[SOUND] One of the most popular applications of natural language processing is in social media analytics. The biggest challenge is going out to the web and pulling in posts from a variety of social media sites. Fortunately, there are companies that aggregate social media posts from many platforms and provide data in a format that is ready to be analyzed. Of course, they charge a fee for this. Once you have the data, what do you do with it? Well, there are two main kinds of analysis that are most commonly done with social media data. First, you look at how often your product or brand is mentioned in peoples posts. This is called buzz. Second, you look at how often your product or brand is mentioned in relation to your competitor's brands or products. This is called share of buzz. Your analytics software then essentially counts how often your product or brand is mentioned and how often your competitors are mentioned. Combined with sentiment analysis in which your analytics package evaluates whether people are saying positive or negative things in their posts, social media analytics can give you a wealth of information about how people are generally perceiving your product or brand. As an additional enhancement, you can combine social media analytics with other kinds of data, such as market research, sales trends, call center activity, promotional activity, and so forth. Social media analytics are especially useful for monitoring long-term trends as your company introduces new products, response to changes in the market and makes business decisions affecting the availability and quality of its products. For example, suppose there is a new trend for people preferring outdoor dining when the weather is pleasant. You might notice an uptick in the buzz surrounding restaurants that introduced this option. People run to social media after their night out and begin making comments. It was a pleasant experience, just like being on vacation in Europe. It was fun to do some people watching in the fresh air. Of course, there can be negative comments as well. It was way too hot and they didn't have enough umbrellas. I got a sunburn and will avoid that place in the future. It was so windy that our umbrella blew over and ruined our meal. Of course, part of your analysis involves comparing comments about your restaurant with comments about other restaurants. You'd like people to be seeing more nice things about your establishment than the competition. With good analytics, you can have useful information that can drive business decision making. One final thing you can do with social media analytics is to see how people are linked to each other. If there's a particular person who makes frequent positive posts about your business and you discover that they are very well connected to other social media users, you might consider offering them product discounts or free samples to encourage them to continue making positive posts. As you know, people tend to follow advice from their friends or from others they admire. As you can see, social media analytics can be an especially valuable tool in your data analytics tool kit. [MUSIC]
