Promotional subspace mining with EProbe framework

Abstract

In multidimensional data, Promotional Subspace Mining (PSM) aims to find out outstanding subspaces for a given object, and to discover meaningful rules from them. In PSM, one major research issue is to produce top subspaces efficiently given a predefined subspace ranking measure. A common approach is to achieve an exact solution, which searches through the entire subspace search space and evaluate the target object's rank in every subspace, assisted with possible pruning strategies. In this paper, we propose EProbe, an Efficient Subspace Probing framework. This novel framework strives to initialize the idea of "early stop" of the top subspace search process. The essential goal is to provide a scalable, costeffective, and flexible solution where its accuracy can be traded with the efficiency using adjustable parameters. This framework is especially useful when the computation resources are insufficient and only a limited number of candidate subspaces can be evaluated. As a first attempt to seek solutions under EProbe framework, we propose two novel algorithms SRatio and SlidingCluster. In our experiments, we illustrate that these two algorithms could produce a more effective subspace traversal order. Being effective, the topk subspaces included in the final results are shown to be evaluated in the early stage of the subspace traversal process.

Maximizing a Record’s Standing in a Relation

“…Although the materialization and exploration approaches of [16] and [15] can directly be used to solve SMP, they have very high storage requirements and their cost is extremely high, as we demonstrate in Section 4. A variant of the problem, where only a fixed number of subspaces is considered instead of the whole set of subspaces, is studied in [17]: an approximate solution to the original problem is derived (considering binary attributes only). All solutions above explore a large number of subspaces, thus they have a high cost.…”

“…Since (v 5 , v 7 ] contains fewer objects than the support threshold, we stop the enumeration of left endpoints for p 7 , and move on to the calculation of the best slope for segments ending at p 7 (lines [11][12][13][14][15][16][17]. The line with the best slope (50%) is p 3 p 7 and the corresponding range (v 3 , v 7 ] is recorded as the current best for A.…”

Maximizing a record's standing in a relation

“…Thus, p 4 is simply enqueued into Q, i.e., Q = {p 0 , p 3 , p 4 }. Since (v 5 , v 7 ] contains fewer objects than the support threshold, we stop the enumeration of left endpoints for p 7 , and move on to the calculation of the best slope for segments ending at p 7 (lines [11][12][13][14][15][16][17]. The line with the best slope (50%) is p 3 p 7 and the corresponding range…”