search

polifonia-project / KG_data_transformation

Script and services to extract data from raw files JSON,csv,tsv and create Polifonia KG
0 stars 0 forks source link

some HarmonicSimilarities contain 4 ChordProgressions #6

Closed phivk closed 2 years ago

phivk commented 2 years ago

see this example in the ttl file I am using for testing: https://github.com/polifonia-project/KG_data_transformation/blob/d4a8a81ec975b8938f6741e88c2015952946a107/harmonic_sim_etl/kg/versions/polifonia-kg-harmonic-test.ttl#L73-L77

The HarmonicSimilarity harm_sim_00424_00339 contains the following ChordProgessions:

I assume this encodes a similarity based on:

The problem with this is that, while creating annotations for the sonar demo, I have no way of telling whether 00339_Fmaj#_G#5_C# is related to 00424_Fmaj_Cmaj_Fmaj or 00424_GmajB_Cmaj_GmajD (likewise for all other combinations).

I might be able to use the ordering of the ChordProgressions, but that feels brittle and hacky.

Could we consider restricting HarmonicSimiliarities to a single pair of ChordProgressions (at least for the time being)? If so, should this be done at the stage of 1) calculating HarmonicSimilarities @andreamust @jonnybluesman? OR 2) KG data transformation @delfimpandiani @ccolonna @FiorelaCiroku? OR 3) am I missing something and I should be able to deal with this at the Sonar data transformation step?

ccolonna commented 2 years ago

Hi @phivk , can you show the query you are using to extract values ? There may be some possibility to map values!

phivk commented 2 years ago

here you go @ccolonna

ccolonna commented 2 years ago

Good catch Philo. I discussed this with @andreamust and we're probably missing an edge

e.g. :chordprogressionA :hasSimilarChordProgression :chordProgressionB

We'll investigate it a bit more and fix this. Then you will be able to relate chord progressions correctly !

ccolonna commented 2 years ago

Hey @phivk ,

you can look at this, the PR should close the issue. https://github.com/polifonia-project/KG_data_transformation/pull/7#issue-1030185954

Now you have a link between related chord progression. In the PR there's the link to the ontology used to model data in the KG.

You can also see rules used here .

I close the issue for the moment in case you see something else feel free to reopen this or a new one.