Inconsistent Databases Research Articles

Combined Approximations for Uniform Operational Consistent Query Answering

Operational consistent query answering (CQA) is a recent framework for CQA based on revised definitions of repairs, which are built by applying a sequence of operations (e.g., fact deletions) starting from an inconsistent database until we reach a database that is consistent w.r.t. the given set of constraints. It has been recently shown that there is an efficient approximation for computing the percentage of repairs that entail a given query when we focus on primary keys, conjunctive queries, and assuming the query is fixed (i.e., in data complexity). However, it has been left open whether such an approximation exists when the query is part of the input (i.e., in combined complexity). We show that this is the case when we focus on self-join-free conjunctive queries of bounded generelized hypertreewidth. We also show that it is unlikely that efficient approximation schemes exist once we give up one of the adopted syntactic restrictions, i.e., self-join-freeness or bounding the generelized hypertreewidth. Towards the desired approximation, we introduce a counting complexity class, called SpanTL, show that each problem in it admits an efficient approximation scheme by using a recent approximability result about tree automata, and then place the problem of interest in SpanTL.

Database Repairing with Soft Functional Dependencies

A common interpretation of soft constraints penalizes the database for every violation of every constraint, where the penalty is the cost (weight) of the constraint. A computational challenge is that of finding an optimal subset: a collection of database tuples that minimizes the total penalty when each tuple has a cost of being excluded. When the constraints are strict (i.e., have an infinite cost), this subset is a “cardinality repair” of an inconsistent database; in soft interpretations, this subset corresponds to a “most probable world” of a probabilistic database, a “most likely intention” of a probabilistic unclean database, and so on. Within the class of functional dependencies, the complexity of finding a cardinality repair is thoroughly understood. Yet, very little is known about the complexity of finding an optimal subset for the more general soft semantics. The work described in this manuscript makes significant progress in that direction. In addition to general insights about the hardness and approximability of the problem, we present algorithms for two special cases (and some generalizations thereof): a single functional dependency, and a bipartite matching. The latter is the problem of finding an optimal “almost matching” of a bipartite graph where a penalty is paid for every lost edge and every violation of monogamy. For these special cases, we also investigate the complexity of additional computational tasks that arise when the soft constraints are used as a means to represent a probabilistic database in the case of a probabilistic unclean database.

Paraconsistent logic and query answering in inconsistent databases

ABSTRACT This paper concerns the paraconsistent logic LPQ⊃,F and an application of it in the area of relational database theory. The notions of a relational database, a query applicable to a relational database, and a consistent answer to a query with respect to a possibly inconsistent relational database are considered from the perspective of this logic. This perspective enables among other things the definition of a consistent answer to a query with respect to a possibly inconsistent database without resort to database repairs. In an earlier paper, LPQ⊃ ,F is presented with a sequent-style natural deduction proof system. In this paper, a sequent calculus proof system is presented instead because such proof systems are generally considered more suitable as the basis of proof search procedures than natural deduction proof systems and proof search procedures can serve as the core of algorithms for computing consistent answers to queries.

Belnap-Dunn Logic and Query Answering in Inconsistent Databases with Null Values

This paper concerns an expansion of first-order Belnap-Dunn logic, called BD^{\supset,F}_{\bot}, and an application of this logic in the area of relational database theory. The notion of a relational database, the notion of a query applicable to a relational database, and several notions of an answer to a query with respect to a relational database are considered from the perspective of this logic, taking into account that a database may be an inconsistent database and/or a database with null values. The chosen perspective enables among other things the definition of a notion of a consistent answer to a query with respect to a possibly inconsistent database without resort to database repairs. For each of the notions of an answer considered, being an answer to a query with respect to a database of the kind considered is decidable.

Consistent Range Approximation for Fair Predictive Modeling

This paper proposes a novel framework for certifying the fairness of predictive models trained on biased data. It draws from query answering for incomplete and inconsistent databases to formulate the problem of consistent range approximation (CRA) of fairness queries for a predictive model on a target population. The framework employs background knowledge of the data collection process and biased data, working with or without limited statistics about the target population, to compute a range of answers for fairness queries. Using CRA, the framework builds predictive models that are certifiably fair on the target population, regardless of the availability of external data during training. The framework's efficacy is demonstrated through evaluations on real data, showing substantial improvement over existing state-of-the-art methods.

Discrepancies between ePrescriptions and dispensing in Belgium, 6 years after the launch of the electronic prescribing – a mixed-method study

ABSTRACT Objectives The number of electronic prescriptions (ePrescriptions) grows steadily in Belgium as in other European countries. In the future, Belgium wants to dematerialize the ePrescription flow, removing all paper trails. A quality check of the digital content and implementation of national ePrescription guidelines in the field was conducted, comparing the content at both prescription and pharmacy side. Methods An explanatory mixed-methods design was applied. In a first phase, potential problems (warning flags) were identified by consulting stakeholders. Secondly, the warning flags were validated to problems (errors) in a random set of ePrescriptions collected in April 2019. In a third phase, explanatory interviews were held with various stakeholders in order to find explanations and to identify the initiators of these errors. Results In the first phase, 15 warning flags were identified to evaluate the quality of an ePrescription. In the second phase, a random selection of 11,798 ePrescriptions was validated. The most prevalent errors found, were the digital construction of the messages (18.88%), combined with lots of necessarily deemed substitutions by the pharmacist (3.39%) not following what was prescribed originally. In the third phase, stakeholders indicated that software of the prescriber and the use of inconsistent databases between prescriber and pharmacy can often be seen as the cause and initiator of these problems. Conclusions Use of authentic medication databases and well-designed software systems have the potential to solve ePrescription problems. Focus should go to prevention instead of detection.

WITHDRAWN: Bhattacharyya coefficient target feature matching based weighted emphasis adaptive boosting classification for predictive analytics with big data

Abstract Data warehousing is the most significant systems that hold large amounts of information. The data warehouse integrates the data from a large and inconsistent database that are located in different locations. Nowadays, data warehouse tools and technologies difficult to handle the analytic process of data since the dimensions of data volumes get increased rapidly. A novel technique Bhattacharyya Coefficient Target Matching Pursuit based Weighted Emphasis Adaptive Boosting Classification (BCTMP-WEABC) technique is introduced for predicting future outcomes with higher accuracy and lesser time consumption. The BCTMP-WEABC consists of two modules namely attribute (i.e., features) selection and classification for handling a large amount of data in the warehouse. In the BCTMP-WEABC technique, Bhattacharyya Coefficient Target Matching Pursuit is applied to select the relevant attributes for performing the classification process based on the similarity measure. After that, Weighted Emphasis Adaptive Boosting Classification (WEABC) process is carried out to perform predictive analytics with the selected attributes for minimizing the time consumption. The WEABC provides an accurate classification by implementing the set of weak learners as a quadratic discriminant classifier. The weak learner performs a likelihood ratio test for distinguishing the given input data into different classes based on mean and deviation. The WEABC combines the weak learners and provides accurate classification and also minimizes the quadratic error. Result evaluation is carried out with factors such as feature selection rate, classification accuracy, classification time, and false-positive rate with respect to the number of features. The evaluation and conferred statistical results notify the development of the proposed BCTMP-WEABC technique in terms of higher classification accuracy, feature selection rate, and minimum time consumption as well as false positives than the conventional methods.

Counting subset repairs with functional dependencies

We study the problem of counting the repairs of an inconsistent database in the case where constraints are Functional Dependencies (FDs). A repair is then a maximal independent set of the conflict graph, wherein nodes represent facts and edges represent violations. We establish a dichotomy in data complexity for the complete space of FDs: when the FD set has, up to equivalence, what we call a “left-hand-side chain,” the repairs can be counted in polynomial time; otherwise, the problem is ♯P-complete. Moreover, the property of having a left-hand-side chain up to equivalence coincides with the condition that the conflict graph of every inconsistent database for the schema is P4-free, and it is polynomial-time decidable.

Specifying and computing causes for query answers in databases via database repairs and repair-programs

There is a recently established correspondence between database tuples as causes for query answers in databases and tuple-based repairs of inconsistent databases with respect to denial constraints. In this work, answer-set programs that specify database repairs are used as a basis for solving computational and reasoning problems around causality in databases, including causal responsibility. Furthermore, causes are introduced also at the attribute level by appealing to an attribute-based repair semantics that uses null values. Corresponding repair-programs are introduced, and used as a basis for computation and reasoning about attribute-level causes. The answer-set programs are extended in order to capture causality under integrity constraints.

The computation of optimal subset repairs

Computing an optimal subset repair of an inconsistent database is becoming a standalone research problem and has a wide range of applications. However, it has not been well-studied yet. A tight inapproximability bound of the problem computing optimal subset repairs is still unknown, and there is still no existing algorithm with a constant approximation factor better than two. In this paper, we prove a new tighter inapproximability bound of the problem computing optimal subset repairs. We show that it is usually NP-hard to approximate it within a factor better than 17/16. An algorithm with an approximation ratio (2 - 1/2 σ-1 )is developed, where σ is the number of functional dependencies. It is the current best algorithm in terms of approximation ratio. The ratio can be further improved if there are a large amount of quasi -Turán clusters in the input database. Plenty of experiments are conducted on real data to examine the performance and the effectiveness of the proposed approximation algorithms in real-world applications.

Approximate denial constraints

The problem of mining integrity constraints from data has been extensively studied over the past two decades for commonly used types of constraints, including the classic Functional Dependencies (FDs) and the more general Denial Constraints (DCs). In this paper, we investigate the problem of mining from data approximate DCs, that is, DCs that are "almost" satisfied. Approximation allows us to discover more accurate constraints in inconsistent databases and detect rules that are generally correct but may have a few exceptions. It also allows to avoid overfitting and obtain constraints that are more general, more natural, and less contrived. We introduce the algorithm ADCMiner for mining approximate DCs. An important feature of this algorithm is that it does not assume any specific approximation function for DCs, but rather allows for arbitrary approximation functions that satisfy some natural axioms that we define in the paper. We also show how our algorithm can be combined with sampling to return highly accurate results considerably faster.

Counting and enumerating preferred database repairs

In its traditional definition, a repair of an inconsistent database is a consistent database that differs from the inconsistent one in a “minimal way.” Often, repairs are not equally legitimate, as it is desired to prefer one over another; for example, one fact is regarded more reliable than another, or a more recent fact should be preferred to an earlier one. Motivated by these considerations, researchers have introduced and investigated the framework of preferred repairs, in the context of denial constraints and subset repairs. There, a priority relation between facts is lifted towards a priority relation between consistent databases, and repairs are restricted to the ones that are optimal in the lifted sense. Three notions of lifting (and preferred repairs) have been proposed: Pareto, global, and completion.In this article, we investigate the complexity of three problems on preferred repairs. The first is the problem of deciding whether the priority relation contains enough information to clean the database unambiguously, or in other words, whether there is exactly one preferred repair. We show that the different lifting semantics entail highly different complexities for this problem. Then, we study the ability to quantify ambiguity, by investigating two classes of problems. The first is that of counting the preferred repairs. We establish a dichotomy in data complexity for the entire space of (sets of) functional dependencies for all three notions. The second class of problems is that of enumerating (i.e., generating) the preferred repairs. We devise enumeration algorithms with efficiency guarantees on the delay between generated repairs, even for constraints represented as general conflict graphs or hypergraphs.

Computing Optimal Repairs for Functional Dependencies

We investigate the complexity of computing an optimal repair of an inconsistent database, in the case where integrity constraints are Functional Dependencies (FDs). We focus on two types of repairs: an optimal subset repair (optimal S-repair), which is obtained by a minimum number of tuple deletions, and an optimal update repair (optimal U-repair), which is obtained by a minimum number of value (cell) updates. For computing an optimal S-repair, we present a polynomial-time algorithm that succeeds on certain sets of FDs and fails on others. We prove the following about the algorithm. When it succeeds, it can also incorporate weighted tuples and duplicate tuples. When it fails, the problem is NP-hard and, in fact, APX-complete (hence, cannot be approximated better than some constant). Thus, we establish a dichotomy in the complexity of computing an optimal S-repair. We present general analysis techniques for the complexity of computing an optimal U-repair, some based on the dichotomy for S-repairs. We also draw a connection to a past dichotomy in the complexity of finding a “most probable database” that satisfies a set of FDs with a single attribute on the left-hand side; the case of general FDs was left open, and we show how our dichotomy provides the missing generalization and thereby settles the open problem.

Foundations of Query Answering on Inconsistent Databases

Notwithstanding the traditional view that database instances must respect all integrity constraints imposed on them, it is relevant to develop theories about how to handle database instances that violate some integrity constraints, and more particularly, how to cope with query answering in the presence of inconsistency. Such a theory developed over the past twenty years is currently known as consistent query answering (CQA). The aim of this article is to summarize and discuss some core concepts and theoretical developments in CQA.

SOME MODEL-THEORETIC RESULTS ON THE 3-VALUED PARACONSISTENT FIRST-ORDER LOGIC QCIORE

AbstractThe 3-valued paraconsistent logic Ciore was developed by Carnielli, Marcos and de Amo under the name LFI2, in the study of inconsistent databases from the point of view of logics of formal inconsistency (LFIs). They also considered a first-order version of Ciore called LFI2*. The logic Ciore enjoys extreme features concerning propagation and retropropagation of the consistency operator: a formula is consistent if and only if some of its subformulas is consistent. In addition, Ciore is algebraizable in the sense of Blok and Pigozzi. On the other hand, the logic LFI2* satisfies a somewhat counter-intuitive property: the universal and the existential quantifier are inter-definable by means of the paraconsistent negation, as it happens in classical first-order logic with respect to the classical negation. This feature seems to be unnatural, given that both quantifiers have the classical meaning in LFI2*, and that this logic does not satisfy the De Morgan laws with respect to its paraconsistent negation. The first goal of the present article is to introduce a first-order version of Ciore (which we call QCiore) preserving the spirit of Ciore, that is, without introducing unexpected relationships between the quantifiers. The second goal of the article is to adapt to QCiore the partial structures semantics for the first-order paraconsistent logic LPT1 introduced by Coniglio and Silvestrini, which generalizes the semantic notion of quasi-truth considered by Mikeberg, da Costa and Chuaqui. Finally, some important results of classical Model Theory are obtained for this logic, such as Robinson’s joint consistency theorem, amalgamation and interpolation. Although we focus on QCiore, this framework can be adapted to other 3-valued first-order LFIs.

5 Conservative treatment for PAD - Risk factor management.

5 Conservative treatment for PAD - Risk factor management.

Errata for trial publications are not uncommon, are frequently not trivial, and can be challenging to access: a retrospective review.

ObjectiveThe research sought to determine the prevalence of errata for drug trial publications that are included in systematic reviews, their potential value to reviews, and their accessibility via standard information retrieval methods.MethodsThe authors conducted a retrospective review of included studies from forty systematic reviews of drugs evaluated by the Canadian Agency for Drugs and Technologies in Health (CADTH) Common Drug Review (CDR) in 2015. For each article that was included in the systematic reviews, we conducted searches for associated errata using the CDR review report, PubMed, and the journal publishers’ websites. The severity of errors described in errata was evaluated using a three-category scale: trivial, minor, or major. The accessibility of errata was determined by examining inclusion in bibliographic databases, costs of obtaining errata, time lag between article and erratum publication, and correction of online articles.ResultsThe 40 systematic reviews included 127 articles in total, for which 26 errata were identified. These errata described 38 errors. When classified by severity, 6 errors were major; 20 errors were minor; and 12 errors were trivial. No one database contained all the errata. On average, errata were published 211 days after the original article (range: 15–1,036 days). All were freely available. Over one-third (9/24) of online articles were uncorrected after errata publication.ConclusionErrata frequently described non-trivial errors that would either impact the interpretation of data in the article or, in fewer cases, impact the conclusions of the study. As such, it seems useful for reviewers to identify errata associated with included studies. However, publication time lag and inconsistent database indexing impair errata accessibility.

Expressive power of entity-linking frameworks

We develop a unifying approach to declarative entity linking by introducing the notion of an entity-linking framework and an accompanying notion of the certain links in such a framework. In an entity-linking framework, logic-based constraints are used to express properties of the desired link relations in terms of source relations and, possibly, in terms of other link relations. The definition of the certain links in such a framework makes use of weighted repairs and consistent answers in inconsistent databases. We demonstrate the modeling capabilities of this approach by showing that numerous concrete entity-linking scenarios can be cast as such entity-linking frameworks for suitable choices of constraints and weights. By using the certain links as a measure of expressive power, we investigate the relative expressive power of several entity-linking frameworks and obtain sharp comparisons.

Quasi-classical reasoning in paraconsistent databases

The well-founded model for any general deductive database computed using the paraconsistent relational model, based on four-valued logic, does not support inference rules such as disjunctive syllogism. In order to support disjunctive syllogism, we utilize the generalization of the relational model to quasi-classic (QC) logic, whose inference power is much closer to classical logic. As the paraconsistent relational model is capable of representing incomplete and inconsistent data, we propose an algorithm to find QC model for inconsistent positive extended disjunctive deductive databases. We also provide the proof for the algorithm.

First-order under-approximations of consistent query answers

Consistent Query Answering (CQA) is a principled approach for answering queries on inconsistent databases. The consistent answer to a query q on an inconsistent database db is the intersection of the answers to q on all repairs, where a repair is any consistent database that is maximally close to db. Unfortunately, computing consistent answers under primary key constraints has already exponential data complexity for very simple conjunctive queries, and is therefore completely impracticable.In this paper, we propose a new framework for divulging an inconsistent database to end users, which adopts two postulates. The first postulate complies with CQA and states that inconsistencies should never be divulged to end users. Therefore, end users should only get consistent query answers. The second postulate states that only those queries can be answered whose consistent answers can be obtained with low data complexity (i.e., by a polynomial-time algorithm or even a first-order logic query). User queries that exhibit a higher data complexity will be rejected.A significant problem in this framework is as follows: given a rejected query, find other queries, called under-approximations, that are accepted and whose consistent answers are contained in those of the rejected query. We provide solutions to this problem for the special case where the constraints are primary keys and the queries are self-join-free conjunctive queries.