RETE Algorithm

One of the major challenges that current P systems simulators have to deal with is to be as efficient as possible. A P system is syntactically described as a membrane structure delimiting regions where multisets of objects evolve by means of evolution rules. According to that, on each computation step, the applicability of the rules for the current P system configuration must be calculated. In this paper we extend previous works that use Rete-based simulation algorithm in order to improve the time consumed during the checking phase in the selection of rules. A new approach is presented, oriented to the acceleration of Population Dynamics P Systems simulations.

We focus on the partitioning of rules for parallel matching in a production system. Our approach, called PARTES, applies the man-cut technique t o a Dataflow Discrimination Network (DDN) that represents th? antecedents of the rules. Our goal is t o rnaxiinize sharing of memory nodes within a partition while minimizing duplication of nodes across partitions. We illustrate the technique using a cost model based on shared memory nodes defined in a DDN created by the R E T E algorithin. A generalization t o any other algorithm utilizing a DDN as straightforward. We provide a performance analysis to show the effectiveness of PARTES.

Software development methodology involves many activities and processes that are carried out when building a software system. There are a lot of available methodologies; one should be suitable for a software project. However, deciding which methodology to be applied requires some assessments on the project nature and characteristics at the early project cycle. This research focuses on finding the relationship between software project factors and methodologies, and further developing a tool that can help software practitioners in choosing the most suitable methodology. Before the tool is developed, a review on several lifecycle models has been carried out in order to examine the relevant software factors. Factors like project size, complexity, requirement stability, duration, performance requirement, modularization, project team, and criticality have the main impact on some methodologies. The relationship between methodologies and software factors is formulated using rule-based approach, and managed properly by a sequence of steps; identifying the initial selection of input variables, counterexamples of bad sub rules, pruning the variables, merging categories, and identifying hypothetical examples. RETE algorithm has been chosen as the problem solving technique in managing the rules. The tool is able to help software practitioners in early decision making process to use appropriate methodology in their software project.

Software development methodology involves many activities and processes that are carried out when building a software system. There are a lot of available methodologies; one should be suitable for a software project. However, deciding which methodology to be applied requires some assessments on the project nature and characteristics at the early project cycle. This research focuses on finding the relationship between software project factors and methodologies, and further developing a tool that can help software practitioners in choosing the most suitable methodology. Before the tool is developed, a review on several lifecycle models has been carried out in order to examine the relevant software factors. Factors like project size, complexity, requirement stability, duration, performance requirement, modularization, project team, and criticality have the main impact on some methodologies. The relationship between methodologies and software factors is formulated using rule-based approach, and managed properly by a sequence of steps; identifying the initial selection of input variables, counterexamples of bad sub rules, pruning the variables, merging categories, and identifying hypothetical examples. RETE algorithm has been chosen as the problem solving technique in managing the rules. The tool is able to help software practitioners in early decision making process to use appropriate methodology in their software project.

We focus on the partitioning of rules for parallel matching in a production system. Our approach, called PARTES, applies the man-cut technique t o a Dataflow Discrimination Network (DDN) that represents th? antecedents of the rules. Our goal is t o rnaxiinize sharing of memory nodes within a partition while minimizing duplication of nodes across partitions. We illustrate the technique using a cost model based on shared memory nodes defined in a DDN created by the R E T E algorithin. A generalization t o any other algorithm utilizing a DDN as straightforward. We provide a performance analysis to show the effectiveness of PARTES.

Cardiovascular diseases become number one cause of death among people. People must be guided based on proper diet. This paper presented current issues on CVDs and proposed He arTy. HearTy is an expert system embedded on Android Operating System for the screening of cardiovascular diseases or CVDs. It is a rule-based expert system on Android phone for screening the risk as well as the likelihood of having CVDs based on the risk factors present and symptoms felt by the patient. The system covers five types of CVDs which includes Stroke, Coronary Heart Disease, Heart Attack, Heart Failure, and Peripheral Vascular Disease. The system applies Backward Chaining and RETE Algorithm. Backward Chaining is a type of reasoning for the inference engine which starts with a list of goals and RETE Algorithm is a pattern matching algorithm which identifies a rule that satisfy the pattern in order to generate the outputs. Result of the study showed accuracy of 68.75%, 66.59% sensitivity, error rate of 31.5% and 71.565% specificity which led to require further development of the developed software.