The OACC provides a comprehensive framework of universal mathematical measures of algorithmic complexity for researchers and professionals. It retrieves objective numerical measures of randomness for potential applications in a very wide range of disciplines, from bioinformatics to psychometrics, from linguistics to economics.
It is based on several years of research devoted to new methods for evaluating the information content and algorithmic complexity. The description of the Coding Theorem method to deal with short strings is described in this paper. It currently retrieves numerical approximations to Kolmogorv complexity and Levin’s universal distribution (or Solomonoff algorithmic probability) for binary strings of short length, for which lossless compression algorithms fail as a method for approximation to program-size complexity, hence providing a complementary and useful alternative to compression algorithms. More algorithmic information measures, more data and more techniques will be incorporated gradually in the future, covering a wider range of objects such as longer binary strings, non-binary strings and n-dimensional objects (such as images).
It also includes a Short String Complexity Explorer (it may take some time to run if you open the link) tool developed in Mathematica, it runs oinline with a free player. It provides a comparison of the estimated Kolmogorov complexity (K), the algorithmic probability (m), Shannon’s Entropy (E) and compressibility (using Deflate) of a string. It also provides an indication of the relative complexity among all other strings for which K has been approximated and its distribution rank.