IEML
Open source artificial method to represent the semantic content of a linguistic sign
From Wikipedia, the free encyclopedia
IEML (Information Economy Meta-Language) is an open source artificial method to represent the semantic content of a linguistic sign. It was designed by Pierre Lévy as an Open collaboration project as part of his works on Collective intelligence[1][2][3] in order to encode meaning in a computer readable way. Its design is based on mathematics and logic abstractions but with a clear inspiration from the organic structure of natural languages.[4]
| Information Economy Meta-Language (IEML) | |
|---|---|
| Original author | Pierre Lévy |
| Written in | Python |
| License | GPL-3.0 license |
| Website | https://intlekt.io |
| Repository | https://github.com/IEMLdev/ieml |
Overview
The goal of the IEML system is to make real-world data machine-readable. It proposes a standard representation that enables the mapping of semantic representations with the data in a computer-friendly way.[5][6]
IEML's design starts with a small amount of primary concepts are arranged in a matrix and composed together in order to create new and slightly more complex concepts, which can be arranged in a new matrix and composed to form even more complex ones, and so on.[7] The arrangement in the form of a matrix and its fractal design make the representation easy to manipulate, quick when calculating the distance between concepts and simple to encode.[8] Each element in each matrix has a unique representation that easily indicates both its location and content.[9] To maintain the integrity of the system, every public submission must pass an automatic analogical verification and must be reviewed by a reliable reviewer before being incorporated or updated into the system.
Challenges
IEML bypasses important challenges of the Semantic web and other semantic representation systems such as vagueness, uncertainty, inconsistency.[10][11] Some of the challenges for IEML include readability, annotation and adoption. Systems that use IEML must deal with these issues in order to work as intended.[12]
- Readability: In order to be computer-readable and semantically connected the system cannot use any one natural language as a representation, which makes it more difficult to be read by a human. Still, the metadata of each element allows user-suggested translations from the IEML concept to any given natural language.
- Annotation: Until more advanced tools are implemented, annotation must be made manually.
- Adoption: In order to grow and improve, the system depends on having an increasing number of users and submitters.