• Aucun résultat trouvé

The Government Data Portal for Germany GovData.de

N/A
N/A
Protected

Academic year: 2022

Partager "The Government Data Portal for Germany GovData.de"

Copied!
13
0
0

Texte intégral

(1)

The Government Data Portal for Germany GovData.de

Konrad Johannes Reiche | September 12, 2013 | Nancy, France

(2)

Open Government Data Government Data

 Motives

– Transparency – Innovation – Participation – Efficiency

 Core Elements

– Machine-readable data – Licenses

– Accessibility

Open Government Data Venn Diagram by justgrimes

(3)

Open Data Example #1 mundraub.de

 Many fruit bushes and fruit trees are unused

 Wild fruits, private grower, organizations

 Data about these plants is collected and published on http://www.mundraub.de

 Data comes from people and administrations who submit their knowledge for the public use

by hybrid.moment

(4)

Open Data Example #2

Glass Recycling Container in Berlin

 Glass Recycling Container in Berlin – City

– Private Organizations

 The Berlin Cleansing Department (BSR) has to clean around them

Problem BSR has no information about the location of many containers

Solution Local administrations do have data about the container’s location and help the BSR by making these data publicly available

by Andreas llerby pixelroiber

(5)

Metadata…

…and Harvesting

 Data is stored and managed distributed

 Why? Centralized data is hardly feasible and beyond administrative

 Heterogeneous data, distributed competence, conflict of interests

 Metadata is used to describe the data

 Distributed data storage with central metadata portal

Harvesting: Copying of metadata for making the data accessible, too

Portal

Document

Dataset

Dataset

Document

(6)

Starting Point

Data Portals of the Länder

Germany is a federalism

Consisting of 16 states (Länder) Administrative power divided

Different Data Provider and Portals Bavaria

Berlin Bremen

Federal Statistical Office (Destatis) Hamburg

Rostock

Environmental Information Portal (PortalU) Geo Data Infrastructure Germany (GDI-DE) Rhineland-Palatinate

and more…

?

DeStatis (David Liuzzo)

(7)

Government Data Portal for Germany GovData.de

 Launch February 19, 2013

 http://www.govdata.de

 Prototyped at Fraunhofer FOKUS

 Different type of data – Datasets

– Documents – Applications

 Focus on free licenses

– German Data License (de-dl,…) – Creative Commons (cc-by,…) – ...

(8)

Quantification in Numbers

 February, 2013

– Datasets: 1,123 – Documents: 12 – Applications: 25 – Daily visitors: 2,000

 March, 2013

– Daily visitors: 500

 August, 2013

– Datasets: 3,797 – Documents: 230 – Applications 15 – Daily visitors: 300

Open Data Licenses on GovData.de

(9)

Building GovData.de Strategy

 Repository software: CKAN (Comprehensive Knowledge Archive Network) – Data catalogue for storing and distributing data

– Developed by the Open Knowledge Foundation (OKFN) – Prevalent format: JSON

– API offers REST Interface

 Metadata Schema (OGD-Metadata)

– Structure used to standardize and unify metadata by data providers – https://github.com/fraunhoferfokus/ogd-metadata

– JSON Schema, keep it simple (few fields), e.g. document data origins – Why the hassle? Different data providers: very heterogeneous data – Make data accessible: unification needed

– Schema not a mere tool, but communicator

(10)

Metadata Schema − Example

Field Subfield Value

Name waste-management-statistics-2013

Title Waste Management: Disposal and Treatment Facility

Author Statistical Office

Maintainer Juliane Sanger

Tags Hessen, Berlin, Visualization, Classification

… … …

Extras

Terms of Use ID: cc-byURL: http://creativecommons.org/licenses/by/3.0 Spatial Coordinates: [[15.02, 47.16], [15.02, 47.16]]

Original Portal http://www.regionalstatistik.de

… ….

(11)

Architecture GovData.de

Portal (Liferay)

Information Pool (Web Portal + CMS)

User Interface for the Data Catalog Indexer + Thesaurus

CKAN

CSW/CKAN Harvester

REST Interface

Browser

Apps

Web Sites of Public Authorities

Subject Catalogs (Geo Data, etc.) Open Data Catalogs (Berlin, Bavaria, Bremen, etc.)

REST Interface

(12)

What’s next?

Outlook

 Open data has to be understood as a process

 Active communication with current, but also to-be data providers to get more data, but especially more interesting data to GovData.de

 Quality of metadata plays a crucial role

– Influences the discoverability and searchability – Needs to be improved constantly

 GovData.de and its metadata schema should not be an isolated application – Schema compatibility with Government Data Austria (data.gv.at)

– DCAT: RDF vocabulary to facilitate interoperability between data catalogs

(13)

Metadata Harvesting Techniques

 CKAN to CKAN

 JSON to CKAN

 ISO 19115 to CKAN

 CKAN API

Références

Documents relatifs

In Section 4, we use a high-performing blocking method called block purging from the original AC paper as a baseline for evaluating the proposed Sorted Neighborhood workflow..

It is also important to remember that due to lack of research value, many open government data are still released alongside other licensed datasets (as is the case with the

A central electronic information and communication system will be developed for storing data on the registration of economic operators and for the exchange of

We introduce a general data exchange (GDE) setting that extends DE settings to allow collaboration at the instance level, using a mapping table M , that specifies for each

representatives of national statistical institutes and national data archives, and Linked Open Data community members have developed the DDI-RDF Discovery

Model The model part of the tool based on our conceptual model [21] consists of classes for each modeled component, such as a class, an association or an attribute on each of

In this paper, we summarise our results on Modelling Dy- namics in Semantic Web Knowledge Graphs published at WWW 2018 where we proposed a novel data-driven schema for graphs and

Fig. 1a shows the RDF dataset publishing process. The requirements of input data are as follows: 1) the dataset must consist of only one table (one spreadsheet),