Skip to content
This repository has been archived by the owner on Nov 5, 2020. It is now read-only.

Latest commit

 

History

History
56 lines (42 loc) · 1.6 KB

README.md

File metadata and controls

56 lines (42 loc) · 1.6 KB

TEA

Teach machines how to understand and analyze texts!

TEA stands for Text Encoder and Annotator. By using TEA, you can enrich digital representations of textual documents by using a Markdown-inspired language.

The language

Spans

The TEA language lets you define sections of interest, called spans, by surrounding text with double angle brackets:

<<Bob>> is a good guy. <<He>> is friend with <<Alice>>.

Spans can be given an alphanumeric string as an identifier:

<b<Bob>b> is a good guy. <b<He>b> is friend with <al<Alice>al>.

In most cases, the first identifier can be omitted:

<<Bob>b> is a good guy. <<He>b> is friend with <<Alice>al>.

In the special case of non-hierarchical overlap though, both identifiers are needed:

This is a case of <1<non-<2<hierarchical>1> overlap>2>.

In the above example, the first span, identified as 1, contains the text non-hierarchical, while the second one (2) contains hierarchical overlap.

RDF triples

When spans are defined, it is then possible to add RDF triples describing the selected portions of text by creating a special code block, delimited by lines with +++:

<<Bob>b> is a good guy. <<He>b> is friend with <<Alice>al>.
+++
al rdf:type foaf:Person
b rdf:type foaf:Person
al foaf:knows b
+++

[FIXME: change the example to use real predicates - rdf:type can't be used like that] [FIXME: these are not real triples - they are more like triple templates]

Contribute to TEA

Configure grunt

npm install grunt
npm install grunt-contrib-coffee
npm install grunt-contrib-watch
npm install grunt-contrib-concat