A notation for expressing frequency of data collection

Hi,


On the time-related front, the vocabulary provides a notation for the duration of the retention period. However, I believe there is another dimension that has to be captured - how often the data are sent.


As an example, compare two IoT devices that measure the temperature in a room. One device sends a data-point every hour, while the other does it every second.


One could argue that in the latter case, the potential for a privacy violation is much higher, considering that an algorithm can use the data to answer questions such as "is there someone in the room?", "how many people are there?" or "what activity are they currently engaged in?".


A straightforward way of expressing it could be a ENUM with the following values, understandable to a non-expert: {multiple times per second, every second, every minute, hourly, daily, weekly, monthly, on}.


A more rigorous notation could leverage `TemporalUnit` (https://www.w3.org/TR/owl-time/#time:TemporalUnit), followed by a number, e.g. `:unitMinute 10` means "the data are sent every 10 minutes".





Some context: the angle from which I approach this matter is that of transparency - how to make it easy for an end-user to understand the implications of data collection practices of a device they own.


We've designed a "privacy facts" label for IoT devices, as well as a GUI with additional visualizations, you can see the prototype and tinker with it here: http://privacy-facts.eu


Our goal was to come up with a usable UI and use an existing vocabulary, instead of inventing our own. DPV is exactly the piece of the puzzle that we need, and we encourage its use in our papers.


Alex