[Tracker] tracker and the semantic desktop [was: tracker 1.2.4]

[mlo <tracker member holoclan de>]

Am 06.11.2014 um 23:43 schrieb John Bestevaar:
> Hi All As a non coding user type of follower of Tracker, this is
> very similar to how i would like to organise my thousands of
> images. Hence text/strings attributes stored within( metadata
> )image files.

regarding images you can imho use shotwell.
I am much more interested in documents (my own as well as collected
ones) of various sort, e.g., Office (odt, ods, ... yes, even
MS-Office) , Layout (i use Scribus, so it's XML), infographics (SVG),
Mindmaps, LaTeX, etc.

> On 06/11/14 23:42, mlo wrote:
> > Am 06.11.2014 um 12:28 schrieb Martyn Russell:
> > > On 06/11/14 11:12, mlo wrote:
> > > > Am 06.11.2014 um 11:36 schrieb Martyn Russell:
> > > Hallo Martin,
> > >
> > > > > Tracker uses W3C standards for RDF ontologies using Nepomuk
> > > > > with SPARQL to query and update the data.


this morning I invested some of my rare spare time to look into
tracker-sparql.

I constructed this query to get an overview of tags tracker had
automatically assigned.

tracker-sparql -q '
           SELECT ?tag ?file WHERE {
             ?f nao:hasTag [ nao:prefLabel ?tag ] ;
                 nie:url ?file .
                 FILTER (!EXISTS { ?f rdf:type nfo:Image } )
           }'

The result was a bit depressing, as it only had considered the
"Keyword" Property of a few PDFs worthy for tags.

How exactly does tracker auto-generate tags?
> From which file-types does it extract tags?

... without the nfo:Image FILTER I got 893 Tags from about 80000 image
files. As I use shotwell to organise my photos, I am quite stisfied
with tagging images (there is always room for improvement).

greets
martin

> > > > Thanks Martyn for that bit of information! I just learned
> > > > from it, that Nepomuk is a common ground in KDE and GNOME. Am
> > > > I right?
> > > That depends a bit.
> > >
> > > > How close is the relation?
> > > Tracker technically uses Nepomuk + extensions to fill the gaps
> > > we need filling, so yes it does use the standard, but not the
> > > library or source works from the KDE folks.
> > >
> > > From what I've heard lately, KDE is moving away from Nepomuk,
> > > so it was common but perhaps not now?
> > >
> > > As for libnepomuk*, I have no idea, but we're not using it.
> > >
> > > > Is there a chance of, e.g., dolphin using the same RDF-store
> > > > as tracker?
> > > Well, tracker is an RDF store. Perhaps you mean "using the
> > > Tracker RDF store" ?
> > >
> > > > Would I break tracker if I tried it out?
> > > Try away, but it's unclear to me what you think you would be
> > > breaking here. If you have indexed content in Tracker and you
> > > change the ontologies installed in to
> > > /usr/share/tracker/ontologies/ then the next restart of
> > > tracker-store might spit errors out at you ;)
> > well, my first attempt lead to reindexing my files with baloo I
> > diden't want that because I would really like to stick with
> > GNOME. dolphin can't (at least on first sight) use the tracker
> > store. I was hoping I could keep tracker as backend and use
> > dolphin to browse and search and (most important for me) to edit
> > tags.
> >
> > even in this very short time I started to really like dolphin.
> >
> > > > I just browsed through the Nepomuk and semanticdesktop.org
> > > > sites. There is a lot of good ideas and promising tools. The
> > > > "DropBox" for one; I'd really like to try that one.
> > > >
> > > > What do you think?
> > > I've not looked into it personally.
> > >
> > > > btw.: I used to do some research in a semantic-web project
> > > > (called SWAP - you can still find it online -
> > > > http://www.aifb.kit.edu/web/SWAP) years ago myself ...
> > > What's your aim here, may I ask?
> > of course you may :-)
> >
> > I very much like the idea of a semantic desktop.
> >
> > Having all your data dynamically grouped or structured by
> > different (combineable) criteria, e.g., (self attached or
> > automatically extracted) tags, timeline, type, location, etc. is
> > really the way I would like to organise myself.
> >
> > unfortunately the (as of today) useable tools, that go in that
> > direction are still quite far away from this vision.
> >
> > SWAP is just one of so many academic projects ... by chance I had
> > a tiny part in this one ;-)
> >
> > Another project I worked with, used to calculate a semantic
> > distance between documents and graphically arranged them
> > according to this distance. So if you selected a document, the
> > "closest relatives" based on a changeable set of criteria were
> > grouped near it, others further away in a hyperbolic tree
> > http://en.wikipedia.org/wiki/Hyperbolic_tree. This one was an
> > internal project at Dresdner Bank, so it is not available to the
> > public.
> >
> > So that was a little trip on my motivation behind this discussion
> > ...