pyerk 0.5.0

Python based 'Easy Representation of Knowledge' (pyERK)

Overview: pyerk

Pyerk is the python implementation of the emergent representation of knowledge framework.

• Designed to formally represent knowledge (including meta levels)
• Implementation-status: "early alpha"
• Inspired by OWL, but much more expressive
  • ... at the cost of guarantied computability
• Inspired by Wikidata, but much simpler
• Inspired by SUO-KIF, but with less brackets
• Represented directly in python: → imperative instaead of declarative knowledge representation

Pyerk originated and is currently (2022) mainly developed with focus on representing knowledge from the domain of control theory as part of the Automatic Control Knowledge Repository (ACKREP). However, in principle, it aims to be applicable wo a wide range of domains.

Assumed Directory Structure

<erk-root>/
├── pyerk/                          ← repo with the code of the core package
│  ├── .git/
│  ├── README.md                    ← the currently displayed file (README.md)
│  ├── setup.py                     ← deployment script
│  ├── src/pyerk/auxiliary.py       ← module containing function get_erk_root_dir()
│  └── ...
│
├── django-erk-gui/                 ← repo with the code for the django gui (project and! app)
│  │                                  (this package is optional)
│  ├── .git/
│  ├── manage.py
│  └── ...
│
├── erk-data/                       ← directory that contains erk-knowledge packages (for actual usage)
│  ├── ocse/                        ← a knowledge package (ontology of control systems engineering)
│  │  ├── .git/
│  │  ├── README.md
│  │  ├── ocse.py                   ← a knowledge module
│  │  │                             (this one is currently used for unit testing)
│  │  └── ...
│  ├── xyz123/                      ← another knowledge package
│  │  └── ...
│  └── ...
├── erk-data-for-unittests/         ← directory that contains erk-knowledge packages
│  └── ...                            (unittest version, probably older and different from
│                                      production data.)
└──...

Documentation

There is not yet "real" documentation available. However, to get an overview of the most importent features you might want to have a look at the source code, especially at the files builtin_entities.py and the test_core.py.

Nevertheless, some topics are explained here until there is some more comprehensive documentation.

Visualization

Currently there is some basic visualization support via the command line. To visualize your a module (including its relations to the builtin_entities) you can use a command like

pyerk -rwd --load-mod demo-module.py demo -vis __all__

Interactive Usage

To open an IPython shell with a loaded module run e.g.

pyerk -i -rwd -l control_theory1.py ct

Then, you have ct as variable in your namespace and can e.g. run print(ct.I5167.R1).

(The above command assumes that the file control_theory1.py is in your current working directory.)

Multilinguality

Pyerk aims to support an arbitrary number of languages by so called language specified strings. Currently support for English and German is preconfigured in pyerk.settings. These language specified strings are instances of the class rdflib.Literal where the .language-attribute is set to one of the values from pyerk.setting.SUPPORTED_LANGUAGES and can be created like:

from pyerk import de, en

# ...

lss1 = "some english words"@en
lss2 = "ein paar deutsche Wörter"@de

where en, de are instances of pyerk.core.LangaguageCode.

The usage inside pyerk is best demonstrated by the unittest test_c02__multilingual_relations, see test_core.py.

Coding style

We use black -l 120 ./ to ensure coding style consistency.