Supervised Semantic Similarity

Abstract

Semantic similarity between concepts in knowledge graphs is essential for several bioinformatics applications, including the prediction of protein-protein interactions and the discovery of associations between diseases and genes. Although knowledge graphs describe entities in terms of several perspectives (or semantic aspects), state-of-the-art semantic similarity measures are general-purpose. This can represent a challenge since different use cases for the application of semantic similarity may need different similarity perspectives and ultimately depend on expert knowledge for manual fine-tuning. We present a new approach that uses supervised machine learning to tailor aspect-oriented semantic similarity measures to fit a particular view on biological similarity or relatedness. We implement and evaluate it using different combinations of representative semantic similarity measures and machine learning methods with four biological similarity views: protein-protein interaction, protein function similarity, protein sequence similarity and phenotype-based gene similarity. The results demonstrate that our approach outperforms non-supervised methods, producing semantic similarity models that fit different biological perspectives significantly better than the commonly used manual combinations of semantic aspects. Moreover, although black-box machine learning models produce the best results, approaches such as genetic programming and linear regression still produce improved results while generating models that are interpretable.

pyRDF2Vec: A Python Implementation and Extension of RDF2Vec

“…In [10], pyRDF2Vec among many other embedding techniques, has been compared to non-embedding methods to better understand their semantic capabilities. In Sousa et al [20] pyRDF2Vec is used to tailor aspect-oriented semantic similarity measures to fit a particular view on biological similarity or relatedness in protein-protein, protein function similarity, protein sequence similarity and phenotype-based gene similarity tasks. Engleitner et al…”

Refining Diagnosis Paths for Medical Diagnosis based on an Augmented Knowledge Graph

“…When it comes to refining existing medical knowledge graphs, the combination of knowledge graph embeddings and machine learning models, similarly to this paper, has been utilized in the past, e.g., for predicting drug-drug interaction [3,13], gene-disease interaction [19], or other tasks, such as proteinprotein interaction, protein function similarity, protein sequence similarity, and phenotype-based gene similarity [26]. Unlike the work presented in this paper, those approaches mostly use a single in-domain knowledge graph for their predictions, while we present a method using an integrated augmented knowledge graph incorporating numerous types of information.…”

Unsupervised Event Graph Representation and Similarity Learning on Biomedical Literature

“…Third, systems should be interpretable in the process of predicting similarities; in the biomedical domain, more than others, reliability is a requirement that cannot be ignored. Fourth, as is often the case with labeled graphs [ 17 , 18 ], subtle nuances in labels can make two events semantically agreeing or not, while events with distinct structures can still be similar. Therefore, to be broadly useful, a successful solution should consider both event graph structure and semantics , fine-grained (e.g., synonym) and coarse-grained (broad scenario).…”