Immutably answering Why-Not questions for equivalent conjunctive queries

Abstract

Answering Why-Not questions consists in explaining to developers of complex data transformations or manipulations why their data transformation did not produce some specific results, although they expected them to do so. Different types of explanations that serve as Why-Not answers have been proposed in the past and are either based on the available data, the query tree, or both. Solutions (partially) based on the query tree are generally more efficient and easier to interpret by developers than solutions solely based on data. However, algorithms producing such query-based explanations so far may return different results for reordered conjunctive query trees, and even worse, these results may be incomplete. Clearly, this represents a significant usability problem, as the explanations developers get may be partial and developers have to worry about the query tree representation of their query, losing the advantage of using a declarative query language. As remedy to this problem, we propose the Ted algorithm that produces the same complete querybased explanations for reordered conjunctive query trees.

A Hybrid Approach to Answering Why-Not Questions on Relational Query Results

“…For instance, if we move the selection σ c.sector>97 up all the way below the projection, the query-based explanation returned by both Conseil and Why-Not would change to include this selection operator only. To remedy this problem, we started studying the problem of generating the same query-based explanations for equivalent query trees (and queries) from a theoretical perspective [Bidoit et al 2014a]; however, complexity results and preliminary evaluation show that a practical solution requires further significant runtime improvements. Figure 9 shows the query tree for MOV2 (with the same additional information as on the query tree for CRIME2).…”

Approximate summaries for why and why-not provenance

“…3 which returns start-and end-points of paths of length 2 in a graph with integer node labels such that the end-point is labeled with a lareger number than the start-point. Evaluating rex over the example instance R from the same figure yields three results: Qex(1, 3), Qex (1,4), and Qex (5,6). In this example, we want to explain missing answers of the form Qex(X, 4), i.e., answering the PQ Φex shown in Fig.…”

“…For this example, the first goal R(2, 2) fails, but the second goal R(2, 4) succeeds. Note that in this process there are two potential ways for why we may fail to produce a derivation of Whynot(Q, D, t): (i) a predicate of the rule may be violated by the bindings generated in this way (e.g., if we would have chosen X = 5, then X < 4 would not have held) and (ii) the derivation may derive an existing answer, e.g., if X = 1 and Z = 3, we get the failed derivation r Φex ex (1,3) of the existing answer Qex (1,4). Analysis of Naive Sampling.…”

“…Based on this observation, we find the subset S lb of pairwise disjoint patterns from S that maximizes completeness, i.e., S lb = argmax S ⊆S∧∀p =p ∈S :p⊥pp p∈S cp(p). 4 We use S lb and S ub to define an lower bound cp(S) and upper-bound cp(S) on the completeness of a pattern set S: Fig. 3: p = (2, Z) − (F , F ), p = (3, Z) − (F , F ), p = (2, 1) − (F , F ).…”

Researching Why-Not Questions in Skyline Query Based on Orthogonal Range

“…Researchers have mainly proposed five explanations to answer why-not questions. First of all, operation positioning [20][21][22] refers to finding out the operation that causes the expected result to be lost. Secondly, data modification [23][24][25][26] refers to inserting new data or modifying existing data to make the missing tuple into query results.…”