OLAP Blitzkreig Introduction 3 characteristics of OLAP cubes

OLAP : Blitzkreig Introduction 3 characteristics of OLAP cubes: Large data sets ~ Gb, Tb Expected Query : Aggregation Infrequent updates Star Schema : Hierarchical Dimensions

Attributes and Measures Attributes are columns with values from a fixed domain (foreign keys). Measures are numerical columns.

Imprecision and Uncertainity Imprecision in a tuple refers to an attribute instantiated by a set of values from the domain instead of a single value. Uncertainity refers to a measure represented by a pdf over the domain instead of a single value.

Hierarchical Domains : Star Schema Location Madhya Pradesh Maharashtra Mumbai Pune Bhopal Indore

Restriction on Imprecision We restrict the sets of values in an imprecise fact to either: 1. A singleton set consisting of a leaf level member of the hierarchy, or, 2. The set of all the leaf level members under some non-leaf level member of the hierarchy.

Cells and Regions A region is a vector of attribute values from an imprecise domains of each dimension of the cube. A cell is a region in which all values are leaf level members. Let reg(R) represent the set of cells in a region R.

Queries on precise data A query Q = (R, M, A) refers to a region R, a measure M, and an aggregate function A. Eg : (<Ambassador, Location>, Repairs, Sum) The result of the query in a precise database is obtained by applying A on the measure M of all cells in R. For the example above, the result is (P 1 + P 2)

Queries on imprecise data Consider the query region <Pune, Model> in the figure. It overlaps two imprecise facts P 4 and P 5. Here, we need to make a decision between 3 strategies : None : Ignore both P 4 and P 5 because of their imprecision Contains : Take P 5 because it is contained inside the query Overlaps : Take P 5, and somehow, P 4 as well.

Contains option : Consistency Intuitively, consistency means that the answer to a query should be consistent with the aggregates from individual partitions of the query. Using the Contains option could give rise to inconsistent results. For example, consider the sum aggregate of the query above and that of its individual cells. With the Contains option, will the individual results add up to be the same as the collective?

None option Essentially, the none option ignores the imprecise facts, even if a fact is completely inside the region. Lays waste to the whole notion of having imprecise facts.

Overlaps option : Possible Worlds

Query semantics on Possible worlds With each possible world, assign a weight wi such that the sum of all weights is 1. Intuitively, the weight of a particular world is like probability that it is the correct underlying data. Given a query Q, we can calculate the result for each vi for each world. Thus, we can return a pdf over the answer Z as P[Z = o] = ∑ i : v_i = o wi A neat short answer could be the expected value of Z E[Z] =∑i wi * vi Problem with this is : number of possible worlds is exponential in number of imprecise facts!

Solution : Extended data model With each cell c in a region r, we add a probability pr, c, called the allocation of r to c. The probability of a possible world becomes the multiple of allocations of ranges to cells that have been populated in the world. This leads to a (reasonable) restriction on the kind of probability distributions on possible worlds.

Advantages of EDM No extra infrastructure required for representing imprecision Efficient algorithms for aggregate queries : SUM and COUNT : linear time algo. AVERAGE : slightly complicated algorithm running in O(m + n 3) for m precise facts and n imprecise facts.

Allocation Policies For every region r in the database, we want to assign an allocation pc, r to each cell c in Reg(r), such that ∑c �Reg(r) pc, r = 1 Three ways of doing so: 1. Uniform : Assign each cell c in a region r an equal probability. pc, r = 1 / |Reg(r)|

Allocation Policies For every region r in the database, we want to assign an allocation pc, r to each cell c in Reg(r), such that ∑c �Reg(r) pc, r = 1 However, we can do better. Some cells may be naturally inclined to have more probability than others. Eg : Mumbai will clearly have more repairs than Bhopal. We can do this automatically by giving more probability to cells with higher number of precise facts. 2. Count based : where Nc is the number of precise facts in cell c

Allocation Policies For every region r in the database, we want to assign an allocation pc, r to each cell c in Reg(r), such that ∑c �Reg(r) pc, r = 1 Again, we can arguably get a better result by looking at not just the count, but rather than the actual value of the measure in question. 3. Measure based : next slide.

Measure Based Allocation Assumes the following model : The given database D with imprecise facts has been generated by randomly injecting imprecision in a precise database D'. D' assigns value o to a cell c according to some unknown pdf P(o, c). If we could determine this pdf, the allocation is simply pc, r = P(c) / ∑ c' in Reg(r) P(c')

Maximum Likelihood Principle A reasonable estimate for this function P can be that which maximises the probability of generating the given imprecise data set D. Example : Suppose the pdf depends only on the cells and is independent of the measure values. Thus, the pdf is a mapping �: C �ℝ where C is the set of cells. This pdf can be found by maximising the likelihood function : ℒ (�) = �r �D ∑c �Reg(r) �(c)

EM Algorithm The Expectation Maximization algorithm provides a standard way of maximizing the likelihood, when we have some unknown variables in the observation set. Expectation step (compute data): Calculate the expected value of the unknown variables, given the current estimate of variables. Maximization step (compute generator): Calculate the distribution that maximizes the probability of the current estimated data set.

EM Algorithm : Example Initialization Step: Data: [4, 10, ? ] Initial mean value: 0 New Data: [4, 10, 0, 0] Step 1: New Mean: 3. 5 New Data: [4, 10, 3. 5] Step 4: New Mean: 6. 5625 New Data: [4, 10, 6. 5625] Step 2: New Mean: 5. 25 New Data: [4, 10, 5. 25] Step 5: New Mean: 6. 7825 New Data: [4, 10, 6. 7825] Step 3: New Mean: 6. 125 New Data: [4, 10, 6. 125] Result: New Mean: 6. 890625

EM Algorithm : Application

Experiments : Allocation run time

Experiments : Query run time

Experiments : Accuracy

Summary Model for ambiguity : Imprecision, Uncertainity Querying on uncertain data : None v/s Contains v/s Overlaps option Consistency, Faithfulness Possible Worlds interpretation : size blowup Extended databases : allocation Aggregation algorithms on Extended databases Allocation policies : Uniform Count Measure : EM algorithm Experiments : Allocation time, query time, accuracy

References : OLAP over uncertain and imprecise data (Doug Burdick et al. ) - The VLDB Journal (2007) 16: 123– 144 OLAP over uncertain and imprecise data(Doug Burdick et al. ) - - The VLDB Journal (2005) http: //en. wikipedia. org/wiki/Expectationmaximization_algorithm