[Repec-data] A first proposal

Thu Sep 12 10:30:15 UTC 2013

To react on the previous exchanges, OECD Publishing would be interested to supply different data types to RePEc: dataset and data tables organised in "series" or "data collection" 

1. We do not currently index each dataset series separately. In terms of granularity we only index each complete dataset of a data collection separately. Even if this may evolve.
So to take the example of "OECD Main Economic Indicators" database (OECD MEI)  mentioned in previous exchanges, we could have in such a case a "series" (in the sense of a data collection) OECD MEI collection (http://dx.doi.org/10.1787/mei-data-en ), and  fill a template for each of its dataset (e.g. key short-term indicators http://dx.doi.org/10.1787/data-00039-en )  

We also index :
2- data tables from our statistical publications.
Again with the example of the "OECD Main Economic Indicators" monthly publication (http://dx.doi.org/10.1787/22195009)  composed of around hundred tables, we could have a "series" OECD MEI  (in the sense of the periodical publication) and fill a template for the latest released data tables (e.g. Gross domestic product, volume, 2005=100 http://dx.doi.org/10.1787/mei-v2013-8-table1-en)

3- thematic key  tables or indicators extracted from our databases organised in "series" typically " Gross domestic product in US dollars" (http://dx.doi.org/10.1787/gdp-cusd-table-2013-9-en)  
In that case it might be interesting to supply the proposed properties as such:
Entity-Class: Country
Entity-Name:  >> multiple entries with all countries covered

Entity-Class: Time Series
Entity-Name: GDP

Entity-Class: Units
Entity-Name: At current prices and current PPPs in billions of US dollars

3. country table or indicators organised in series,  typically "Country statistical profile: United States"  (http://dx.doi.org/10.1787/csp-usa-table-2013-1-en) 
In that case it might be interesting to supply the proposed property as such:
Entity-Class: Country
Entity-Name:  US

I agree that it should be up to the provider to decide the level of granularity which can be supplied depending on the organisation of  its data catalogue.‌
We would prefer using the same template for the different data type: dataset, table, time series...
Based on the above, the proposed template sounds fine as a starting point . 
A few comments /suggestions:
- do not understand what is "Frequency-update" property compared to "Frequency-data". 
- It should be possible to limit "Range start" and "Range End" to a year (YYYY format), otherwise a property "time range" (years covered) would be useful to add (YYYY-YYYY).
- provide the ability to link to related items not listed in RePEc.

Thanks,
Pascale 

Pascale Cissokho Mutter
OECD Publishing
Project and Information Management

‌

-----Original Message-----
From: Repec-data [mailto:repec-data-bounces at lists.openlib.org] On Behalf Of David K.Levine
Sent: 30 August, 2013 10:35 AM
To: repec-data at lists.openlib.org
Cc: repec-data at lists.openlib.org
Subject: Re: [Repec-data] A first proposal

Then I think entity-name/class needs to be handled in a more flexible way.

taking the example

US GDP would have Entity-Class: US and Entity-Name: United States

first I think the Entity-Class should be "Country" not "US"
second, there needs to be a place to indicate "GDP" plus indicate if it is nominal, real, PPP adjusted, what the base year is for price adjustment and so forth

can we allow repeated entries, with sequential Entity-Class/Entity-Name pairs be associated, so

Entity-Class: Country
Entity-Name: United States

Entity-Class: Time Series
Entity-Name: GDP

Entity-Class: Units
Entity-Name: real 1966 PPP adjusted dollars

----- On 8/30/2013 12:25 am repec-data at lists.ope wrote -----

Yes, as long as appropriate metadata is provided. Hence the need to define it adequately.

On Thu, 29 Aug 2013, David K. Levine wrote:

> So then the idea would be for software to provide the higher level organization by reading the templates and allowing them to be searched?
>
> ----- On 8/29/2013 11:20 am repec-data at lists.ope wrote -----
>
> That is really up to the data provider. You want the granularity to would
> help users find what they want. To take the OECD MEI example again, users
> generally do not look for national accounts or the MEI, they look for a
> specific series of a specific country.
>
> So yes, I would create 148,000 templates, ne

==

David K. Levine

Professor of Economics and Joint Chair RSCAS
Department of Economics                    http://www.dklevine.com/
European University Institute              phone:  [+39] 055 468 5954
Villa San Paolo                                     office: VSP 45
Via della Piazzuola 43
I-50133 Firenze - Italy

John H. Biggs Distinguished Professor
Department of Economics
Washington University in St. Louis     phone: [+1] 314 935 9529
Campus Box 12081 Brookings Dr.
St. Louis MO 63130-4899  

_______________________________________________
Repec-data mailing list
Repec-data at lists.openlib.org
http://lists.openlib.org/cgi-bin/mailman/listinfo/repec-data