Concept of Relational Similarity Search

For decades, the success of the similarity search has been based on a detailed quantification of pairwise similarity of objects. Currently, the search features have become much more precise but also bulkier, and the similarity computations more time-consuming. While the k nearest neighbours (kNN) search dominates the real-life applications, we claim that it is principally free of a need for precise similarity quantifications. Based on the well-known fact that a selection of the most similar alternative out of several options is a much easier task than deciding the absolute similarity scores, we propose the search based on an epistemologically simpler concept of relational similarity. Having arbitrary objects q, o(1) , o(2) from the search domain, the kNN search is solvable just by the ability to choose the more similar object to q out of o(1), o(2) - the decision can also contain a neutral option. We formalise such searching and discuss its advantages concerning similarity quantifications, namely its efficiency and robustness. We also propose a pioneering implementation of the relational similarity search for the Euclidean spaces and report its extreme filtering power in comparison with 3 contemporary techniques.