Narrow Results

In this paper, we present a method for improving the performance of classical symbolic rules. This is achieved by introducing a type of hybrid rules, called neurules, which integrate neurocomputing into the symbolic framework of production rules. Neurules are produced by converting existing symbolic rules. Each neurule is considered as an adaline unit, where weights are considered as significance factors. Each significance factor represents the significance of the associated condition in drawing the conclusion. A rule is fired when the corresponding adaline output becomes active. This significantly reduces the size of the rule base and, due to a number of heuristics used in the inference process, increases efficiency of the inferences.

Neurules are a kind of hybrid rules integrating neurocomputing and production rules. Each neurule is represented as an adaline unit. Thus, the corresponding neurule base consists of a number of autonomous adaline units (neurules). Due to this fact, a modular and natural knowledge base is constructed, in contrast to existing connectionist knowledge bases. In this paper, we present a method for generating neurules from empirical data. To overcome the difficulty of the adaline unit to classify non-separable training examples, the notion of 'closeness' between training examples is introduced. In case of a training failure, two subsets of 'close' examples are produced from the initial training set and a copy of the neurule for each subset is trained. Failure of training any copy, leads to production of further subsets as far as success is achieved.

This paper presents the design and implementation of a knowledge-based tool for performance tuning of the UNIX operating system. The tool, called BMS, provides an intelligent support and maintenance for identifying performance bottlenecks in UNIX and recommending solutions to the problems. Currently, it handles problems in UNIX resource management, such as memory utilization, disk utilization, CPU scheduling and I/O devices. BMS has been implemented in the EXSYS environment and tested on UNIX V.3. Preliminary results have indicated that such a knowledge-based tool to operating system performance tuning (1) is viable; (2) increases the productivity of system maintenance personnel and reduces the cost of training; and (3) offers a better service to operating system users by providing prompt recommendations to solutions of their system performance problems.

Selection of die components is an important activity for the design of a compound die. The traditional methods of carrying out this task are complex, time-consuming, and skill-intensive. This paper describes the research work involved in the development of an expert system for selection of components of compound die. The proposed system is constructed in the form of eight modules. For development of system modules, domain knowledge acquired from different sources of knowledge acquisition is refined and then framed in the form of "IF-Then" variety of production rules. System modules are coded in AutoLISP language and user interface is created using Visual Basic (VB). The system is capable of imparting expert advices for selection of type and size of all major die components including die block, die gages, stripper, stripper springs, punches, punch plate, die-set, fasteners, and knockout bar. The proposed system can be implemented on a PC having VB and AutoCAD softwares. Therefore, the implementation cost of proposed system is low and hence easily affordable for small scale stampingindustries.

This paper presents an extension to a database rule language compiler to support parallelism. The rule language is called RDL//C and is based on an extended version of tuple relational calculus. The grain of parallelism introduced into RDL//C is the production rule. The user is made aware of parallel processing by having to specify in a sub-language of control which sets of rules are to run in parallel and in which order. From the user view point, the parallelism is managed as a partial order over a set of rules. The compiler translates rule programs into C based applications which run over the DBMS. Our implementation is over a set of UNIX workstations which corresponds to a MIMD shared-nothing architecture. Parallelism is achieved by having the application connect to different DBMS servers residing on workstations on a local network. Initial performance results are presented. They show the advantages and limitations of parallel execution in this architecture. The main contributions of the paper are (i) the definition of a sub-language of control to specify a parallel execution of a rule program and (it) to show how parallel extensions can be brought to a rule language compiler using a standard distributed environment with a relational DBMS.