3 ms·
I'm working on something akin to this, although my version is somewhat more ambitious[1]. So far I have a "Knowledge Engine" that can answer (some) questions i
by nl 13y ago
I'm working on something akin to this, although my version is somewhat more ambitious[1].
So far I have a "Knowledge Engine" that can answer (some) questions in natural language by using a process similar to how IBM Watson Deep QA works (an extremely simplified version atm, though).
I'm currently working on integrating SymPy for symbolic calculation[2]. SymPy already has a web interface[3] which is broadly similar to what WolframAlpha can do. That relies on Google AppEngine to provide security sandboxing.
My version is designed to be completely private, open source and self-hostable, including all the data[1], so I can't rely on AppEngine. Instead, I have SymPy running inside ZeroVm + LXC (Docker) as a sandbox. I'm currently working out how to get that to communicate with the front end (which is non-trivial, because ZeroVm is pretty much undocumented).
Obviously SymPy isn't as "Natural Language Like" as the WolframAlpha language, but arguably Python is actually more powerful.
After that I'll do OpenStreetMap integration (ie, self hosted OSM), which will give it better geographic understanding that it has.
[1] Every time I mention this I need to include a standard disclaimer that I do realize this is completely and utterly crazy.
[2] http://sympy.org/en/index.html http://sympy.org/en/index.html
[3] http://live.sympy.org/ http://live.sympy.org/
- taliesinb 13y agoHow do you represent queries? That's really, for me, the most interesting question.
- nl 13y agoI run multiple parsing strategies in parallel. One is based on Quepy[1] (which generates a parse graph, then runs RegEx on the parsed objects), and a simple thing I wrote using a combination of NLTK and a Java processing thing. For "question answering" (the bit I have working) both end up being converted into SPARQL queries and then run against a local RDF store. For SymPy the idea is that my frontline parallel pipeline thing will throw the query at SymPy as well as the other parsers, but SymPy will be the only thing that knows what to do with it. [1] http://quepy.readthedocs.org/en/latest/ http://quepy.readthedocs.org/en/latest/