Deductive Databases Research Articles

COMPUTING PERFECT AND STABLE MODELS USING ORDERED MODEL TREES

Ordered model trees were introduced as a normal form for disjunctive deductive databases. They were also used to facilitate the computation of minimal models for disjunctive theories by exploiting the order imposed on the Herbrand base of the theory. In this work we show how the order on the Herbrand base can be used to compute perfect models of a disjunctive stratified finite theory. We are able to compute the stable models of a general finite theory by combining the order on the elements of the Herbrand base with previous results that had shown that the stable models of a theory T can be computed as the perfect models of a corresponding disjunctive theory ɛT resulting from applying the so called evidential transformation to T. While other methods consider many models that are rejected at the end, the use of atom ordering allows us to guarantee that every model generated belongs to the class of models being computed. As for negation‐free databases, the ordered tree serves as the canonical representation of the database.

Multiple Query Optimization with Depth-First Branch-and-Bound and Dynamic Query Ordering

In certain database applications such as deductive databases, batch query processing, and recursive query processing etc., usually a single query gets transformed into a set of closely related database queries. Also, great benefits can be obtained by executing a group of related queries all together in a single unified multi-plan instead of executing each query separately. In order to achieve this Multiple Query Optimization (MQO) identifies common task(s) (e.g. common subexpressions, joins, etc.) among a set of query plans and creates a single unified plan (multi-plan) which can be executed to obtain the required outputs for all queries at once. In this paper a new heuristic function (hc), dynamic query ordering heuristics, and Depth-First Branch-and-Bound (DFBB) are defined and experimentally evaluated, and compared with existing methods which use A* and static query ordering. Our experiments show that all three of hc, DFBB, and dynamic query ordering help to improve the performance of our MQO algorithm.

A paraconsistent relational data model

We present a generalisation of the relational data model based on a 4-valued paraconsistent logic. Our data model is capable of manipulating incomplete as well as inconsistent information. For this model, we define algebraic operators that are generalisations of the usual operators, such as union, selection, join, on ordinary relations. Our data model can underlie any database management system that deals with incomplete or inconsistent information. As another application of our model and its algebra, we present a bottom-up method for constructing the weak well-founded model of general deductive databases. This method can be very simply extended to construct the well-founded model.

Incremental view maintenance in extended relational databases

This paper presents a method for efficiently computing updates to a view when relations appearing in the view's definition are modified. Views that are defined using all of the relational operators, stratified negation and general recursion are considered. As a result, the method can be employed in both a traditional relational database system and also a deductive database system. The method can compute the updates without requiring access to a materialization of the view and is shown to be more efficient than related methods. The method can compute the updates even when the view's definition is modified.

The query clustering problem: a set partitioning approach

In this research, we address the query clustering problem which involves determining globally optimal execution strategies for a set of queries. The need to process a set of queries together often arises in deductive database systems, scientific database systems, large bibliographic retrieval systems and several other database applications. We address the optimization problem from the perspective of overlaps in data requirements, and model the batched operations using a set-partitioning approach. In this model, we first consider the case of m queries each involving a two-way join operation. We develop a recursive methodology to determine all the processing strategies in this case. Next, we establish certain dominance properties among the strategies, and develop exact as well as heuristic algorithms for selecting an appropriate strategy. We extend this analysis to a clustering approach, and outline a framework for optimizing multiway joins. The results show that the proposed approach is viable and efficient, and can easily be incorporated into the query processing component of most database systems.

The 1993 International Logic Programming Symposium

The 1993 International Logic Programming Symposium was held in Vancouver, British Columbia, on 26-29 October. It presented the state of the art in logic programming, emphasizing the deliberate interaction with other fields, in particular, humanistic fields. Topics covered at the symposium included algorithmic analysis, programming methodologies, semantic analysis, deductive databases, and programming language design.

AN OBJECT-ORIENTED RULE EVALUATION SCHEME FOR LOGICAL DATABASES IN MASSIVELY PARALLEL ENVIRONMENT

This paper extends an object-oriented approach to query processing in a deductive database with the consideration of deductive rules. In the object-oriented approach, each deductive rule and fact is represented as an active object, which acts with its own dedicated control sequences according to its functionality. All operations in objects are proceeded asynchronously. Since no object has global knowledge about the database, cooperation among objects is necessary in answering a query. While most of the conventional search strategies focus on some parts of the AND/OR tree expanded by a query at a time, the object-oriented approach can search down the AND/OR tree with a breadth-first strategy in parallel. As a result, all parts of the tree can be considered globally so that queries do not have to be evaluated in a tuple-by-tuple, operation-by-operation fashion; and the need of maintaining temporary relations can be eliminated.

Rule allocation in distributed deductive database systems

Allocation of rules to sites in a distributed deductive database system is an important and challenging task especially for a large knowledge base. We identify communication cost in rule execution to be the primary basis for decomposing a global knowledge base into clusters for their allocation to sites. We show that the problem of optimal allocation is a 0–1 quadratic programming problem, which has prohhbitive execution times for large knowledge bases. We propose an efficient heuristic algorithm for rule allocation and study its performance experimentally. We represent a knowledge base as a hierarchy and characterize it in terms of height and inherent clusters with overlaps. The experimental results of the heuristic algorithm on random hierarchies as well as on hierarchies with varying heights and overlaps are seen to be close to the optimal solution.

QUERY EVALUATION IN PARTITIONED DISJUNCTIVE DEDUCTIVE DATABASES

Query evaluation in disjunctive deductive databases is in general computationally hard. The class of databases for which the process is tractable is severely limited. The complexity of the process depends on the structure of the database as well as on the type of query being evaluated. In this paper we study the issue of simplified query processing in disjunctive deductive databases. We address the possibility of evaluating general queries by independently processing their atomic components and describe the class of databases for which this approach is possible. We also discuss the issue of dividing a disjunctive deductive database into a set of disjoint components then answering queries and computing database completions by combining the results obtained against the individual components. Some practical special cases are considered. The methods developed in this paper can be utilized to introduce parallelism into the query evaluation process.

Mixed integer programming methods for computing nonmonotonic deductive databases

Though the declarative semantics of both explicit and nonmonotonic negation in logic programs has been studied extensively, relatively little work has been done on computation and implementation of these semantics. In this paper, we study three different approaches to computing stable models of logic programs based on mixed integer linear programming methods for automated deduction introduced by R. Jeroslow. We subsequently discuss the relative efficiency of these algorithms. The results of experiments with a prototype compiler implemented by us tend to confirm our theoretical discussion. In contrast to resolution, the mixed integer programming methodology is both fully declarative and handles reuse of old computations gracefully. We also introduce, compare, implement, and experiment with linear constraints corresponding to four semantics for “explicit” negation in logic programs: the four-valued annotated semantics [Blair and Subrahmanian 1989], the Gelfond-Lifschitz semantics [1990], the over-determined models [Grant and Subrahmanian 1989], the Gelfond-Lifschitz semantics [1990], the over-determined models [Grant and Subrahmanian 1990], and the classical logic semantics. Gelfond and Lifschitz[1990] argue for simultaneous use of two modes of negation in logic programs, “classical” and “nonmonotonic,” so we give algorithms for computing “answer sets” for such logic programs too.

How to forget the past without repeating it

Bottom-up evaluation of deductive database programs has the advantage that it avoids repeated computations by storing all intermediate results and replacing recomputation by table lookup. However, in general, storing all intermediate results for the duration of a computation wastes space. In this paper, we propose an evaluation scheme that avoids recomputation, yet for a fairly general class of programs at any given time stores only a small subset of the facts generated. The results constitute a significant first step in compile-time garbage collection for bottom-up evaluation of deductive database programs.

Intuitionistic interpretation of deductive databases with incomplete information

The aim of this paper is to build the relationship between deductive databases with incomplete information and hypothetical reasoning using embedded implications. We first consider the semantics of deductive databases with incomplete information in the form of null values. We motivate query answering against a deductive database with nulls as the problem of extracting the maximal information from a (deductive) database in response to queries, and formalize this in the form of conditional answers in a (syntactic) higher-order logic. We give a fixpoint semantics to deductive databases with nulls, and examine the relationship between existing recursive query processing techniques and the proof procedure for deductive databases with nulls. We then examine hypothetical reasoning using embedded implications and develop an intuitionistic model semantics for embedded implications with integrity constraints. Finally, we illustrate by example a method for transforming embedded implications into deductive databases with nulls. This result shows that the important functionality of hypothetical reasoning can be implemented within the framework of deductive databases with null values.

A database of groups of prime‐power order

AbstractThe design, implementation and performance of TwoGroups, a deductive database for the 58,761 groups of order 2n, (n ≤ 8), is described. The system is implemented in NU‐Prolog, a Prolog system with built‐in functions for creating and using deductive databases. TwoGroups has a set‐theoretic query language, which provides users with a familiar notation to access the data. The paper describes the data and its representation, the set‐theoretic query language, its translator and optimiser, and the experiments on the performance of the database.

A three-valued semantics for deductive databases and logic programs

This paper proposes two principles, justifiability and minimal undefinedness , for a three-valued model-theoretic approach to semantics of logic programs and deductive databases (also called disjunctive logic programs). The former is intimately related to the concept of labeling-based justification in Doyle's truth maintenance system while the latter requires the use of the truth value undefined only when it is necessary. We examine the question why and in what circumstances the undefined is needed under these two principles. We show that these two principles yield a declarative semantics for deductive databases and logic programs, which is called the regular model semantics . Program properties in this semantics are analyzed and results obtained concerning the relationship among regular, stable, and well-founded semantics, which show that the regular model semantics is a natural extension of the latter two semantics.

A selective enumeration based heuristic for rule allocation

A distributed deductive database system consists of many autonomous deductive database systems connected by a computer network to facilitate sharing of both database and rules. The design of a distributed deductive database system differs from the design of conventional nondistributed deductive database systems. The former requires design of distribution of both the database and rulebase, whereas the latter requires design of data distribution. The paper addresses the problem of allocating rules in a distributed deductive database system, where rulebase (a collection of rules) and database are shared across autonomous sites. The problem is relevant and challenging in applications involving large rulebases. We have identified communication cost as the primary consideration in allocation of rules. The optimal allocation of rules over a set of sites is an NP-complete problem. It has prohibitive execution time for large rulebases. In this paper, we propose an efficient heuristic algorithm for non-replicated rule allocation and study its performance vis-à-vis the enumerative algorithm for optimal allocation. A heuristic for replicated allocation of rules is also presented which starts from the non-replicated allocation.

A new fragmentation scheme for recursive query processing

Data fragmentation is a well-known technique used in distributed database design to support efficient query processing. In deductive databases and knowledge bases, one distinct operation is the computation of transitive closures. In this paper, we employ a fragmentation approach to deductive databases. Fragmentation techniques are investigated in distributed database environments for parallel recursive query processing. Based on the novel fragmentation scheme, a new parallel transitive closure algorithm is proposed. Compared with previous algorithms, our algorithm can explore a much higher level of parallelism with very little communication overhead increase.

Mid-term report on the Arden Syntax in a clinical event monitor

In implementing a clinical event monitor (CEM), a decision-support system, we worked with an existing repository of clinical data (Keystone), fed from ancillary systems using HL7. The rules are written in the Arden Syntax, an ASTM standard for expressing medical knowledge as medical logic modules (MLMs). The Arden Syntax leaves unspecified the clinical data model and deductive database access language; we briefly describe our query language and the related medical concepts dictionary (MCD). This paper gives an overview of our implementation of the Arden Syntax, the MCD and the deductive database access language, with reasons for the major design decisions. Overall, less than a quarter of the development effort has gone into the Arden compiler and interpreter; the rest has focused on accessing the data and integrating with other systems. We feel that the Arden Syntax has proved its worth in writing rules; effort should now be focused on medical vocabularies and data models.

Deriving incremental production rules for deductive data

We show that the production rule mechanism provided by active database systems can be used to quickly and easily implement the logic rule interface of deductive database systems. Deductive rules specify derived relations using Datalog with built-in predicates and stratified negation; the deductive rules are compiled automatically into production rules. We present a materialized approach, in which the derived relations are stored in the database and the production rules automatically and incrementally propagate base relation changes to the derived relations. We also present a non-materialized approach, in which the production rules compute the derived relations on demand.

A pragmatic approach to software synthesis

We present a practice oriented tool for software synthesis that supports the interface-correct configuration of complex systems from a library of reusable software components. Besides simply checking the interface-correctness of a link by means of type constraints, the tool is also designed to propose software components for solving a (loosely) specified problem within a certain context. In particular, it identifies possible interfacing modules that in case of an interface-conflict may serve for the right conversion, transformation or parameter configuration. We illustrate our tool, which is based on the deductive database system LOLA , in three application specific settings.

A deductive database system for analyzing human nucleotide sequence data

The analysis of the human genome is one of the most significant topics in both biology and medical science. There is a growing need for a well-designed database system for searching and analyzing the human genome data. We developed a deductive database system to search and analyze nucleotide sequence data derived from the GenBank primates data. A deductive database system is a next generation one and it contains an inference mechanism that can handle problems beyond the capabilities of classical database systems. Database queries are described in logical rules. These rules are simple even for molecular biologists who are not experts in computer programs because they are declarative and do not require the procedural commands that are usually used in computer programs. Furthermore, queries based on logical rules are powerful enough to express complicated biological problems. Particularly, recursive rules are suitable for examining secondary structures of nucleotide sequences. In our analysis of TfR's IRE, we noted five stem-and-loop structures.