Currently aether is a modest collection of experiments in text analysis and data visualisation. Its main focus is on data available on or retrieved from the internet.
Note: See the '/notes' directory for much more in depth documentation.
Development of code to:
- investigate and determine the most efficient/accurate way to rank texts based on similarity.
- develop best-practices for writing higher-order functions to facilitate extremely modular code.
- provide simple visualization of the internet as a graph.
The eventual intent is for Aether to become a platform for providing novel approaches to searching, exploring and analysing the
internet itself and the wealth of data it contains, and for serving applications built from these endeavours.
Interaction with the platform will hopefully be structured in three tiers:
A comprehensive and modular 'engine' that allows development of programs which:
- leverage core libraries/algorithms for data-mining, machine-learning, data-visualization, etc.
- manipulate, combine, and modify algorithms and any other code written by other users.
- make use of any data collected by users.
- follow a set of interface guidelines so that all code is easy to recycle and all data collected is easily accessible.
Due to the vast amounts of data involved, much code on this tier will likely execute on the cloud and not be directly a part of any specific application. Rather, this level will provide the primary substrate for expanding Aether's functionality. All tested and conforming code can be pulled into the main code base.
An API that allows developers to access community data and/or call functions which have been included into the platform. Through this API, applications can be developed that will be able to run on Aether. These can then be submitted and run on the cloud where they will be accessible over the internet, whether in a browser or otherwise.
The Aether website. Here end-users can use developer applications which may either be free or paid.