Using NooJ to Process Satellite Data

Abstract

This paper describes the processing of satellite telemetry data and its conversion from reduced form into full, orthographically correct word sequences for further use in text-to-speech synthesis or other applications.

The purpose of a telemetry system is to reliably and transparently convey measurement information from a remotely located data-generating source to users located in space or on Earth. For the purpose of this paper, by satellite telemetry data, we are also referring to the results of measurements made by nanosatellites in space and received with antennas.

In this study, we used a text corpus which was created with the data from satellite telemetry converter software. The data is provided in the form of abbreviations, numbers, and measurement units, along with orthographical words (e.g. Voltage of 5 V system is 4.904 [V]; Temperature of the 145 MHz TX: 16 °C). Telemetry data is mostly collected in English as an international language for science, yet Belarusian national space tradition requires this information to be available in one of the national languages of Belarus, Belarusian or Russian.

As input, the grammar takes as text sequences like 0.9708 A and transforms first the numerical part, then measurement units, into a Belarusian phrase: нyль цэлыx дзeвяць тыcяч ceмcoт вoceм дзecяцiтыcячныx aмпepa ‘zero point nine thousand seven hundred and eight ten-thousandths of an ampere’. Due to the fact that Belarusian and Russian are both synthetic languages, the declension paradigm of such numerals can be very complex.