The Internet and the World Wide Web have changed mankind, forever. It is to early too tell, but their impact may be as great as the combustion engine or the introduction of electric devices. The Internet gave universal access to information, not just information that broadcasters or newspapers thought that was important, but information that interests the ‘websurfer’. However the Internet is not a one way street, every Internet user is also a producer: people make websites, maintain blogs, post tweets, and socialize via social networks.

Since a substantial part of the world has Internet access, and every user is also a producer, there is an enormous amount of information available. Some of it is of peer-reviewed and of a high quality, most of it is unchecked and biased. Still, every single piece of information can contain valuable information. For instance, as a vacuum cleaner producer, you might think that social media are not so interesting. However, the contrary is true: between billions of messages there may be hundreds expressing sentiment about your product. Such messages can answer questions about how your brand is conceived, what problems people commonly have with your product, etc.

Obviously, it is out of anyone’s reach to manually analyze a significant portion of the information that is available on the Internet. The amount is just too overwhelming. Classic data analysis tools may not suffice either, most of the information is seemingly unstructured, and consist of blobs of natural language sentences. However, language is also structure. It is just not the kind of structure that computers can normally deal with. Computers deal with neat XML files, fragments of JSON, or comma separated values. No, it is the structure that we humans use to convey and transfer meaning. This book is about that type of information. We will go into many of the techniques that so-called computational linguists use to analyze the structure of human language, and transform it into a form that computers work with.