Search engines used to have it so easy. Their bots could crawl all around the internet and just index everything, paying no heed to whether the information was an accurate representation of what each particular website was trying to say.
Then came social media, and everything changed. Suddenly, you had millions of different web commenters posting their opinions on virtually everything. Now the web crawlers were indexing information from every random Joe, many of them with wildly divergent opinions on each issue.
So what’s a web crawler to do in this crazy world of over-opinion?
To solve this problem, Visible Technologies developed a sentiment analysis tool that is a crucial component to social media monitoring.
Sentiment analysis is an automated service which determines the attitude, opinion, or even emotional state of a given speaker or writer. It can be very useful for companies who want to quickly get broad information about the general feeling towards their brand or products in the social media sphere.
Sentiment analysis isn’t just as simple as “good” or “bad,” however. At Visible, our sentiment analysis is guided by these principles:
- Sentiment is up to interpretation- In some cases, only a human can accurately gauge the sentiment of a given comment. We account for this ambiguity in our analysis.
- Varying degrees of sentiment- Sure, a program can determine if something is good, but how good? Are people raving about your product or brand, or just OK with it?
- Context is everything- There are certain subjects (like movie or theater reviews) which are inherently opinionated. There are other subjects which don’t necessarily cause opinionated reactions. Context is always included in our analysis.
Of course, this only is just a tiny amount of the complexities within Visible’s sentiment analysis technology. For a more in-depth look at this exciting concept, check out our sentiment analysis white paper.
