Those who have been chomping at the bit to use IBM's Watson machine-intelligence service with their apps need gnaw no longer. Watson APIs are now available for public use, albeit only through IBM's Bluemix cloud services platform.
The eight services currently available include:
- Language Identification, which can determine what language a given text is written in (from a predetermined set of 25).
- Machine Translation, which translates text between multiple language pairs.
- Concept Expansion, which can take a colloquialism (say, "tri-state area") and map it to a set of meanings based on context (New York, New Jersey, and Connecticut).
- Message Resonance, which can determine the popularity of a given word with a predetermined audience.
- Question and Answer, which provides "direct responses to user inquiries fueled by primary document sources." Topics in health care ( "What causes scabies?") and travel ("Which museums are in Manhattan?") are offered as the first two knowledge bases.
- Relationship Extraction, which can parse sentences to determine the relationships between components as a way for other analytic systems to better understand the significance of what's being discussed. For example, if fed "Mark Wahlberg spoke yesterday about his new film," it would understand that "Mark Wahlberg" is a person, "yesterday" is a time reference, and "film" is the object of the sentence.
- User Modeling, which employs linguistic analysis to make predictions about someone's social characteristics from a supplied text.
- Visualization Rendering, which generates data visualizations from different kinds of data -- not merely pie or bar charts, but also the likes of flow charts and node graphs.
Of the services offered so far, Visualization Rendering seems the most immediately useful and powerful, since it isn't limited by data training many of Watson's other services rely on. Most of the services rely on a "corpus," or cultivated body of data that Watson can use as raw material, so the breadth of several Watson offerings is limited by the size of their existing corpora.
Watson's Machine Translation system, for instance, only translates between a few pre-selected pairs of languages: English, Brazilian Portuguese, Spanish, French, and Arabic. Even then it doesn't always go both ways. For instance, English to French is supported, but Arabic to French is not.
The Relationship Extraction system seems less limited by available data than Machine Translation, but it is limited in different ways. When the Relationship Extraction system is fed the sentence "Nick Cave's new film '20,000 Days on Earth' debuted yesterday," it understood that "Nick Cave" was a person and that "yesterday" was a date, but didn't understand that "20,000 Days" referred to the title of a work. Maybe that level of sophistication is planned for a later release.
Apart from a generic analytics market, sales and marketing stands out as a target audience for the Watson APIs. IBM dropped hints of this when it unveiled a clutch of mobile apps designed to use Watson's services, one of them a customer profiling and recommendation engine. The Message Resonance and User Modeling APIs are further advances in that direction, although they hardly seem like the kinds of algorithms that require the vaunted strength of Watson.
Since Watson seems like a plum way for IBM to promote use of Bluemix, it seems unlikely Big Blue will offer access to the Watson APIs outside that service anytime soon. But the real measure of Watson's worth will be determined by how quickly the individual offerings expand (such as more languages for Machine Translation) and how many more APIs are rolled out in the months to come.