Semantic Models in practice A how to guide

Semantic Models in practice A “how to guide” Living Labs meeting 23 February 2021

A story in 3 acts I. Semantic Web a. k. a. Federated Network of Platforms? II. The world we know (and the other one) III. The Round Trip (but will we get home? )

Semantic web a. k. a. Federated network of platforms…?

Where data is heading

Where database technology is heading ?

Graph database is just a database, but bigger SQL SELECT xx FROM yy SPARQL SELECT xx WHERE yy

Federated network of platforms ≈ Linked Data ≈ Graph Forwarder Road carrier Customs Dangerous Goods packing

What do we need for this graph? • A common semantic model and an ontology • so we can understand each other’s language and the meaning • We need linked data to create the network • use URI’s in our data • We need a common syntax • Use RDF (Resource Description Framework) which has different interchangeable syntaxes like XML, JSON-LD, TTL (‘Turtle’) • OWL (Web Ontology Language) for more complex semantics • SHACL for managing data constraints

Starting from the world we know Let’s face it, it all started with clay…

Early messaging and documents (Babylon, 18 th century BC) Now, when you had come, you spoke saying thus: 'I will give good ingots to Gimil. Sin'; this you said to me when you had come, but you have not done it. You have offered bad ingots to my messenger, saying 'If you will take it, take it; if you will not take it, go away. ' Who am I that you are treating me in this manner -- treating me with such contempt? and that between gentlemen such as we are. I have written to you to receive my money, but you have neglected [to return] it. Repeatedly you have made them [messengers] return to me empty-handed through foreign country. Who is there amongst the Dilmun traders who has acted against me in this way? You have treated my messenger with contempt. And further with regard to the silver that you have taken with you from my house you make this discussion. And on your behalf I gave 18 talents of copper to the palace, and Sumi-abum also gave 18 talents of copper, apart from the fact that we issued the sealed document to the temple of Samas. With regard to that copper, as you have treated me, you have held back my money in a foreign territory, although you are obligated to hand it over to me intact. You will learn that here in Ur I will not accept from you copper that is not good. In my house, I will choose and take the ingots one by one. Because you have treated me with contempt, I shall exercise against you my right of selecting the copper.

Paper: 100% interoperable Where does the data come from?

Where does the data come from? Electronic Documents GUI Corporate System

We’re back to a database But where does a database come from?

You have to choose: red or blue pill…? Show me the Matrix I don’t want to know. I just want to get and send data and do my job

Blue pill: just give me the data Business as usual: schemas and electronic documents

Red pill: data is everything • Data schemas are derived from semantic models • Semantic models are derived from the ontology of logistics & transport, i. e. , the knowledge we have of our domain that allows us to reason about this. • This knowledge is expressed in “knowledge graphs”. • Using tools we can query that knowledge and reason and make inferences. We can ask questions like: “Is there a dangerous goods packing service near Dortmund that can handle vaccines? ” directly from the knowledge graph. • Sorry, you took the red pill…

Physical Infrastructure Logistics Ontology Digital Twin Network of 5 ontology modules Product Node / Hub / Place Business and Compliance Business Service Event Digital Twin Physical infrastructure Business and compliance Cargo Legal or natural Person Event Classification Equipment Custom item Transport Means Event

Digital Twin Physical Infrastructure Digital. Twin_detail Event Physical. Infrastructure_detail Classifications Service_detail Classification Business and Compliance

Ontology in action Data: Container type x is loaded on the vessel name x in the port name x.

Design aspects of the model • Open world assumption – extensibility to future data sharing requirements • Standards perspective • Standards of different modalities as input to the model • Standards perspective • Support of all types of code lists (UN Recommendations, ISO code lists, • etc. ) Standards of different modalities as input to the model • Support of different standards and technologies as data carriers (e. g. APIs, JSON(-LD), • messaging, Support of. XML, all types of RDF) code lists (UN Recommendations, ISO code lists, etc. ) • Support of different standards and technologies as data carriers (e. g. messaging, APIs, XML, JSON(-LD), RDF) • Alignment with different models (e. g. IATA One Record, WCO, UN CEFACT) HM

Structure of the model Technology Structure Knowledge graph – concepts with their data properties Shape graph – value constraints, rules HM 01/01/2022

Modeling examples Logistics. Roles: Roles Logistics. Roles: Financial. Roles Logistics. Roles: Commercial. Roles Logistics. Roles: Roles Logistics. Roles: Business. Roles Logistics. Roles: Financial. Roles Logistics. Roles: Payer Logistics. Roles: Roles Logistics. Roles: Financial. Roles Logistics. Roles: Payee Logistics. Roles: Roles Logistics. Roles: Financial. Roles Logistics. Roles: Insurer Logistics. Roles: Roles Logistics. Roles: Financial. Roles Logistics. Roles: Bank Logistics. Roles: Roles Logistics. Roles: Commercial. Roles Logistics. Roles: Contact. Person Logistics. Roles: Roles Logistics. Roles: Commercial. Roles Logistics. Roles: Service. Provider Logistics. Roles: Roles Logistics. Roles: Commercial. Roles Logistics. Roles: Customer Logistics. Roles: Roles Logistics. Roles: Commercial. Roles Logistics. Roles: Authority Logistics. Roles: Roles Logistics. Roles: Dispatch. Party Logistics. Roles: Roles Logistics. Roles: Carrier Logistics. Roles: Roles Logistics. Roles: Notify. Party Logistics. Roles: Roles Logistics. Roles: Operator. Roles Logistics. Roles: Delivery. Party Logistics. Roles: Roles Logistics. Roles: Consignee Logistics. Roles: Roles Logistics. Roles: Logistic. Service. Provider_ Logistics. Roles: Roles Logistics. Roles: Forwarder Logistics. Roles: Roles Logistics. Roles: Shunting. Yard Logistics. Roles: Roles Logistics. Roles: Consignor Logistics. Roles: Roles Logistics. Roles: Business. Roles Logistics. Roles: Issuer Logistics. Roles: Roles Logistics. Roles: Business. Roles Logistics. Roles: Importer Logistics. Roles: Roles Logistics. Roles: Business. Roles Logistics. Roles: Exporter Logistics. Roles: Roles Logistics. Roles: Business. Roles Logistics. Roles: Declerant Concepts and structure Digital. Twin: Cargo Digital. Twin: Dangerous Digital. Twin: package. Level. Type xsd: boolean xsd: string Digital. Twin: Cargo Digital. Twin: natureof. Cargo Digital. Twin: Modularity classifications: nature. Of. Cargo xsd: integer Digital. Twin: Cargo Digital. Twin: Excise Digital. Twin: Number. Of. TEU Digital. Twin: cargo. Temperature Digital. Twin: cargo. Maximum. Volume xsd: boolean xsd: integer xsd: string xsd: integer Digital. Twin: Cargo Digital. Twin: liquid. Bulk xsd: boolean Digital. Twin: Cargo Digital. Twin: is. Empty Digital. Twin: cargo. Max. Size Digital. Twin: Goods. Description xsd: boolean xsd: integer xsd: string Digital. Twin: Cargo Digital. Twin: cargo. Min. Weight Digital. Twin: location_detail Digital. Twin: cargo. Minimum. Size Digital. Twin: cargo. Minimum. Temperature xsd: decimal pi: Location xsd: integer xsd: string Digital. Twin: Cargo Digital. Twin: Reefer xsd: boolean Digital. Twin: Cargo Digital. Twin: Waste xsd: boolean Digital. Twin: Cargo Digital. Twin: type. Of. Cargo classifications: cargo. Type Digital. Twin: Cargo Digital. Twin: dry. Bulk xsd: boolean Data Properties Rules WH

e. FTI support is similar to support by UN CEFACT MMT WH

All that just to exchange data? No…! • A semantic model or ontology allows us to interact with other ontologies. • Logistics with logistics • Federated network of platforms • Logistics subdomains with other subdomains • maritime and road and rail and air and inland waterways • FEDe. RATED and DTLF and UNCEFACT and WCO and ….

All that just to exchange data? No…! • A semantic model or ontology allows us to automate the data infrastructure: • Generate APIs and associated documentation • Populate search engines and indexes. • Generate class models and other system components • Build mobility Data Spaces • Build industry wide AI services • Build … the Matrix!

No Matrix without awesome tools • Editing semantic models and ontologies: tools like Protégé • Visualizers such as Webvowl but better • Ontology browsers such Widoco • Query languages such as Sparql of Graphql

The Round Trip Go back to where you started and end up at the same place

s Thank IMO - Maritime UN/CEFACT - Example Road L! E I R CEF IATA - Air DTLF – SG 1 EFTI

s Thank L! E I R CEF IATA - Air IMO - Maritime FEDe. RATED Semantic Model UN/CEFACT - Example Road DTLF – SG 1 EFTI

s Thank e. CMR the Round Trip e. AWB L! E I R CEF

Your mission Take the blue pill for now… … and take the red pill later!

Build your Living Labs and make sure you are in control of your schema’s, in particular for the exchange of data. Be ready to change your schema to a schema generated by the FEDe. RATED ontology. Learn about the semantic web. Look into RDF and JSONLD. Form an idea about data spaces.

Semantic Models in practice A “how to guide” Living Labs meeting 23 February 2021