Master’s Thesis Semantic Query Caching in Mobile Environments By: Jekkin Shah Advisor: Dr. Konstantinos Kalpakis

Master’s Thesis 2 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 3 Introduction Disparate works and progresses in :  Geographic Information System (GIS)  Global Positioning System (GPS)  Wireless Technology  Handheld devices Convergence to Mobile Geographic Information System (mobile GIS) Rapid growth in mobile GIS applications in all walks of life Emphasis on spatial data, its storage, retrieval and manipulation

Master’s Thesis 4 Mobile GIS GISGPS WirelessHandheld Convergence

Master’s Thesis 5 Growing List of Applications Car navigation systems Emergency services Real time stock quotes Field services Real time tracking and routing of shipments Environmental surveys and the list is growing rapidly …

Master’s Thesis 6 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 7 Motivation Hungry !!! Lets find a nearby restaurant query Q1: FIND restaurants WHERE location = “ nearby ” McDonalds2 miles Dominos Express2.4 miles Taco Bell3 miles Subway4 miles McDonalds5 miles ……. Found 37 matches

Master’s Thesis 8 Example 1 (cont.) Wait …We also need some gas !!! Lets see if we can find a gas station near McDonalds. query Q2: FIND “McDonalds” WHERE gas Station = “ nearby ” McDonalds5 miles McDonalds12.4 miles Found 2 matches

Master’s Thesis 9 Shouldn’t we speed up the process ? Query Q1 is in local cache Query Q1 subsumes query Q2 Why do we need to execute query Q2 from scratch ?? We need a technique to determine and extract Q2 from Q1 Unfortunately, traditional techniques like page caching do not provide much help in this case Q 1 Q2Q2

Master’s Thesis 10 A new approach – Semantic Caching Along with query results, store the queries also in cache Use these queries (query descriptors) to determine if and how a new query can be answered from cache  Check if the required data is present in cache.  Extract the data from cache Add, remove, merge data by performing corresponding operation on query descriptors Manage cache by managing the query descriptors Think of query descriptors as intelligent pointer references that implicitly contain some information about the data they refer to

Master’s Thesis 11 Problems with traditional caching Pointer references do not contain any implicit information Q1  p1,p2,p3,p4,p5,p6 Q2  p7,p8,p9,p10,p11,p12 Q3  all the pages Space constraints will make it difficult to store all the pages in cache. p1 p3 p5 p7p8 p9p10 p11p12 p2 p4 p6 data3

Master’s Thesis 12 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 13 Contribution An architecture for Semantic Caching in mobile environments A system prototype as a “proof-of-concept” with the following building blocks  Query parser and validator  A Solver for determining query satisfiability  An Executor for processing partial and remainder queries  A Cache manager for efficiently managing the cache A cache replacement algorithm Techniques for query processing

Master’s Thesis 14 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 15 Issues in semantic caching Although the idea of semantic caching is straight forward, store query descriptors along with their results, the issues involved are much harder !! Simple concept but Difficult Implementation Issues 1. We need to decide if the answer is present in cache 2. If present, do we have sufficient information to extract it ?

Master’s Thesis 16 Answering Queries from Cache Q1 Select * from db where A > 50 Q2Select * from db where B < 550 Q3Select * from db where (A > 200 and B < 300) Is result of Q3 present in (Q1 + Q2) ?

Master’s Thesis 17 Solving the implication problem Let T = { Q1, Q2 } be a set of query descriptors already in cache We need to show that Q  T We show that ¬ (Q  T) is FALSE ¬ (Q  T)  ¬ (¬ Q  T)  Q  (¬T)  Q  ¬(T1  T2  T3  T4)  Q  (¬T1)  (¬T2)  (¬T3)  (¬T4) This is the primary technique used in our thesis. The algorithm is adopted from [LY85].

Master’s Thesis 18 Solving the implication problem (Cont.) Exponential growth in the number of equations to be solved. Solution:  Clustering based on Signatures  Signature created by taking into account the predicate attributes present in the query  Restriction on the number of clusters created  Signature used in indexing the query descriptors Attr A, B Attr X, D

Master’s Thesis 19 Data Extraction problem Select * from db where A > 50 Select * from db where B < 550 Select * from db where (A > 200 and B < 300 and C = 100 ) Data1 Data2 Data3 Can we extract Data3 ? We fetch attribute C from remote source and take a Cartesian product with the data already present in cache

Master’s Thesis 20 Answering Partial Queries What happens if Q  T is FALSE ? There may be a non empty intersection set between Q and T Answer (Q  T) locally (Partial match) Send (Q  ¬ T) to the server (Remainder Query) T1 T2 Q

Master’s Thesis 21 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 22 Semantic Caching Architecture Solver (Query implication) Query parser and Validator Cache manager Remote db Local Cache Executor results query

Master’s Thesis 23 Cache Structure Local Cache is implemented as relational database structures Query descriptors are stored in one table indexed by their signatures Corresponding query results (data) are stored in another table An auxiliary table associates the query descriptors with its corresponding data Cache manager interacts with query descriptor table Manipulation of data is achieved through the manipulation of query descriptors

Master’s Thesis 24 Cache: Operations and Management Cache Manager: Replacement module :  Replacement : Determines what needs to be cached and what can be purged out Management module :  Addition : Granularity of addition is a semantic region  Deletion : Removal of region, though not necessarily leading to the removal of data  Merge : To simplify query processing, two or more regions can be merged  Decomposition : A very large region, can be decomposed for efficiency reasons

Master’s Thesis 25 Cache Replacement Theory and Assumptions What is the performance metric ? Conventional caching schemes optimize one or more of the following parameters with the goal of improving the performance  Hit ratio  Response time  Data transmission time Due to the dynamics of our application domain, none of these parameters truly reflect the performance of our applications

Master’s Thesis 26 Theory and Assumptions (Cont.) Cache Hit Rate : how do we define hit rate ?  One: At least one data record obtained from cache  All: All data records to be obtained from local cache  Mid: 50% of data records to be satisfied from local cache Response time:  Partially answered queries make it difficult to accurately define the response time Data transmission time:  Lot of dependence on the actual network parameters like latency and bandwidth

Master’s Thesis 27 Theory and Assumptions (Cont.) Mobile environments: Premium on bandwidth Our goal: “To minimize the cost of servicing the requests that cannot be answered from the local cache” Cost is measured in terms of time Performance metric is Byte hit rate (BHR):  Ratio of actual amount of data served from local cache to the amount of data transferred from the remote source Assumptions:  Negligible query execution time  Uniform latency and bandwidth across the network

Master’s Thesis 28 Replacement Algorithm Guiding Action Selection function (GAS) to assign a value to each semantic region GAS value = a + (s * f * b)  s = size of data transferred from the remote source  f = frequency of access of the query  a, b are domain specific parameters  a = freshness count of each query  b = 1/S d, where S d is the distance between the current location of the moving object and the location of query Using the GAS function the value of each semantic region is calculated

Master’s Thesis 29 Replacement Algorithm (Cont.) For each query in cache we have,  GAS value (V i )  Weight (W i )  Also, we have a limit on the total size of the cache (W) and also the total number of queries (K) that can be admitted Problem definition:  “Given a set of rectangles with a weight and a value, choose at most K rectangles that gives maximum value, provided the weight does not exceed W” The problem can be formulated as the 0-1 Knapsack problem with additional cardinality constraint

Master’s Thesis 30 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 31 Experiments (Setup) Requirements  Workload (datasets and queries)  Modeling the behavior of the moving object  Query execution guidelines Real datasets  Hard to obtain  Complexity in processing due to complex structures of spatial objects Synthetic dataset generator  Easily generated  Various parameters can be controlled

Master’s Thesis 32 Workload Query load selection  Tables Restaurants: LocX, LocY, Name, ID, tables, City, Zip Gas Stations: LocX, LocY, Name, ID, Low, Mid, High  Query specifications: Rectangular queries (select and project only) Number of queries issued per trip : Type of queries: Location aware, location dependent and non- location related Frequency of issuance: Selected randomly ranging from 5 ms to 100 ms Overlap rate: 10-25%

Master’s Thesis 33 Experiments (Moving Object) Behavior of Moving Object  Generating Spatio-Temporal Dataset (GSTD) [PT00]  Moves in a 2D space  Static points and regions called infrastructure emulate real life objects like buildings, rivers, roads etc.  Trajectories are generated using specific guidelines Initial statistical distribution of infrastructure objects Source and destination location Speed of moving object Direction of motion Duration of journey

Master’s Thesis 34 Query Execution Guidelines Controllable parameters  Type of queries: Location dependent, Location aware, Non-location related  Frequency of query issuance  Selectivity of chosen queries  Query overlap rate Parameters are chosen in a variety of combinations  Random  Gaussian distribution  Skewed distribution

Master’s Thesis 35 Results Cache Size Vs Hit Rate ( NEW vs m-LRU) The NEW replacement scheme compares roughly equal to modified LRU replacement scheme BHR increases upto 70% when cache size is progressively increased

Master’s Thesis 36 Results Hit rates Vs Number of queries (NEW scheme) Increasing the number of queries in the system does not substantially increase the hit rates. Byte hit rate performs nearly equal to Hit rate Mid

Master’s Thesis 37 Semantic Query Caching in Mobile Environments Introduction Motivation Contribution Concept of Semantic Caching Issues involved in semantic caching System Architecture Prototype and Experiments Conclusion and further work

Master’s Thesis 38 Conclusion  No assumption made on Spatial Locality of Reference  Query descriptors act as “Intelligent” References  Can support Content Based Reasoning  Ability to take advantage of Schema Knowledge Page / Tuple caching schemes do not scale well in our GIS domain Reasons: “Unintelligent” pointer references Questionable assumption of Spatial Locality of Reference Inability to take advantage of Semantic Overlaps

Master’s Thesis 39 Advantages of Semantic Caching Benefits of Semantic Caching  Leverages semantic locality found in typical mobile GIS applications  Adapts dynamically to the patterns of user queries rather than caching static clusters of tuples  Minimizes cost of cache lookup due to compact representation of query descriptors  Capable of providing partial and/or approximate answers to queries quickly

Master’s Thesis 40 Conclusion (Cont.) Shortcomings of Semantic Caching  Complicated cache management schemes  Too restrictive. Solver can process only simple type of queries  Captures the semantics of the query and not the result objects. Hence, fails to utilize cached objects when the semantics of the query do not match

Master’s Thesis 41 Conclusion (Cont.) Future work … Lots of things  Make the solver more general to handle different types of queries  Make the caching scheme flexible enough to capture the semantics of the query descriptors as well as the result objects  Simpler cache management  Ability to share cache with peers