Impact of answer number inter-arrival time of attribute- attribute-changed eventattribute-changed event

Figure 5.10 shows the results of the simulations with different mean of inter-arrival time.

Because our simulation time is set to 3000 seconds, there more attribute-changed events arrive

0

(a) 5% data variation rate

0

(b) 10% data variation rate

0

(c) 15% data variation rate

0

(d) 5% data variation rate

0

(e) 10% data variation rate

0

(f) 15% data variation rate

Figure 5.9: Simulation with different k

when mean of inter-arrival time is smaller. During the simulation, the response time of the first-time evaluation is significantly large because we have to sort all the attributes in every dimension. When we have more attribute-changed event arrived, the response time of first-time evaluation can be amortised to the other evaluations. Therefore, we can see the average response time of all approaches increase sightly when mean of inter-arrival time increases in Figure 5.10(a)(b)(c). In Figure 5.10(d)(e)(f), the total packet bytes decreases when mean of inter-arrival time increases. This is because more attribute-changed event arrived make the chance of reevaluation increases and the total packet bytes also increase.

0

(a) 5% data variation rate

0

(b) 10% data variation rate

0

(c) 15% data variation rate

0

Arrival time of attribute-changed event(sec)

Number of packet bytes(K)

FKNMatchAD CFKNMatchAD-C CFKNMatchAD-C with SR CFKNMatchAD-D

(d) 5% data variation rate

0

Arrival time of attribute-changed event(sec)

Number of packet bytes(K)

FKNMatchAD CFKNMatchAD-C CFKNMatchAD-C with SR CFKNMatchAD-D

(e) 10% data variation rate

0

Arrival time of attribute-changed event(sec)

Number of packet bytes(K)

FKNMatchAD CFKNMatchAD-C CFKNMatchAD-C with SR CFKNMatchAD-D

(f) 15% data variation rate

Figure 5.10: Simulation with different mean of inter-arrival time

Chapter 6 Conclusion

In this thesis, we consider the problem of continuous k-n-match search. We propose a algo-rithm CFKNMathAD to compute a safe region for every attribute of points in high dimensional databases. We do not perform the query reevaluation if fluctuated attribute is within its safe region. We reduce the query response time without doing unnecessary query reevaluation.

Furthermore, we also apply our algorithm in de-centralized environment to balance the sys-tem workload. Our experiments show that CFKNMatchAD has better performances than FKNMatchAD in different data variation rates. Finally, we conclude that CFKNMatchAD reduce the query response time and balance system workload.

Bibliography

[1] R. Agrawal, K.-I. Lin, H. S. Sawhney, and K. Shim. Fast similarity search in the pres-ence of noise, scaling, and translation in time-series databases. In Proc. of the 21th International Conference on Very Large Data Bases (VLDB), pages 490–501, 1995.

[2] S. Arya, D. M. Mount, N. S. Netanyahu, R. Silverman, and A. Y. Wu. An optimal algorithm for approximate nearest neighbor seraching fixed dimensions. Journal of the ACM(JCAM), 45(6):891–923, 1998.

[3] A. Badel, J. P. Mornon, and S. Hazout. Searching for geometric molecular shape comple-mentarity using bidimensional surface profiles. Journal of Molecular Graphics, 10(4):205–

211, 1992.

[4] N. Beckmann, H.-P. Kriegel, R. Schneider, and B. Seeger. The r*-tree: an efficient and robust access method for points and rectangles. In Proc. of the 10th ACM Conference on Management of Data (SIGMOD), pages 322–331, 1990.

[5] S. Berchtold, C. B¨ohm, D. A. Keim, and H.-P. Kriegel. A cost model for nearest neighbour search. In Proc. of the 16th ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems(PODS), pages 78 – 86, 1997.

[6] S. Berchtold, D. A. Keim, and H.-P. kriegel. The x-tree: An index structure for high-dimensional data. In Proc. of the 22th International Conference on Very Large Data Bases (VLDB), pages 28 – 39, 1996.

[7] N. Bruno, L. Gravano, and A. Marian. Evaluating top-k queries over web-accessible databases. In Proc. of the 18th International Conference on Data Engineering(ICDE), page 369, 2002.

[8] R. Fagin, A. Lotem, and M. Naor. Optimal aggregation algorithms for middleware.

In Proc. of the 20th ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems(PODS), pages 102–113, 2001.

[9] L. Gao, Z. Yao, and X. Wang. Evaluating continuous nearest neighbor queries for stream-ing time series via pre-fetchstream-ing. In Proc. of the 11th International Conference on Infor-mation and Knowledge Management (CIKM), pages 485 – 492, 2002.

[10] A. Guttman. R-trees: A dynamic index structure for spatial searching. In Proc. of the 4th ACM Conference on Management of Data (SIGMOD), pages 47–57, 1984.

[11] H. Hu, J. Xu, and D. L. Lee. A generic framework for monitoring continuous spatial queries over moving objects. In Proc. of the 25th ACM Conference on Management of Data (SIGMOD), pages 479–490, 2005.

[12] N. Katayama and S. Satoh. The sr-tree: An index structure for high-dimensional near-est neighbor queries. In Proc. of the 17th ACM Conference on Management of Data (SIGMOD), pages 369 – 380, 1997.

[13] F. Korn, N. Sidiropoulos, C. Faloutsos, E. Siegel, and Z. Protopapas. Fast nearest neigh-bor serach in medical image databases. In Proc. of the 22th International Conference on Very Large Data Bases (VLDB), pages 215–226, 1996.

[14] E. Kushilevitz, R. Ostrovsky, and Y. Rabani. Efficient search for approximate nearest neighbor in high dimensional spaces. In Proc. of the 30th annual ACM symposium on Theory of computing(STOC), pages 614 – 623, 1998.

[15] K. Mouratidis, M. L. Yiu, D. Papadias, and N. Mamoulis. Continuous nearest neigbor monitoring in road networks. In Proc. of the 32nd International Conference on Very Large Data Bases (VLDB), pages 43–54, 2006.

[16] T. Seidl and H.-P. Kriegel. Optimal multi-step k-nearest neighbor search. In Proc. of the 18th ACM Conference on Management of Data (SIGMOD), pages 154–165, 1998.

[17] T. Sellis, N. Roussopoulos, and C. Faloutsos. The r⁺-tree: a dynamic index for multi-dimensional objects. In Proc. of the 13th International Conference on Very Large Data Bases (VLDB), pages 507–518, 1987.

[18] A. P. Sistla, O. Wolfson, S. Chamberlain, and S. Dao. Modeling the querying moving objects. In Proc. of the 13th International Conference on Data Engineering(ICDE), pages 422–432, 1997.

[19] Z. Song and N. Roussopoulos. K-nearest neighbor search for moving query point. In Proceedings of the 7th International Symposium on Advances in Spatial and Temporal Databases (SSTD), pages 79–96, 2001.

[20] Y. Tao and D. Papadias. Time parameterized queries in spatio-temporal databases. In Proc. of the 22nd ACM Conference on Management of Data (SIGMOD), pages 334–345, 2002.

[21] A. K. H. Tung, R. Zhang, N. Koudas, and B. C. Ooi. Similarity search: A matching based approach. In Proc. of the 32nd International Conference on Very Large Data Bases (VLDB), pages 631 – 642, 2006.

[22] R. Weber, H.-J. Schek, and S. Blott. A quantitative analysis and performance study for similarity-search methods in high-dimensional spaces. In Proc. of the 24th International Conference on Very Large Data Bases (VLDB), pages 194 – 205, 1998.

[23] D. A. White and R. Jain. Similarity indexing with the ss-tree. In Proc. of the 12th International Conference on Data Engineering, pages 516 – 523, 1996.

[24] J. Xu, X. Tang, W.-C. Lee, and M. Wu. Top-k monitoring in wireless sensor networks.

IEEE Transactions on Knowledge and Data Engineering(TKDE), 19(7):962–976, 2007.

[25] K. Yi, H. Yu, J. Yang, G. Xia, and Y. Chen. Efficient maintenance of materialized top-k views. In Proc. of the 19th International Conference on Data Engineering(ICDE), pages 189–200, 2003.

[26] C. Yu, B. C. Ooi, K.-L. Tan, and H. V. Jagadish. Indexing the distance: An efficient method to knn processing. In Proc. of the 27th International Conference on Very Large Data Bases (VLDB), pages 421 – 430, 2001.