suneal / helion-1

Helion

Helion is a data-driven framework that models the regularities of user-driven home automation, generates natural home automation scenarios, and provides stakeholders with tools to use the scenarios and obtain actionable outcomes.

Getting Started

1. Download necessary files:

• MITLM - MIT Language Modeling Toolkit
• Python Daemon 1.5.5 - Library to implement a well-behaved Unix daemon process.
• Brain Files - The language model server script which reads/writes JSON documents to named pipes.

2. Untar MITLM and Python Daemon:

Execute the following command to un-tar the the MITLM and python daemon packages:

$ tar-xzf <package_name.tar.gz>

3. Add proper directories to your path:

Execute the Following commands or add them to your .bash_profile to set up your $PATH. Be sure to replace the paths with local paths on your machine

export PYTHONPATH=$PYTHONPATH:/Path/to/Python/Daemon/python-daemon-1.5.5/
export PYTHONPATH=$PYTHONPATH:/Path/to/Python/Daemon/python-daemon-1.5.5/daemon/ 
export PYTHONPATH=$PYTHONPATH:/Path/to/Python/Daemon/python-daemon-1.5.5/daemon/version/ 
export PATH=$PATH:/Path/to/MITLM/mitlm-0.4.1/ 
export PATH=$PATH:/Path/to/Brain/Files/

4. Build MITLM:

Navigate to the mitlm-0.4.1/ directory and execute the following commands:

$ ./configure
$ make

After both of these commands have been executed, you should be able to see the estimate-ngram, evaluate-ngram, and interpolate-ngram executables in the mitlm-0.4.1/ directory.

5. Instantiate the Brain:

Navigate to the folder where your training and vocabulary files are located. Then instantiate the brain by running:

$ braind data/helion.train data/helion.vocab

For more detailed instructions on how to interact and instantiate the Brain, see the README.

Execution Engine

Coming soon..

Snapshot Module

Coming soon..

Acknowledgements:

• Kramer