SLiDInG6

Purpose

SLiDInG6 aims to bring together organisations from industry, government and academia from around Scotland to share different perspectives around Linked Data, and to inform participants of current research on Linked Data and its uptake in industry and government.

We plan to have flash talks to highlight current areas of activity and deployment of Semantic Web technologies. We then plan to have discussion groups focusing on:

1. What are the killer apps of the Semantic Web?
2. What are the challenges from industry and government?
3. Where are the synergies in Semantic Web research in Scotland?
4. Can we identify projects to push forward for funding?

We hope that the outcome will be the identification of future collaborations and projects that will be developed after the event.

Programme

Live minutes from the event are being collected in this document. Please help by adding your comments and questions into the document.

Lunch, tea, and coffee will be provided. Please let the organisers know by email of any dietary requirements.

Please upload your talks to this folder.

Flash talks: (Draft)

1. Jeff Z. Pan: Knowledge Graph: Enriching Your Data for Learning and Reasoning
2. Roderic Page : Towards a biodiversity knowledge graph
   • One way to think about core biodiversity data is as a network of connected entities, such as species and their names, scientific publications, people, DNA sequences, evolutionary trees, images, and museum collections. Many tasks in biodiversity informatics correspond to tracing paths in this "knowledge graph". Despite repeated calls for a biodiversity knowledge graph, and over a decade of publishing taxonomic data in RDF, the biodiversity informatics community still struggles to achieve the dream of a cross-linked, queryable synthesis of basic biodiversity information. The talk explores some reasons for this.
3. Kemas Rahmat Wiharja: Revisiting the Notion of Correctness in Knowledge Graph Completion
4. Chenxi Li: Using Knowledge Graphs for Content-based Fake News Detection
5. Aba-Sah Dadzie: Using ontologies as a backbone for visual data analytics - brief summary of two use cases
6. Ningxi Li: Knowledge Graph as Common Sense Knowledge for Object Detection and Visual Relation Detection
7. Yangmei LI: CURIOS: A Knowledge Graph Enabled Content Management System
8. Fiona McNeill: Communication during Emergency Management through Data Integration and Trust
   1. I will introduce our work on the CEM-DIT project, looking at how to get data from multiple heterogenous sources to decision makers during crises so that is it is both relevant and understandable (by using semantic matching techniques) and reliable (by analysing the provenance of the data).
9. Diana Bental: CHAIn: Integrating data from heterogenous sources to support decision making in crisis management
   • The CHAIn system is designed to facilitate fast data sharing between disparate organisations during emergency response situations. When queries fail because of incompatibilities between the query and the schema of the queried datasource, CHAIn uses structured semantic matching to reformulate the failed queries dynamically and retrieve responses that are relevant to the failed query. CHAIn allows these organisations to collaborate and respond to queries without extensive data pre-alignment. CHAIn has been applied to relational and RDF data sources. We are now developing CHAIn’s user interface, considering issues of trust and data provenance, and seeking datasets with which to evaluate CHAIn.
10. Kobby K.A. Nuamah: Functional Inference for Query Answering using FRANK
    • In this talk, we discuss FRANK (Functional Reasoning for Acquiring Novel Knowledge), an inference framework, for answering queries using data from heterogeneous web sources. The large amounts of data on the Internet presents a good opportunity to answer queries that go beyond simply looking up facts and returning them. We are, interested in queries for which no suitable answer is readily contained in any available data source. This requires a method that identifies facts needed for inference, searches for them, and then integrates these facts and their diverse formalisms into the local query-specific inference tree. However, such web data is often vague and noisy and therefore requires uncertainty to be handled during the inference. We show that the estimates of uncertainty derived by our method are well-calibrated and correlate with the actual deviations from the true answer.
11. Peter Winstanley: DCAT Update Activity
    • The W3C Data Exchange Working Group are in the process of revising DCAT to address the short comings identified with the initial version. This talk will provide an update on this process and the progress on the new version of DCAT.
12. Oluwaseun Bamgboye: Semantic approach in dealing with real life sensor data
    • The ubiquitous environment including the smart spaces is often characterised with deployment of multimodal or heterogeneous sensors and related technologies. These sensors are capable of producing insights about the environment they are deployed while also supporting certain automated applications. However, there are instances in which these sensor readings are delivered to consuming applications such as safety critical and decision support systems with missing data points in a spatial-temporal settings. To be able to deal with this problem, the semantic modelling and reasoning technique is considered appropriate for this task.
13. Alasdair Gray: FAIR data: Successes and Challenges
    • The FAIR data principles aim to make data reusable beyond the original publishers of the data. In this talk I will discuss various projects I'm involved with and the challenges they raise for publishing FAIR data.
14. Xue Li: Repair faulty Datalog-like theories by combining belief revision, abduction and reformation
    • These three techniques are complementary, but they have not previously been combined. Abduction adds the axioms that representing the explanation for observed assertions, and belief revision deletes the unwanted ones. Both of them change an axiom as a whole, while reformation changes the language of the theory. This project aims to establish a framework for aligning these three techniques in one system for repairing Datalog-like logical theories.
15. Liam Cavin: Scottish Government Statistics
16. Ric Roberts: Querying RDF without SPARQL
    • How can we allow users to query RDF without needing to learn SPARQL?
17. Ric Roberts: Linked Data ETL
    • How do you make well structured linked data without knowing anything about RDF?

When and Where

• Tuesday 29th May 2018
• 10:00 – 16:30
• 2.33 Earl Mountbatten Building, Heriot-Watt University, Edinburgh, EH14 4AS

Getting to Heriot-Watt by public transport:

By train: The closest stations to Heriot-Watt are Curriehill (where some trains to/from Glasgow Central stop), Edinburgh Park (where trains to/from the west, including some Glasgow trains, and Stirling trains stop) and Edinburgh Gateway (where trains to/from the north, including Aberdeen, Inverness and Perth trains stop). There is no public transport from any of these stations to the university, so you will need to take a taxi (5-10 minute journey) from the station. You may be able to pick one up at the station, but it is advisable to book one in advance (see below).

If your train doesn’t stop at any of these stations, you can get off at Haymarket and take the 25 bus from there. If you are coming from the south, you can get the 25, 34 or 45 from near Waverley.

By bus: Heriot-Watt is served by numerous buses. 25, 34 and 45 serve the route between the city centre and the university, with the 25 passing Haymarket Station and all three stopping near Waverley station. The cost is £1.70 per trip, and you need to have the exact fare.

By tram: You can get the tram to Edinburgh Park but would need a taxi to take you on from there.

Taxi:

• City Cabs: 0131 228 1211;
• Central Taxis: 0131 229 2468;
• Capital Cars: 0131 777 7777.