July 3rd internal meeting

IMG_0869

 

Present : Alexandre Aubourg, Loïc Charles, Guillaume Daudin, Pierre Gervais, Paul Girard, Jérémy Hervelin, Demba Konate, Emmanuelle Lavallée, Ivan Ledezma, Sandrine Mesplé-Somps, Laura Moutou, Guillaume Plique

 

1) Discussion on RICARDO

The transition from Access to SQL Lite and Matteo Azzi’s work were first described by Paul and Beatrice. Paul also introduced the interface and its 5 “views”. Depending on the “view”, we will be able to observe the flows between two countries and their trade declarations, compare global trade between entities or follow the evolution of the share of each country.

The data is still being cleaned up and a large part of the work should be done by the end of the month. Furthermore, the database should be put online in October. The objective is to make the project publicly available. Some issues regarding the use of API to download the data by researchers were also raised (e.g. we need to make them available in csv rather than JSON).

 

2) Work with Guillaume Plique

If it is accepted by Europe, Guillaume Plique will start working with us in April 2015. Transcription and brainstorming about the datascape should then be finished. He will work till October on developing it.

 

3) Online storage of the data

Pictures and databases can be found on www.livedrive.com. If you wish to access the files, make comments or contribute, do not hesitate to contact me to obtain the username and the password.

 

4) Transcription of the data

Demba and Laura are currently working on La Rochelle. National and regional data are also being retranscribed by the two service companies.

6 years will be delivered by SMSI – only 1788, 1789 and the 19th century will be remaining from the national data. Regarding the regional data, Thasectra will send a first batch (Bordeaux, Nantes) in September and the data from Marseille, Bayonne and Rouen should be ready by the end of the year. Thanks to the regional sources, we will be able to reconstitue the French national trade from 1725 onwards.

Prices and transcription errors will be checked by Jérémy, Demba and Alexandre. Everything should be imported into Stata by the end of January 2015.

 

5) Expenditure

They are now all available : thanks !

 

6) Discussion on the database

The database is made up of roughly 120 000 observations giving the details of bilateral flows by products. It is linked to a “Country database” (300 observations) and a “Unit database” converting quantities to the Metric system – the structure of which has to be defined. Other relational databases aim at standardizing (8000 observations) and classifying the names of goods.

The history of local nomenclatures, the links between regional and national sources and the issues related to quantity units were discussed. Regarding the names of goods, Paul insisted on the importance of classification levels and described how we could create networks with the names. We discussed the differences between administrative and business practices in naming goods. In any event, it will be essential to :

– work on the spelling of information units in the name of goods (Open Refine is our friend)

– find a way to exploit the original categories (notably regarding the use of  “;”)

– rethink the organization of the relational databases.

– we need to think about research questions ASAP. Ivan will collate what we promised and what he believes we can do

 

7) Next meeting

September 18th at 10am at SciencesPo médialab (13 rue de l’université)

 


Laisser un commentaire

Votre adresse e-mail ne sera pas publiée.

Ce site utilise Akismet pour réduire les indésirables. En savoir plus sur comment les données de vos commentaires sont utilisées.