Abstract
It is shown in the literature that the multiple attribute tree outperforms inverted files in terms of worst-case complexities for partial-match queries. In this paper, we estimate the expected values for the complexities of both complete match query and range query on multiple attribute tree and inverted files. We use a uniform probabilistic model for the input data space. We show that the multiple attribute tree is more efficient than the inverted file in terms of these expected value measures.
Original language | English |
---|---|
Pages (from-to) | 251-266 |
Number of pages | 16 |
Journal | Theoretical Computer Science |
Volume | 62 |
Issue number | 3 |
DOIs | |
State | Published - Dec 1988 |
Externally published | Yes |