“This week I had the pleasure of speaking at the Cambridge Computer Laboratory for the Artificial Intelligence Research Group. The talk was titled “How Can We Complete a Knowledge Graph?”. I spent the afternoon in the department, and the talk of Knowledge Graphs and learning inspired conversations with 3 PhDs working across rule mining for scene understanding, multiple choice question answering and graph pattern-matching via reasoning.
Thanks so much to Mateja Jamnik and Zohreh Shams for inviting me to speak, and also for sharing their own strifes working with complex biomedical data!”
2019 Project Participants
6 students who are currently 2nd year students of Computer Science, divided themselves into two groups to try to implement two different methodologies of automatically testing a query language - Graql.
"Our support is basically answering their questions and guiding them during the implementation. It’s been a pleasure working with them throughout the project and always get a little boost of energy whenever we spend time with them.
The first approach was to find an analogous system serving as a test oracle. The system was chosen to be a particular implementation of Datalog which is closely related to Graql.
The second approach is to test the language against itself using the metamorphic testing approach - we defines a pair or relations that need to hold between corresponding input and output pairs and then by generating large volumes of inputs we verify the respective relation holds between the outputs."