1 / 30

QUERYING IMPRECISE DATA IN MOVING OBJECT ENVIRONMENT

QUERYING IMPRECISE DATA IN MOVING OBJECT ENVIRONMENT. REYNOLD CHENG DMITRI V.KALASHNIKOV SUNIL PRABHAKAR PRESENTED BY: VISHNU VANDANA.DANDU. ABSTRACT . Data uncertainty Probabilistic Queries Range Queries Nearest Neighbor Queries. INTRODUCTION .

theta
Download Presentation

QUERYING IMPRECISE DATA IN MOVING OBJECT ENVIRONMENT

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. QUERYING IMPRECISE DATA IN MOVING OBJECT ENVIRONMENT REYNOLD CHENG DMITRI V.KALASHNIKOV SUNIL PRABHAKAR PRESENTED BY: VISHNU VANDANA.DANDU

  2. ABSTRACT • Data uncertainty • Probabilistic Queries • Range Queries • Nearest Neighbor Queries

  3. INTRODUCTION • Systems for continues monitoring or tracking of mobile objects receive updated locations of objects as they move in space. • Due to limitations of the bandwidth and battery power of mobile devices, it is infeasible for database tracking the movement of object. Example: If there is a time delay between capture of location and its script at database, location values received by object may be different from actual locations.

  4. This can be explained using • Nearest Neighbor Query.(1a) • Range Queries.(1b)

  5. Fig. 1. (a) A nearest-neighbor query for a point q can yield false results by using the data values stored in the database. (b) Imprecision can be used to provide answers with guaranteed certainty.

  6. Contributions of this paper: • Formal notation of probabilistic nearest neighbor queries, • Algorithm for answering probabilistic nearest neighbor queries under a general model of uncertainty, • Solutions to probabilistic nearest neighbor queries for two of the most important moving –object models, • Methods for efficient execution of algorithms

  7. UNCERTAINITY MODEL AND PROBABILISTIC QUERIES Uncertainity of an object can be characterized as: • Definition 1. An uncertainty region of an object Oi at time t, denoted by Ui(t), is a closed region such that Oi can be found only inside this region. • Definition 2. The uncertainty probability density function of an object Oi, denoted by fi(x; y; t), is a probability density function of Oi’s location (x; y) at time t, that has a value of 0 outside Ui(t).

  8. Definition 3: Probabilistic Range Query (PRQ). Given a closed region R and a set of n objects O1;O2; . . .;On with uncertainty regions and probability density functions at time t0, a PRQ returns a set of tuples in the form of (Oi; pi), where pi is the nonzero probability that Oi is located inside R at time t0.

  9. Definition 4: Probabilistic Nearest-Neighbor Query (PNNQ). For a set of n objects O1;O2; . . .;On with uncertainty regions and probability density functions given at time t0, a PNNQ for a point q is a query that returns a set of tuples of the form (Oi; pi), where pi is the nonzero probability that Oi is the nearest neighbor of q at time t0.

  10. EXAMPLE:

  11. EVALUATING QUERIES FOR IMPRECISE DATA EVALUATION OF PRQ: A PRQ returns a set of tuples (Oi; pi), where pi is the nonzero probability that object Oi is located in the query region R at time t0. Let S=(O1; . . O|S|) be the set of all moving objects that have to be considered by the PRQ and let X be the set of tuples returned by the PRQ. The algorithm for evaluating the PRQ at time t0 is described in Fig. 3, which basically integrates the probability distribution function in the overlapping area of Ui(t0) and R to compute pi.

  12. PRQ ALGORITHM

  13. EVALUATION OF PNNQ: The solution consists of 4 four steps: • Projection phase • Pruning phase • Bounding phase • Evaluation phase

  14. EXAMPLE: Fig. 4. An example of a PNNQ processing: (a) Locations of objects, (b) uncertainty regions and distances from q, (c) bounding circle, and (d) bounded regions.

  15. ALGORITHM FOR PRUNING PHASE

  16. ALGORITHM FOR EVALUATION PHASE:

  17. QUERYING WITH LINE SEGMENT AND FREE MOVING UNCERTAINITY Two important types of uncertainty are: • Line-Segment Uncertainity • Free-Moving Uncertainity

  18. Examples: Example of uncertainty model under the assumption that objects specify a maximum speed with each update. (a) Line-moving objects. (b) The uncertainty regions of objects a, b. (c) Free-moving objects. (d) The uncertainty regions of objects a, b, and c.

  19. PRQ FOR LINE SEGMENT AND FREE-MOVING UNCERTAINTY • Line moving uncertainty: Pi= length of Ui(t0) that overlaps R length of Ui(t0) • Free Moving uncertainty: Pi=area of Ui(t0) that overlaps R area of Ui(t0)

  20. PNNQ FOR LINE SEGMENT AND FREE MOVING UNCERTAINTY : • The parameters that need to be found to adapt the generic PNNQ solution to a particular uncertainty model for every object Oi are: • Ui(t0) and fi(x,y,t0), • Ni and fi, the shortest and longest distance of Ui(t0) from q, respectively and • Pi(r) and pri(r)

  21. Parameterizing Generic PNNQ Solution for Line segment Uncertainty: • Obtaining ni and fi • Obtaining Pi(r) and pri(r) which is given by Pi(r)= length of Ui(t0) inside Cq(r) Length of Ui(t0)

  22. Parametizing Generic PNNQ Solution for free-Moving Uncertainty • In this case we can obtain Pi(r) as Pi(r)=Overlapping area of Cq(r) andUi(t0) Area of Ui(t0)

  23. Efficient Query Processing: • Efficient execution of the Pruning Phase Using VCI • Efficient Execution of the Evaluation Phase.

  24. PERFORMANCE STUDIES: • Simulation Model(Parameters used)

  25. Execution time versus e".

  26. PERFORMANCE RESULTS: • Effect of € • PNNQ versus NNQ_old Quality of PNNQ versus uncertainty. Execution time versus uncertainty.

  27. Breakdown of execution time.

  28. RELATED WORK: • Future Temporal Logic (FTL) Spatio-temporal Query language • Pfoser and Jensen’s Uncertainty model (I) possible semantics (II) surely semantics (III) probably semantics • Dynamic Query

  29. CONCLUSION: • Defined Generic Model of uncertainty • Presented algorithms • Evaluation of Queries • Uncertainty Models

  30. THANK YOU

More Related