The type of data collected from the base platforms include data of companies, service providers and products & services offered by them. The base platforms have different data schemas to define these entities. Inorder to capture different attributes of the entities (companies & products/services) these base platform data schemas need to be consolidated into a common schema/ontology.
To enable an effective federated search in EFPF, we designed a common ontology called EFPF Manufacturing Ontology (EFONT) that includes the following concepts (see Figure 1):
- A Class/ Category of a product/ service/ partner’s capability has 0 or more properties;
- A Property describes the product/service class in detail, e.g. length, height, certificates;
- An Item is an instance of a Class/ Category. Each Class/ Category has 1 or more item instances representing the actual product/service or partner’s capability that will be manufactured/ provided by a party/ company;
- A Party has attributes such as a legal-name, keywords and activity sector that extend a variety of attributes for matchmaking processes.
Figure 1 : EFPF Matchmaking Ontology Basic Version
The concepts and their attributes defined above are inspired by the Universal Business Language (UBL) specifications on Supplier PartyType and CatalogueType.
We extended the above ontology with more attributes which will be useful for matchmaking transactions and the extended version of the EFONT is depicted below in Figure 2. The additional relations/attributes were added mainly to the Item and Party concepts by analysing the different schemas used across the base platforms. Some of the main attributes of the concepts are as below.
- Party : Legal name, brand name, business type, origin, address, website, projects, trust score
- Item : Label, manufacturerID, price, packageUnit, packageAmount, totalCapacity, deliveryTime, certificateType
The extended full EFONT ontology is depicted below in Figure 2.
Figure 2 : EFPF Matchmaking Ontology Extended Version