IPLodB Aligning

Large amounts of semantic data are available in LOD format(s) in many fields, including that of IP. However, to realize the fusion of multi-source data as needed is a non-trivial task connected to alignment of data between the databases .

Aligning data requires a diligent disambiguation process. The disambiguation aims at identifying and connecting same real-life entities (persons and organizations). Results below are generated following the IP LODB disambiguation and alignment process. The alignment exploration of linked open data (LOD) datasets focuses on potential EP LOD-compatible datasets. Our initial focal sources are thus the EP LOD, with information on patents, and the SN SciGraph data, with information on scientific publications.

The disambiguation process for entities consisted of several main steps: cleaning and parsing, blocking, creation of entity cards, similarity profile, computing a priori matches and merging of identity cards and correcting pairwise estimates associated with triplet violations.

Unique Person Card Example

Search persons...


Oberländer, Kai Daniel jSON

IPLodB ID: 95017a92-710a-4d6a-8fcf-b5419bbe14e5
Date Aligned: 07. 02. 2020

Collected from entities:

Unique Person Map Example