Categorical skylines for streaming data

Abstract

The problem of skyline computation has attracted considerable research attention. In the categorical domain the problem becomes more complicated, primarily due to the partially-ordered nature of the attributes of tuples.In this paper, we initiate a study of streaming categorical skylines. We identify the limitations of existing work for offline categorical skyline computation and realize novel techniques for the problem of maintaining the skyline of categorical data in a streaming environment. In particular, we develop a lightweight data structure for indexing the tuples in the streaming buffer, that can gracefully adapt to tuples with many attributes and partially ordered domains of any size and complexity. Additionally, our study of the dominance relation in the dual space allows us to utilize geometric arrangements in order to index the categorical skyline and efficiently evaluate dominance queries. Lastly, a thorough experimental study evaluates the efficiency of the proposed techniques.

Understanding the meaning of a shifted sky: a general framework on extending skyline query

“…Another instance employing the traditional skyline is the recent k most representative skyline operator [16] [12,9] ---+ + k-dominance [6] + + -+ + Table 3 Revisiting existing dominance relationships introduces a constraint on the number of skyline points to be returned, and selects from the traditional skyline those points that maximize the total number of dominated points. The skyline query on partially ordered domains [5] and categorical domain [20,17] are special cases of the traditional skyline query, but the properties of scaling robustness or shifting robustness are not applicable because partially ordered domains do not support scaling or shifting operations.…”

Efficient Skyline Maintenance for Streaming Data with Partially-Ordered Domains

“…Recent research effort has shifted towards query processing in online environment, where a skyline result is dynamically maintained for a long-standing skyline query over continuous streaming data [10,12].…”

“…Thus, the skyline is computed over all non-expired tuples and is updated whenever a new tuple arrives or an existing tuple expires [12]. In the count-based window model, the skyline is maintained for the most recent N tuples [10]. Thus, the skyline is updated whenever a new tuple arrives, and the arrival of the new tuple may also cause the oldest existing tuple to expire if there are already N tuples before the new arrival.…”

“…Several recent work on skyline queries have broadened the scope to include categorical attributes with partially-ordered domains in both offline [2,9] as well as online [10] environment. Categorical attributes are more general than numerical attributes as the dominance relationships among the domain values for a categorical attribute are based on a partial ordering instead of a total ordering.…”

“…In this problem, a fixed-size buffer of the N most recent tuples is dynamically maintained and the key challenge is how to efficiently maintain the skyline of the sliding window of N tuples as new tuples arrive and old tuples expire. The state-of-the-art approach for this skyline problem is the STARS method [10], which is based on two key data structures: a multi-dimensional grid to organize the tuples in the buffer, and a geometric arrangement structure to organize the skyline tuples.…”