Optimal Route Queries with Arbitrary Order Constraints
read more
Citations
Spatial keyword search: a survey
Efficient Clue-Based Route Search on Road Networks
Optimal route search with the coverage of users' preferences
Finding Top-k Optimal Sequenced Routes
Finding the minimum spatial keyword cover
References
R-trees: a dynamic index structure for spatial searching
Efficient retrieval of the top-k most relevant spatial web objects
Keyword Search on Spatial Databases
Efficient query processing in geographic web search engines
On trip planning queries in spatial databases
Related Papers (5)
Frequently Asked Questions (15)
Q2. What are the future works mentioned in the paper "Optimal route queries with arbitrary order constraints" ?
In the future, the authors plan to study alternative definitions of the optimal route query, that have temporal constraints ( e. g., have lunch at a specified period ) or maximize the number of categories to be visited given a total travel length budget.
Q3. What is the main reason why backward search methods prune?
Since backward search methods prune based solely on the greedy bound, they tend to maintain and process a large number of useless sub-routes.
Q4. What is the main reason for SBS to prune?
The main reason is that SBS prunes with only the bound θ obtained by the greedy algorithm, whereas PLUB tightens the the bound whenever a total-ordered subquery returns a better result.
Q5. What is the way to solve the optimal suffix table?
During the computation of the optimal suffix table, RLORD uses a pruning technique to eliminate sub-routesthat cannot participate in the optimal solution.
Q6. What is the effect of the greedy algorithm on the performance of the optimal route query?
the greedy algorithm is more likely to identify a good route whose length is close to the optimal one, meaning that the candidate set CS becomes smaller, leading to decreased join costs.
Q7. How many experiments have confirmed the proposed methods are efficient and practical?
Extensive experiments, using large-scale real and synthetic datasets, confirm that the proposed methods are efficient and practical.
Q8. What is the way to decompose a general optimal route query?
PLUB [11] decomposes a general optimal route query to multiple total-order queries and processes them individually, e.g., using R-LORD.
Q9. What is the main reason for the performance gap between SBS and PLUB?
As m grows, the accuracy of the greedy method worsens, and, consequently, the performance gap between SBS and PLUB gradually closes.
Q10. What is the main drawback of the backward join method?
The batch backward search (BBS) method, shown in Algorithm 3, improves SBS by employing batch processing in the backward join operations.
Q11. How is the optimal route query proven?
In fact, the optimal route query is proven to be NP-hard [13], and heuristics-based algorithms such as Greedy cannot guarantee optimality of the result.
Q12. What is the way to prune a prefix?
A simple idea for pruning is to backtrack whenever the length of the current prefix reaches or exceeds the upper bound θ, since subsequent searches based on this prefix cannot possibly lead to the optimal solution to the query.
Q13. What is the way to prune sub-routes?
Such sub-routes are pruned in SBS and BBS for the original optimal route query, since they can only lead to complete routes (i.e., those covering all 3 categories) that visit a pub before a restaurant, violating GQ.
Q14. What is the order of the new clusters to be added to the current prefix?
The order of new clusters to be added to the current prefix is based on their MBRs’ minimum distances to the MBR of the current last cluster P (line 7).
Q15. What are the solutions for the optimal route query?
Section 3 and 4 present solutions for the optimal route query, following the backward search and forward search frameworks, respectively.