US 11,941,056 B2
Method for weighting a graph
Martin Oberhofer, Sindelfingen (DE); Mike W. Grasselt, Leinfelden-Echterdingen (DE); Claudio Andrea Fanconi, Celerina (CH); Thuany Karoline Stuart, Nice (FR); Yannick Saillet, Stuttgart (DE); Basem Elasioty, Regensburg (DE); Hemanth Kumar Babu, Böblingen (DE); and Robert Kern, Karlsruhe (DE)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Apr. 20, 2021, as Appl. No. 17/234,835.
Claims priority of application No. 20171979 (EP), filed on Apr. 29, 2020.
Prior Publication US 2021/0342397 A1, Nov. 4, 2021
Int. Cl. G06F 16/901 (2019.01); G06F 16/36 (2019.01)
CPC G06F 16/9024 (2019.01) [G06F 16/367 (2019.01)] 12 Claims
OG exemplary drawing
 
1. A computer implemented method for pre-processing a graph comprising:
obtaining a graph comprising nodes representing entities and edges representing relationships between entities in accordance with one or more domains;
for each edge in the graph, determining whether a data quality problem associated with a specific domain of the one or more domains exists between one or more attributes of the edge and one or more attributes of the node connected to the edge, wherein the determining is based on whether the one or more edge attributes and the one or more node attributes fulfill a set of ontology rules descriptive of an ontology of the specific domain;
pre-processing the graph by assigning a weight to each of the nodes and each of the edges in the graph based on the determined data quality problems, the weight of each node or edge indicating whether there is a domain specific data quality problem with one or more attributes of that node or edge;
processing requests according to the assigned weights in the pre-processed graph to reduce an impact of the determined data quality problems;
receiving a request from a database matching engine, the request indicating the specific domain; and
in response to receiving the request, performing the pre-processing, and integrating the pre-processed graph in the database matching engine; and
utilizing the pre-processed graph as a basis for a matching decision.