I want to access data about my data!

Hello everyone,

I would like to suggest a goal for Nextcloud which in my opinion could complement really well the already existing ones.
For me as a user, the main reason to use Nextcloud is to have the full control of my data. Yet up to now there is some data that I cannot use (and so fully control): the data about my data. All the information on my files, my contacts, my emails,ā€¦ could be really interesting if more easily accessible. I know that at first glance this can be seen as some speculation of a data geekā€¦ but Iā€™ll try to explain why this is not true

There are two simple objections to an easy access to this data:
security:
it is clear that everything is simple if we are talking about a single person cloud. If we are talking about a big multiuser installation we will have to define who can access the data (I guess @LukasReschke could give useful insight)

usability:
the ā€œnormalā€ user does not want to play with the data files of his/her own files (I think that @jan would say thisā€¦). True, but the target is different:

  • all the ā€œcuriousā€ (statisticians, social scientists or whatever) who are willing to read the data
  • all the developer who can reuse the data to give an easy reading to the normal user

But the developer can access the data already! I have the impression that this is not so simple, correct me if I am wrong. I think that up to now the only developers who can easily access the data about my
contacts are the ones working on the contacts part of the core and on the contact app. They do not want to provide too much information that could not useful or disturbing for many users; this is correct.
Yet, if it would be possible to ā€œqueryā€ the contact app and the email app, a developer (who does not belong to the ā€œcontacts app teamā€) could build a widget for the dashboard (see @Holger_Dehnhardt work) that summarizes which are my contacts to which I write more emails: another could use the position to calculate the distance between contacts (maybe I am interested in that because I am a sales
representativeā€¦); a third one could build a widget or an app that shows me my social graph. And so on. Clearly different users have different needs and it
should be possible for them to use the data in the best way for them.

I think that this could bring to a lot of useful services; it could have also an interesting ā€œpedagogical side effectā€: a lot of people could not fully realize the wealth of information that this kind of data bring; using the information for themselves and their own goals could be eye-opening.

I do not have the possibility to contribute to the programming but I am more than willing to contribute with discussion and data visualization proposals. I also have some knowledge of statistical programming, if useful.

Thank you for reading!!!

1 Like

I think it is interesting that the forum item opened just after this one (Wholesome user statsistics would be very interesting to have is linked to this: having user statistics is a sort of subset of this suggestion.

Ooh, yes. This absolutely falls into what I was talking about, plus a lot more information on what we as admins could really use.

I am going to add some further contribution: here are some issues that deal with access and reuse of metadata

  1. the creation of an extended files metadata feature by @alenkovich (who offers development/funding support)
  2. the creation of a MediaMetadata app which [quote=ā€œoparoz, post:1, topic:1601ā€]
    will automatically extract meta data from media files in order to be able to offer some enhanced features in media apps
    [/quote]
    If feasible, using a common approach for all the metadata could be useful; the ā€œIf feasibleā€ issue is big: the same @oparoz states: [quote=ā€œoparoz, post:5, topic:1601ā€]
    Iā€™d like to support as many types as possible, but the biggest problem always comes from having enough resources to maintain the project, so past a certain project size, it will depend on the number of contributors.
    [/quote].
    The MediaMetadata app will be presented at a talk in the Nextcloud conference by the same @oparoz

Hope this helps.

Youā€™ll like some of the features in Nextcloud 10 :wink:

Thank you very much for answering!!! :slight_smile:

Iā€™ll wait with curiosity!!! :wink:

I happened to read the page:

Fair Web Services

one of the points is on access to metadata

Meta Data
There is not only the actual data users upload to services, but also
meta data such as logs, the information who is looking at which data,
what users did when, etc. Who owns this data? Who has access to it? How
do users control what is happening with their meta data?

I would still love to have this functionality!!!

The Dashboard app is not good at this.

This issue is still pending: Visualize useful synthetic information when user clicks ā€˜Tagsā€™

i keep hoping :slight_smile:

This feature would still be interesting.

It would be covered by @Rello 's Analytics app (https://apps.nextcloud.com/apps/analytics) if/when the use of Nextcloud internal tables as Datasource will be implemented.

This is listed in the roadmap of the app:

Hi,

happy to see any usecase and provide something for that.
Analytics was enhanced that Datasources can be added easily - or also via 3rd party apps.

But do develop it so generic, that a normal frontend-user can do ā€œanythingā€ will hardly be possible. He needs to be guided with pre-combinded information - without you would end up with a database-scripting-tool.

any specific questions you would like to get answered?

talking about social graphs:

I am waiting for chartjs 3.0 to reach a proper state. then something like Graph/Trees will be possible:

or wordclouds

Sorry, @Rello, it took a bit to answerā€¦

You made a difficult question. I get that a complete access to all the tables of the DB is technically challenging and not user friendly at all.

I guess that most of the answers could be provided by three synthetic tables:

  • the sessions
  • the object opened / modified
  • the interactions with other users

The first table (sessions) should already exist, as you can find that information in the security section of the settings. By looking at it you could infer a lot of information about your use of Nextcloud.

The other two tables do not exist; they would be a data extraction from the different apps. This would give the possibility to know which files, emails, contacts, ā€¦ you read and modified (objects) or who you interacted with via chats, comments, emails, shares, ā€¦ (interactions). I get that these tables are not easy to create, also because different apps have different kind of objects (for example deck has boards, stacks and cardsā€¦).

Maybe providing access to your sessions table could be a start, do you agree?

Thanks for listening!