Researchers who study Canada have generated large quantities of geohistorical data for many years. While we reflect on the creation of a national geohistorical infrastructure, it is pertinent to identify datasets at different scales which can become a part of such a portal. We are therefore trying to enhance the discoverability of existing and available datasets. In the long run, it would be preferable to enumerate and describe each layer and each attribute  table, it is not necessary, for the moment, to delve at such a level of detailed granularity. We hope, at this stage, to identify collections which have emerged from different research projects or from the online deposit of previously georeferenced digital data such as:

• raster geographic maps
• aerial photographs
• vector layers
• attribute data linked to vector layers

We have already identified datasets offered by different types of creators so that we can present diversity in the nature and the type of data which can interest researchers. We have therefore identified:

• quality international data (FAO)
• data from collaborative mapping projects (Open Street Map, Natural Earth)
• data available on GIS company web sites  (ESRI)
• national data (government of Canada, Géogratis)
• provincial or territorial data (British Columbia, Yukon, Québec, Nova Scotia, Prince Edward Island, New Brunswick)
• municipal data  (Toronto, Montréal, Sherbrooke)
• research team data (CIEQ, NICHE, LHPM, MAP, VIHistory)
• data from map library and archive centres (Scholars’ Geoportal, MADGIC, GéoIndex+)
• personal initiative data  (historical railway lines )

Choosing what type of metadata to associate with each dataset has meant achieving a compromise. An insuficient level of detail would prevent effective searches while requirements for overly detailed metadata could discourage data creators who are not trainted to  create metadata which meet international standards. According to Rodolphe Devillers, we can use six criteria to define the quality of a geospatial dataset¹.

i. Definition : Allows the user to evaluate if the nature of a datum and of the object it describes, i.e. “what”,  meets his or her requirements (semantic, spatial and temporal definitions);

ii. Coverage : Allows the user to evaluate if the territory and the period for which the data exists, i.e. the “where” and the “when”, meet his or her requirements ;

iii. Genealogy : Allows the user to know where the data came from, the project’s objectives when the data was acquired, the methods used to obtain the data, i.e. the “how” and the “why” and to verify if this meets the user’s requirements ;

iv. Precision : Allows the user to evaluate the data’s worth and if it is acceptable for the user’s requirements (semantic, temporal and spatial precision of the object and of its attributes);

v. Legitimacy : Allows the user to evaluate the official recognition and the legal standing of the data and if it meets the user’s requirements (de facto standards, recognised good practices, legal or administrative recognition by an official agency, legal garantee by a supplier, , etc.);

vi. Accessibility : Allows the user to evaluate how easily the user can obtain the data (cost, delays,format, privacy, respect of recognised good practices, copyright, etc.).

A metadata standard which would meet all of these criteria may seem overwhelming for may people who would like to make their data available. We therefore propose to use the format defined by the Dublin Core Metadata Initiative, an international standard for which the types of fields are easier to understand for people less familiar with metadata. We have applied and interpreted the DCMI based upon its general definition available on Wikipedia² and on the interpretation of a few fields proposed by the Bibliothèque nationale de France³. This approach can certainly be criticised, because it is geared towards a simple application rather than perfection. Based on how metadata will be entered in this list, we can refine these principles to improve this compromise. The fields do not appear in the same order as in the DCMI and some are subdivided to provide for a slightly finer level of granularity.

Table 1. List of fields used to describe datasets

A list of identified resources is available here:  http://bit.ly/2rlIkRC. Some of the notices are incomplete and we are working on completing them. If you would like to propose a dataset, you can fill out the form available here: http://geohist.ca/donnees-sigh-hgis-data-form

1  DEVILLERS, Rodolphe (2004). « Conception d’un système multidimensionnel d’information sur la qualité des données géospatiales », [En ligne], Ph. D., Université Laval <http://theses.ulaval.ca/archimede/fichiers/22242/22242.html>.

This post is also available in: French