An Ontology Based Dynamic Traceabililty Of Requirements

Requirements traceability supports many of the key activities in software engineering,and has an irreplaceable role to understand the software system,and its importance has been widely recognized.When the project requirements change,you can find the related code to change it according to the requirements traceability chain.As the software developer is of high liquidity,for a long life cycle project,establishment of the traceability of the requirements to the code is crucial,and it can help the subsequent engineers to better develop and maintain software.Requirements traceability associates the requirements and code implementation in the life cycle of the software system.Requirements traceability in the life cycle of the software system and the associated code implementation are very important in software project management and effective methods to improve the quality of software.Requirements traceability can be divided into two categories: static requirements traceability and dynamic requirements traceability.Static requirements traceability needs to be handled manually,and when requirements changes,it needs to modified manually.This makes the static requirements traceability time-consuming,laborious and cannot meet the requirements for tracking effect.Dynamic Requirements Tracking can automatically establish traceability relationships,and has the ability to track the relationship between requirements and changes,to remind the updates automatically,which fortunately make up for static requirements tracking defects.Most of current requirements for dynamic tracking research are based on the dynamic requirements traceability information retrieval(IR)models.The accuracy of the models is based on the calculation of matching index words when requirements and code using different indexing,even when words between indexing terms are similar or related,match will be lost.This seriously influences the accuracy of dynamic requirements traceability.Because that the dynamic requirements traceability process has the latent semantic matching problem and produces the low accuracy,we propose a new dynamic requirements tracing mechanism based on ontology in the existing requirements traceability technology and ontology matching technology.Our method makes the standardization of requirement documents and codes through the construction of requirement ontology and code ontology,and matching the requirements ontology nodes and code ontology nodes,and mutual establishesrequirements traceability relationship between requirements and codes.The main contents of this thesis include:(1)Analysis of the original requirements decomposition method and proposing requirements to ontology mapping model and mapping rules,and how to use the ontology to verify the requirements decomposition.(2)Comparison of object-oriented language similarities with the code ontology,and propose the code to ontology mapping model and mapping rules.(3)On the basis of the requirements ontology model and code ontology model,we establish the requirements and the code association model.This model implements semantic similarity calculation between requirement ontology node and code ontology node,and still achieving the associated ontology node similarity calculation and internal similarity calculation.Association rules are broken down into requirements to the children requirements to construct the ontology,at the same time corresponding functional group code to construct ontology for code,through the node matches between the two ontologies,and finally find the functional requirements in the code implementation classes and methods,in order to achieve requirements traceability purposes.Lastly experimental results are conducted to validate the effectiveness and feasibility of the tracking method.The results show that the model is feasible and effective,and compared with the conventional model based on information retrieval;the requirements of the model tracking precision and recall rate have been improved.