Through mCLOUD, the Federal Ministry of Transport and Digital Infrastructure (BMVI) provides open data on a central platform, and from this a variety of innovative projects and companies have already benefited. However, the open data has its weaknesses. These include the fact that data sets are often provided and viewed as monolithic and independent of each other. This makes searching for relevant data difficult, let alone using it in innovative, data-driven applications. The OPAL project seeks to change that.
The OPAL project
OPAL strives to research and develop an integrated portal for open data, using existing open data from mCLOUD and MDM. Unlike the majority of existing open data portals, OPAL will refine metadata and transform it into 5-Star Linked Open Data. The portal will thus ensure that data records can be easily found and accessed by both people and software agents, e. g., other data portals, data-driven applications, etc. In addition, OPAL will also enable the search via content facets (such as automatically generated topics, location information and time constraints) instead of just metadata.
OPAL’s unique selling points are: 1) its technical basis in the form of linked open data technologies, which makes it possible to connect individual currently independent data sets to each other. 2) Its automatic extraction of metadata by machine learning and focused crawling, necessary for this purpose, enables an extensive search for content and metadata. 3) This is also the basis for innovative search functionality such as Question Answering.
Search and further use cases
In addition to the search function integrated in the Web portal, a mobile app and a social bot are to be implemented as further use cases that indicate suitable data records. Close cooperation with the LIMBO project is planned to identify additional metadata on the contents of the data sets, such as topics and schema properties.
The University of Paderborn carries out the OPAL project and is represented by the Data Science Group, which has emerged from the AKSW Group. Its approximately 50 scientists establish theoretical results and scalable implementations for the realization of the Semantic Data Web under the leadership of Prof. Dr. Axel Ngonga. They pay particular attention to the fields of knowledge extraction, information search and knowledge and data integration in the Linked Data Web.