Personal Data Locker? What is it and Why?
Issue/Topic: Personal Data Locker? What is it and Why? (F3D)
Convener: Jim Fenton
Notes-taker(s): Justin Tormey
Tags for the session - technology discussed/ideas considered: Personal Data Stores, Personal Data Locker, PDX
Discussion notes, key understandings, outstanding questions, observations, and, if appropriate to this discussion: action items, next steps:
A key take-away was that the term "Personal Data Locker" or "Personal Data Store" is a bit of a misnomer; they might more accurately be described as "Personal Data Brokers" because the data doesn't have to reside there, but they manage access to the data. The term "PDX" is sometimes used, where X can stand for any number of terms.
Resident PDX expert Phil Windley detailed (and later blogged about) the essential characteristics of a Personal Data Store.
A personal data store without any automation just becomes, "yet another thing for people to manage." In addition to the essential characteristics outlined in Phil's blog a true PDX system would need to handle inputs and events on a rules based system.
As discussed in other sessions, the question was posed, "Isn't XYZ already a data store?" Again, there was some confusion over the concept that a data store is a singular location that houses ALL user data. The answer is, yes, sort of. Any single system could be a data store, but the data could also be stored in other locations and should be easily transferable / sync-able between any n number of systems containing the data. Essentially, just because a system has some data doesn't automatically make it a PDX.
The discussion moved onto topics metadata and standardization. The microformats issue still remains - just because data is stored in a similar fashion doesn't mean it's "meaning" is shared between systems. There are two approaches that can be taken, "standardization" and "descriptions / dictionaries." Dictionaries have the advantage of being bi-directional - this means this over here and it means that over there.