Projects
Fundamental
Language Detection
Text Comprehension cannot be envisaged without language detection. The same goes for Automatic Translation.
We needed that feature in order to conduct text analysis which help us understand which topics a text covers. Per se it has never been considered as a prime target; it is more a by-product or prerequisite for text analysis, lato sensu.
Language Detection is performed by statistics: we use a limited lexicon of ± 500 words/language to detect what language a text is written in.
Our algorithms proved to be correct 98% of the time, which is considered more than enough for the usage we make of it.
Text Comprehension
Sentiment & Opinion Analysis
Text-to-Speech
Geo Data
Automatic Translation
Generic Concepts
Most generic concepts we have worked on are extracted from schema.org and Wikidata
Mother
All concepts have a common ancestor: Mother
Mother is an abstract
class: it cannot be
instantiated per se, except through a class that depends
on it.
Our computer code is shared as Open Source on GitHub and we keep updating it regularly.
WebPage
Each page of the TRQL Radio site is known as a
WebPage
object. Our definition is based on the
one provided by schema.org (https://schema.org/WebPage).
The definition of WebPage
is itself based on
a parent concept, CreativeWork
(https://schema.org/CreativeWork). In turn,
CreativeWork
depends on a higher concept:
Thing
.
As we can see, all these concepts are a bit of a nest of concepts and lead to many basic definitions. At this point, we have worked on almost 1000 different concepts.
Our computer code is shared as Open Source on GitHub and we keep updating it regularly.
WebSite
All the pages of the TRQL Radio website are part of the
… of the TRQL Radio website. It was therefore important to
define what we meant by website. To do so, we took as a
starting point the definition that comes, again, from
schema.org (https://schema.org/WebSite
).
This standard definition is suitable for many projects that are related to the creation of a websites : websites for artists, websites for radio stations, …
Our computer code is shared as Open Source on GitHub and we keep updating it regularly.
Radio
Being a radio makes this notion a cornerstone of the code of TRQL Radio.
Our computer code is shared as Open Source on GitHub and we keep updating it regularly.
Doc: trql.radio.class.php
Weather Report
More info to be provided soon
Weather Forecast
More info to be provided soon
Applied Research
Automated News
Automated News is all driven via interests that are gathered in (1) sources (2) topics
All sources we collected are RSS sources we have come across and have considered reliable; topics have been carefully selected manually upon a number of interesting keywords.
For example,
https://www.researchsquare.com/rss.xml
is
considered a valid source for the science
topic.
News are extracted and categorized as to be presented in an ordered manner in our Vae Soli! Player and our News page
Wheather Reports & Wheather Forecasts
A Weather report is known as a report about aspects of the weather (https://www.wikidata.org/wiki/Q67179467).
A Weather forecast is known as the application of science and technology to predict the conditions of the atmosphere for a given location and time (https://www.wikidata.org/wiki/Q182868).
Both notions refer to high-level objects that we have infered either from Schema.org or from Wikidata, or both: Weather Report and Weather Forecast.
Such notions cannot be tackled without a perfect understanding of geo data for which we use the geonames API services.
Traffic Info
For everything related to traffic information, we use the TomTom API (https://developer.tomtom.com). The traffic API requires an extensive comprehension of geo-data.
Artist Info
Today in History
We Are One, You And Us, We Are TRQL Radio, open radio from Belgium, fully driven by Artificial Intelligence.