Skyline queries in crowd-enabled databases

Abstract

Skyline queries are a well-established technique for database query personalization and are widely acclaimed for their intuitive query formulation mechanisms. However, when operating on incomplete datasets, skylines queries are severely hampered and often have to resort to highly error-prone heuristics. Unfortunately, incomplete datasets are a frequent phenomenon, especially when datasets are generated automatically using various information extraction or information integration approaches. Here, the recent trend of crowd-enabled databases promises a powerful solution: during query execution, some database operators can be dynamically outsourced to human workers in exchange for monetary compensation, therefore enabling the elicitation of missing values during runtime. Unfortunately, this powerful feature heavily impacts query response times and (monetary) execution costs. In this paper, we present an innovative hybrid approach combining dynamic crowdsourcing with heuristic techniques in order to overcome current limitations. We will show that by assessing the individual risk a tuple poses with respect to the overall result quality, crowdsourcing efforts for eliciting missing values can be narrowly focused on only those tuples that may degenerate the expected quality most strongly. This leads to an algorithm for computing skyline sets on incomplete data with maximum result quality, while optimizing crowd-sourcing costs.

Crowdsourcing for Query Processing onWeb Data: A Case Study on the Skyline Operator

“…This paper mainly summarizes and unifies previous work discussed in [2] and [3], but also provides new insights, especially with respect to processing skylines heuristically. The rest of this paper is organized as follows: in the next section we discuss related work and give an overview of the theoretical foundations of skyline queries.…”

“…This approach, first introduced in [2], surrogates missing values by (potentially unreliable) predictions before a skyline is computed. Tuples that are considered high impact (risk) have their predicted values superseded with real values via crowdsourcing Eventually, the skyline can be computed with high accuracy.…”

“…Different strategies for selecting a suitable k are detailed in [2], where k could be chosen in accordance to quality requirements or additional meta-data provided by the user.…”

“…Moreover, even within a dataset, some attributes can be predicted more accurately than others, e.g., games played in the NBA set has a mean error of ∼ 0.34, while most other attributes of the NBA set have mean errors of ∼ 0.07. Evaluations with other prediction algorithms can be found in [2]. Table 3.…”

“…Instead of naively crowdsourcing all missing data, we aim at selective hybrid approaches which rely on crowdsourcing only those tuples where heuristics will most likely introduce errors, but fall back to using (cheap) heuristics when their application is safe. We present two different strategies, as studied in previous works a) surrogating missing data with predicted values [2], b) surrogating missing data with minmax values [3].…”