Hello everyone,

I am the CTO of LinguaSys. We recently released a public cloud version of our NLUI Server framework (available at http://nlp.linguasys.com, direct link: https://nlp.linguasys.com/docs/services/547d8bd70644500e9493e492/operations/547d9399547de90a50d1f9d3).

It is also a markup script-based chatbot engine, with a couple of differences:

[ul]
[li]* semantic and morphological awareness for the supported 20+ languages. You don’t work with strings that you look up in the user’s input; you actually look for concepts. E.g. you can add a condition like “any kind of Asian” and it will match “Chinese, Japanese, Indian, Korean, Cantonese, ...” etc. - far beyond the omnipresent WordNet. In highly-inflected languages like Russian, German, Arabic, it will match the concept regardless of the inflection, number, case, etc.[/li]

[li]* crosslinguality. As you can see in the documentation, the scripts contain ID numbers of concepts. Since the ID numbers are the same across languages, once you wrote your script, it supports all the languages that LinguaSys supports. [/li]

[li]* mixed-initiative. It’s not a tree or a hard-coded flow. The engine has objectives to achieve, and the user can provide them at once, or be guided through the process. You, as a script writer, don’t even have to think about how to guide them.[/li]

[li]* easily integrated with external web services, libraries, databases, etc.[/li]

[li]* use it as a web service in the cloud, on premises, or embedded (the last one is .NET only but only for now)[/li]

[/ul]

You can access other services, as well, like low-level parsing for all these languages, morphological analysis, sentiment analysis, etc.

Any feedback is appreciated.