Version 1.3.2

Changed the method name of the betweenness function from "minimum" to "infimum".

This reflects the fact, that this variant of betweenness actually computes the infimum (i.e., the limit) rather than the minimum (i.e., the actual value) over all alpha-cuts, because the latter may not exist. No changes have been made to the code's logic, this is merely a more consistent naming. The old value "minimum" can still be used (i.e., the code is backwards compatible).

Version 1.3.1

The prior versions of this code did not compute the correct intersection results in cases where the cores of the two concepts overlap, but the parameter mu of the two concepts differ.

This update fixes this bug by providing a partial re-implementation of the respective code and additional test cases in order to ensure that it functions correctly.

Version 1.3.0

Migration to Python 3.6 (kindly provided by @marius-pol )

Version 1.2.0

Version 1.2.0 of our conceptual spaces implementation contains the following changes:

• Additional method crisp_subset_ofwhich checks whether a given concept is a crisp subset of another concept.
• Revised implementation of conceptual similarity and betweenness (only two variants each, but now fulfilling some mathematical criteria).

Visualization, Sampling, Similarity & Betweenness

Version 1.1.0 of our conceptual spaces implementation contains the following additional functionality:

• Visualization: A new module called concept_inspector provides interactive 2D and 3D visualizations of the concepts in the current conceptual space
• Sampling: You can now sample points from a concept based on its membership function. If you want to, you can also export these points in a csv file.
• Similarity & Betweenness: The code now contains quite a number of options for computing the similarity and betweenness of concepts.

Version 1.0.0

First stable release of our implementation.