Download now Free registration required
The authors formulate three intuitive semantic properties for top-k queries in probabilistic databases, and propose Global-Topk query semantics which satisfies all of them. They provide a dynamic programming algorithm to evaluate top-k queries under Global-Topk semantics in simple probabilistic relations. For general probabilistic relations, the authors show a polynomial reduction to the simple case. Their analysis shows that the complexity of query evaluation is linear in k and at most quadratic in database size.
- Format: PDF
- Size: 242.5 KB