Ranked in these QuestionsQuestion Ranking
Pros
Pro Easy to set up
Watson is a bit of a one stop shop for tone analysis. It's packages are self contained, so setup is simple and can be injected into nearly any app.
There aren't any dependencies and setup can be done via the command line by simply setting up an endpoint and a user engagement point. All setup steps are well documented here.
Pro Globally used with a good community
Watson is one of the more common tone analyzers and is used worldwide. The community surrounding it is fairly large, so any questions you have can be directed to an educated userbase. You can find Watson developers on StackExchange, the IBM forums, and they even have a Slack channel.
Pro Can detect emotion through both slang and emoticons
Pro Developer friendly
IBM provides SDK's for Android, Java, Node.js, Python, Salesforce, Swift, .NET, OpenWhisk, Unity, Go, PHP, Ruby, and Scala.
Using the service is easy, simply passing the input text to the service via GET or POST, and the service returns an analysis in JSON format.
Pro Good accuracy
Watson was tested against standard emotion sets like SEMEVAL and ISEAR, and was found to be around 41-68% accurate, beating out the best reported accuracy of the state of the art models which came in at around 37-63%.
You can read more about Watson's accuracy here.
Cons
Con Limited non-English support
Some users have reported that Watson has trouble with other languages. Watson only officially supports English and French, so its support for other languages can be limited. Users have reported trouble with the Spanish tool specifically, saying that it's slow and unresponsive.
Commonly Compared
Alternative Products
