Until recently, text analysis, sometimes known as natural language processing, was a little-known specialty, but I’ve been working in this area since the 1980s. Suddenly, text analytics is the new hot area, so after 30 years I am an overnight success.
Text analytics is an umbrella term for a broad range of techniques that help computer programs “understand” text. Some uses for text analytics include:
- Part-of-speech analysis. Sometimes, it can be extremely helpful to know the difference between a noun and verb. Text analytics can identify each word’s part of speech in dozens of national languages.
- Entity extraction. Knowing a word is a noun isn’t always enough. It can be very important to identify a word as a name of a company, or a city, or the name of a person, or a place. Entity extraction can pull those words out of running text and identify exactly what kind of proper noun they are.
- Sentiment analysis. It’s one thing to find text that mentions your company, but is it a positive or negative opinion? Sentiment analysis tells you.
Over my career, text analytics has undergone a huge shift from being based completely on linguistic rules to adding great reliance on machine learning algorithms that provide for more accurate answers.
Text analytics technology has many uses for businesses–everything from identifying your brand’s popularity in social media to automatically processing resumes for job seekers. Some uses for text analytics identify sales leads to increase revenue, while other uses focus on reducing the costs of jobs that human beings had to perform. I also have lots of experience with hybrid text analytics systems, where the technology does 70% of the job and humans clean up the parts that computers don’t yet do well.
If you have lots of text that you think a computer could analyze, I’d love to help. If you’re interested in talking to me about a consulting opportunity with your company, I’d like to hear from you, so please contact me.