting services.
Exploratory Data Analysis
After harvesting and enriching data, it’s important to take the final steps: analyze and visualize. That’s what takes our work beyond basic statistics to the next level of exploratory data analysis.
But it’s difficult to comprehend this process if you don’t already have a sense of data literacy. That’s why we created an example around public opinion and the 2016 presidential election to help you make sense of the process.
Public Opinion on the Presidential Race
We collected data to gauge public opinion on each of the candidates in the 2016 race. We harvested more than 14 million global news articles and Tweets to identify public opinion on the candidates in both news and social media.
With the entities (candidates and campaign issues) extracted, we had a full list of structured data to determine what’s the public opinion of these candidates and the issues.
For our webinar, we used a visualization tool called Tableau to create a full dashboard with data about specific candidates, where they are being mentioned, and in what context.
Within this dashboard, we can also track the polarity and salience of the candidates in each article, meaning their importance to the piece and the light in which their mentioned (i.e. is it a positive or negative piece).
In our research, we found a few important pieces of information. Major events in our world sway mentions away from the candidates, such as the shooting that occurred in Oregon in early October. And international coverage varies greatly from that of the American media, showing which issues are valued most by different cultures.
Take Advantage of Data Resources
The political campaigns are dominating the public consciousness, but BrightPlanet and Rosoka’s technologies can help anyone use large amounts of Web data to better understand public opinion.
This could help you gain insight into the value of your brand or the future of your company. It’s time for companies of all shapes and sizes, from all industries, to take advantage of Web data.
Are you interested in improving your data literacy and improving your business? Contact our team for a free demo and we’ll show you how BrightPlanet can help harness the endless power of the Internet.