by Jason Baldridge and Lillian Lee
Language is the holy grail of artificial intelligence. When we imagine sharing a world with smart machines, we don't think about logic, or problem solving, or winning at chess. We hear HAL-9000 declining to open the pod bay doors, and the Terminator saying he'll be baaack. Researchers have been working on building computers we can talk to for 60 years; in the 1990s, Bill Gates predicted that speech would soon be “a primary way of interacting with the machine”. So why aren't we talking to our computers yet ....Or are we? Thanks to new developments in human language technology (also known as "natural language processing") and text analytics, computers are analyzing everything from e-mail and tweets to clinical records and and speed-date conversations. How does the technology work, when does it work well (and when not), what's it doing for us, and where is it headed?
by Bruce Smith
Social media includes lots of free-form textual data in "natural languages", the languages people speak. Natural Language Processing (NLP) helps you analyze that data. Some NLP problems are very hard, but a number of lightweight NLP techniques are available in open-source tools. You can use these to improve your social media applications, even if you're not a computational linguist. In this session, I will introduce some of these techniques and tools, and I will give hints on getting started using NLP in your social media applications. Many NLP techniques require training corpora, sets of annotated documents. I will talk about constructing and maintaining a training corpus. And, I will talk about some of the ways we use NLP at Lithium Technologies.
Electronic health records have the potential for enormous good, but in order for them to live up to their full potential, information about patients -- their symptoms, diagnoses, allergic reactions, medical backgrounds, family histories -- must take the form of standardized, structured, easy-to-manipulate data. One obvious way to get there is to tightly structure the way that doctors create the medical record. As a result, physicians are under increasing pressure to abandon unrestricted natural language and the clinical narrative, and turn the medical documentation process into a jungle of pull-down menus, checkboxes, and restricted vocabularies. In this presentation I argue that the results could be catastrophic, I make the case for preserving the clinical narrative, and I argue for a practical way out of the dilemma: using natural language processing technology to produce the structured records we need, while still allowing physicians the freedom of unrestricted clinical language.
9th–13th March 2012