| CPC G06F 16/2455 (2019.01) [G06F 16/285 (2019.01)] | 9 Claims |
|
1. A filtered search method, for performing a search in a data set, and the data set comprises a plurality of data points, and the filtered search method comprising:
dividing the data set into a plurality of clusters based on a similarity of the data points;
dividing each of the clusters into an inlier part and an outlier part based on a distribution density of the data points;
performing a coarse search on all of the inlier parts, to filter out the inlier parts of the first candidate number;
designating a representative point in each of the clusters;
calculating a first distance between each of the representative points and a target point;
obtaining a difficulty level of each of the clusters for the search based on each of the first distances, wherein each of the difficulty levels of the clusters is a value positively related to the first distance;
setting a first candidate number and a second candidate number based on the difficulty level of each of the clusters;
performing a fine search on the inlier parts of the first candidate number, to search the data points of the second candidate number; and
obtaining a search result used for applications of similarity search based on the data points of the second candidate number;
wherein the data points of the second candidate number represented by the search result are close to the target point.
|