Peter Neubauer

In diesem Vortrag wird auf die Daten-Pipeline des Mapillary-Projektes eingegangen. Es werden auch die unterschiedlichen FOSS-Komponenten beleuchtet (OpenSfM, MapillaryJS, iOS-SDK, Python-Tools, JOSM/iD-Integrationen).

Es soll in diesem Vortrag darum gehen, die FOSSS-Komponenten vorzustellen, die auch außerhalb des Mapillary-Projekts genutzt, geändert und integriert werden können. Dies umfasst die APIs, Vektortiles und Exportformate bis hin zu völlig freistehenden Projekten wie MapillaryJS und OpenSfM. Auch die neuen Android- und iOS SDKs werden kurz beleuchtet, zusammen mit den Python-basierten Skripts unter Mapillary-Tools. Als Integrationspunkte werden QGIS und OpenStreetMap, insbesondere der iD-Editor und JOSM, eingesetzt.

In this talk, Peter will describe the implementation and working of http://gist.neo4j.org. It is based on ASCIIDOC, Opal.js, Heroku and Neo4j and rendered all client side. Also, Peter will show some of the examples community members have been contributing - everything from Chess play graphs to product configurations.

Peter Neubauer presents the LDBC EU-project which is setting out to provide comparative graph-database performance tests for both property-graph and RDF databases/triple-stores.

The LDBC benchmarks will span four main areas of Linked Open Data management: complex query execution, transactionality in graphs, RDF inference and RDF support for ETL/data integration. They will:

target hard problems and choke-points that mostly affect users and hence slow down the uptake of RDF and graph database technology, encourage innovative performance and scalability improvements that directly benefit users, be open, community generated, liberally licensed (open-source/creative commons) and target real-world usage scenario, become the de facto standard for publishing performance results and enable objective judgements about the performance and functionality of competing vendor offerings. points that mostly affect users The LDBC Foundation will work in the same spirit as the Transaction Processing Council (TPC) that has established a widely accepted by the industry, set of benchmarks for relational database management systems.It will be responsible for:

• specifying benchmarks, benchmarking procedures and verifying/publishing results,
• providing a TPC-style auditing service for certifying results published by vendors for benchmarks endorsed by LDBC,
• training auditors for its benchmarking, creating a long lasting business model for auditing benchmark results.