Which model is better for data translation in Power BI ?

[DidierTerrien]

Hello,

I'm working hardly on translations as I think a data culture adoption needs reports to be translated when we want to reach more users. I wrote a bunch of blog posts explaining how to translate model and visuals using Excel and a plugin I wrote. I'm focusing now on translating data but I run into problems once I draw the relationships. 

Which of these models is better according to your experience ?

Language	Value
En	Car
Es	Coche
Fr	Voiture

or

Value En	Value Es	Value Fr
Car	Coche	Voiture

My favorite is the first one but I'm open to other point of view as the report becomes complex when several dimensions need to be translated and when I avoid many to many & bidirectional relationships.

My test report is here 

Thanks in advance

Didier Terrien

[amitchandak]

@DidierTerrien , the first one.

[DidierTerrien]

Hello,

Thank you for your quick reply !

Can you developp a bit ? Could you translate several dimensions avoiding many to many & bidirectional relationships ? Is the report easy to maintain ?

[v-deddai1-msft]

Hi @DidierTerrien ,

I also agree with the first one.  In my opinion, the first one is more conducive to creating a star schema. And Using star schema can avoid many to many & bidirectional relationships. Please refer to https://docs.microsoft.com/en-us/power-bi/guidance/star-schema

Best Regards,

Dedmon Dai

[DidierTerrien]

Thank you Dedmond,

The first one brings a many to many relationship for each dimension table so we need a bridging table for each dimension table. I think we are going out of the star schema. The Many-to-many relationship guidance doesn't consider this case :

• Relate two dimension-type tables
• Relate two fact-type tables
• Relate higher grain fact-type tables

Regards

Didier terrien

[v-deddai1-msft]

Hi @DidierTerrien ，

These need to be adjusted in specific circumstances, you cannot force the star model to your data model.  There are also many functions in DAX that can avoid you from creating complex relationships such as USERELATIONSHIP ,TERATAS, CROSSFILTER and so on.

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

Best Regards,

Dedmon Dai

[DidierTerrien]

Thank you Dedmond,

As far as I understand, we need relationships to be able to slice by dimensions which is obviously necessary. The functions you mentioned don't help for slicers. Please tell me if I am wrong.

It would be very helpful if Power BI team could provide some example with good practices for data translations. 

Regards

Didier Terrien

[v-deddai1-msft]

Hi @DidierTerrien ,

Maybe you can get some sample from :https://docs.microsoft.com/en-us/power-bi/create-reports/sample-datasets

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

Best Regards,

Dedmon Dai