Pairwise similarity is an essential operation in multidisciplinary fields of study. However, its operation is expensive due to its complexity as well as the evolution of big data. Besides, similarity scores without semantics leads to neither accurate nor practical results. In this paper, we study the problem of pairwise similarity in terms of semantics, efficiency, and increment. More concretely, we build our own synonym database and organize it in a reflexive synonym index to support us in semantic-aware pairwise similarity. In addition, we employ cache to perform incremental similarity computing in the process of pair-by-pair similarity search. Moreover, we design our feature-based hierarchy to eliminate irrelevant candidate objects before doing pairwise similarity. Furthermore, we analyze a practical case and apply our method into such social network-based applications in order to demonstrate its feasibility and efficiency in reality.
CITATION STYLE
Phan, T. N. (2019). Efficiently Semantic-Aware Pairwise Similarity: an Applicable Use-Case. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11814 LNCS, pp. 226–238). Springer. https://doi.org/10.1007/978-3-030-35653-8_15
Mendeley helps you to discover research relevant for your work.