HyperCausalLP: Causal Link Prediction using Hyper-Relational Knowledge Graph

TL;DR

Results show that the inclusion of knowledge about mediators in causal link prediction using hyper-relational knowledge graph improves the performance on an average by 5.94% mean reciprocal rank.

Abstract

Causal networks are often incomplete with missing causal links. This is due to various issues, such as missing observation data. Recent approaches to the issue of incomplete causal networks have used knowledge graph link prediction methods to find the missing links. In the causal link A causes B causes C, the influence of A to C is influenced by B which is known as a mediator. Existing approaches using knowledge graph link prediction do not consider these mediated causal links. This paper presents HyperCausalLP, an approach designed to find missing causal links within a causal network with the help of mediator links. The problem of missing links is formulated as a hyper-relational knowledge graph completion. The approach uses a knowledge graph link prediction model trained on a hyper-relational knowledge graph with the mediators. The approach is evaluated on a causal benchmark dataset, CLEVRER-Humans. Results show that the inclusion of knowledge about mediators in causal link prediction using hyper-relational knowledge graph improves the performance on an average by 5.94% mean reciprocal rank.

Figures (6)

• Figure 1: Causal link. (A) A serial causal connection where A causes B and eventually B causes C. The node A is known as a cause, C is known as an effect, and B is known as a mediator. (B) A serial causal link, the link is encoded as a knowledge graph link using RDF format, (C) Causal link as a hyper-relational link where the mediator entity is represented a hyper-relation with the hyper-relation predicate, hasMediator. The link is encoded as a knowledge graph link using RDF-Star format
• Figure 2: HyperCausalLP has four primary phases: 1) encoding the causal associations in data as a causal network, 2) translating the causal network into a causal knowledge graph, 3) learning knowledge graph embeddings (CausalKG-Base and hyper-relational graph based embedding CausalKG-M with mediators as hyper-relations) from the causal knowledge graph, and 4) using the knowledge graph embeddings for causal link prediction tasks.
• Figure 3: The figure shows reified causal relations, causesType and causedByType. The causedByType is a reified relation from an effect-entity instance to the type of a cause-entity. The causesType is a reified relation from a cause-entity instance to the type of an effect-entity. It also illustrates the two qualifier relations associated with causes relation: hasMediator and hasMediatorType. The qualifier relations are also associated with the causedBy relation, which is an inverse of the causes relation.
• Figure 4: StarE encoder, which encodes a hyper-relations for the causal relation StarEGalkin2020message. The hyper-relation qualifier pairs (or mediator pairs) are passed through a composition function $\phi _q$, which are summed together and transformed by weights $W_q$. The transformed vector is merged with $\gamma$ and $\phi _r$. The final node i.e. cause entity combines messages from all the hyper-relation. [Note: As specified in StarE- 1) $\phi$ is a composition function of a node with its respective relation, 2) $W_{\gamma(r)}$ is a direction-specific shared parameter for outgoing, incoming, and self-looping relations, 3) $\gamma$ is a function that combines the main relation, ($r_c$) representation with the representation of its qualifiers, ($Q$)
• Figure 5: A snapshot of the CausalKG-Base and CausalKG-M representation. (A) A snapshot of collision events in a video at time t-1, t, and t+1 from the CLEVRER-Humans. There are three consecutive collision events that occur: A: the red cube collides with the yellow ball, B: the yellow ball hits the blue cylinder, and C: the blue cylinder moves. The A, B, C are causal entities. A.Type is Collide, B.Type is Hit, and C.Type is Move. (B) The causal event graph of the above snapshot. (C) The causal and mediator (qualifier pairs) links representation in the two different CausalKG.