Winners 2011

Winners 2011

1. Food and Agriculture Organization of the United Nations, Italy
Team contact: Martin Kaltenböck, SWCG, Austria

Abstract:

The LOD2 PubLink project at the Food and Agricultural Organization of the United Nations (FAO) located in Rome, Italy is managed by the Semantic Web Company (SWCG) and supported by the Free University of Berlin (FUB) and the University of Leipzig (ULEI). The project’s focus is on issues around the data.fao.org Data portal. The issues that are tackled within the project are of three kinds:

A) Know-How Transfer in Linked Open Data (LOD)

FAO seeks to learn more about the Open Data movement and about the movement, principles and technologies of Linked Open Data (LOD). This issue of know-how transfer is managed by LOD2 mainly through webinars on special topics in which FAO is particularly interested (e.g. provenance, open data governance, etc.).

B) Management of Equivalent Entities

The management of equivalent entities in the FAO portal (e.g. apple), as well as conceptually broader/narrower entities, will rely partially on a batch discovery process and partially on crowd sourcing efforts (i.e. user contribution). In both cases the FAO data portal will benefit from a mechanism to keep track of the equivalence and hierarchical relationships formulated during the process. Especially when the users propose any of the three relationships, the mechanism needs to keep also track of the provenance (i.e. who did it) for the sake of generating social surplus value.

In this respect, FAO has decided that the LOD2 mini-project has to deliver an RDF model that captures these requirements. After the model is populated, i.e. evolving while the users provide their input, it will be the basis to feed the portal with entities related in terms of equivalence or in terms of a narrower or broader order. The model is also basis for a scoring mechanism that captures the strength of the connection among terms as the user reviews it (e.g. G+1, like), as well as for a provenance mechanism that informs who (in your network) has approved the relationship.

C) Modelling Landing Pages

The LOD2 PubLink mini-project can also be pilot work for modelling the Open RDF of one landing page category. FAO has decided not to take country as an example as it provides little room for action and hence for success; FAO opted for the category of concepts. Even if the Open RDF model for a landing page is independent from the current RDF being produced for the FAO catalogue, it will have to take into consideration the same constraints of description currently observed in the catalogue.

The result of the LOD2 PubLink mini-project will thus be a model that makes use of well-known/standard (de-facto) vocabularies as proposed by the experienced LOD2 project participants, and include only a portion of the data currently in the catalogue.

2. European Environment Agency, Denmark
Team contact: Bastiaan Deblieck, Tenforce, Belgium

Abstract:

The EEA, European Environment Agency, is an EU agency responsible for aggregating and publishing information on the environment. Main clients are the European Union institutions - the European Commission, the European Parliament, the Council — and the EU member states

LOD2 has been interacting with the “IT networking and public systems” team. This team has been publishing open data for quite some time already and is up-to-speed with regards to triple stores and metadata management. (See: http://www.eea.europa.eu/data-and-maps/data) However, they still face challenges in the area of legacy databases and the ETL Process. Several meetings took place and experience and expertise have been exchanged. Below is a summary of the outcomes:

  • PubLink intensified the contact between LOD2/LATC and a public agency.
  • LOD2/LATC provided information on and discussed how a triple store could replace/integrate with existing legacy databases.
  • LOD2/LATC provided information on and discussed how technology from the LOD2 stack can support the ETL process. (See: http://www.semantic-web.at/linked-data-manager for more info on the technology)

PubLink helped the EEA to refine its strategic course by challenging their decisions and assumptions. As a result the EEA was well prepared for delivering data to the EC Open Data Portal.

 

3. Birmingham City Council, Great Britain
Team contact: Hugh Williams, OGL, Great Britain

Abstract:

The Digital Birmingham project (http://www.digitalbirmingham.co.uk/) was established in 2006 as the driving force to ensure that the benefits of digital and emerging technologies are available to all in the city. As part of this initiative the council is currently investigating the procurement of an open data platform for the city to encourage its reuse further, in line with their open data policy. The Birmingham Civic Dashboard (http://civicdashboard.org.uk/) is one such service that receives a report of the requests that come in from members of the public for services from Birmingham City Council each day, with the data currently being available in CSV and JSON formats. It then produces a number of visualizations of that data, such as showing them on a map.

The main goal of the engagement with the PubLink project is to assist in devising a structure and strategy for publishing this data as Linked Open Data, starting with the civic dashboard site. A key requirement is to publish the daily updates the site receives, for which an initial sample static dataset has been sponged and published at: http://demo.openlinksw.com/c/ZCSJG

4. Municipality of Udine, Italy
Team contact: Sebastian Hellmann, ULEI, Germany

Abstract:

The Municipality of Udine is a local institution including 950 human resources, 900 PC workstations and an internal ICT system management. One of the main goals of the administration is to stimulate citizens’ active participation in the decision-making process, which is why many e-Government solutions are promoted such as open databases, web 2.0 applications or wikis. Their implementation is to help introducing new ways of working, a new language for administrative proceedings, and innovative tools to avoid inefficiency, to evaluate the municipal services, to propose ideas and to offer online services.

The main goal of the PUBLINK project is to provide a better structure of the open data sources, to find relationships between them and finally to publish them as Linked Open Data. In detail, the focus is on the datasets belonging to the spendings of the Municipality of Udine. Apart from a specifically developed visualization of the data, it would be possible to introduce this data into projects like http://openspending.org. In the end, these efforts will lead to more transparency for the citizens.

Currently, the first data sources have been translated into RDF and will now be published on top of OntoWiki.

5. Statistical Office of the Republic of Serbia, Serbia
Team contact: Valentina Janev, IMP, Serbia

Abstract:

The Statistical Office of the Republic of Serbia (SORS) is the main producer of official statistics in Serbia, publishing a wide range of documents including press releases, monthly statistical bulletin, quarterly reports on different statistical indices, statistical yearbook, working documents, methodologies and standards, trends, etc.  While most of the publications are free of charge and downloadable in PDF format, some of them are accompanied with their raw data that is stored in an MS SQL-based dissemination database.

With the aim of achieving better transparency and improving the interoperability of data with similar governmental agencies, in the framework of the LOD2/PUBLINK activities, the SORS started to publish the data trapped in the dissemination database in RDF, a machine-processable format. The final goal is to make the reports interlinked in the LOD2 cloud via the PublicData.eu portal. Publication of the SORS reports as RDF using standard vocabularies such as the RDF Data Cube vocabulary is crucial for SORS, as the published data is compatible with the cube model that underlies SDMX (Statistical Data and Metadata eXchange), an ISO standard for exchanging and sharing statistical data and metadata among organizations including EUROSTAT.

You can see the result of the support activities by LOD2/PUBLINK here: http://rs.ckan.net/group/rzs

6. Bezirksverordnetenversammlung Berlin-Kreuzberg, Germany
Team contact: Sebastian Hellmann, ULEI, Germany

Abstract:

Friedrichshain-Kreuzberg is a district in central Berlin comprising parts of both former East and West Berlin. It numbers 271.823 inhabitants and is characterized by its diverse population with outstanding creativity, subculture, and progressive politics. The Bezirksverordnetenversammlung (BVV) is less a classical parliament rather than a body of public administration, which is a hybrid between the legislative and executive powers. Much of the BVV’s work is at the administrative level. Citizens interact with the BVV through the deputies, who channel 'Anfragen' (requests) and 'Anträge' (applications) in the form of 'Drucksachen' (printing matters) to the BVV. The documents detailing the requests (or 'Drucksachen') are electronic, but they are not machine-readable, have completely intransparent filenames and are generally unusable to the extent that the deputies themselves prefer to look up hard copies in the physical archive.

The data we aim to publish is the complete record of requests and answers as they emerge. The data is mostly of textual nature, but graphical items like maps are also found. Data frequently includes references to recurring entities such as laws, administrative and geographical entities and the like. These recurring entities will be identified and interlinked, thus allowing an easier exploration of the data. Interlinking based on these entities will facilitate data mining by interested parties and will furthermore enhance democratic control of the political parties. The timely publication of these datasets will allow to identify and reconcile divergent interests in the very early stages of public projects and help to satisfy the growing need for participation in a democratic local society.


LOD2 Demos