Based On Ontology Component Annotating Algorithm Research

Nowadays, a great quantity of component library and scattered componentsresources are emerged on the Internet, however, these components resources present partof the focus, general dispersion characteristics, as well as the different description of thecomponent information among the various component library, as a result, it difficult forusers to retrieve the required components quickly and accurately. To solve this problem,it is a great practical significance of making use of the abundant knowledge of ontologiesfor semantic annotation and shielding the differences among the different componentlibraries.According to the study of the current ontology, domain ontology and semanticannotation technology, with a view to the features of the domestic mainstreamcomponent library and the need of real-world applications. The component concept iscollected by the domain experts who analysis the relationship between these conceptsand the attributes of the concept. The Protégé tool is used to build the completecomponent ontology knowledge base, The wealth knowledge of the componentontologies is used to mark components. Based on the characteristics of the descriptioninformation of components, we will divide the label of components into two parts, i.e.the basics information annotation and the label of descriptions of component function.this article is focused on the label of descriptions of component information. At theprocess of component labeling, it mainly related to the following key technologies:(1)build domain ontologies;(2) segmentation based on natural language processingtechnologies;(3) associated word merge based on semantic classification;(4) calculatethe semantic similarity of concepts and ontologies with improved edit distance andGoogle distance.In view of the current status of the component library, this article present acomponents labeling algorithm-OBSCA (Ontology-Based Software Component Annotation, OBSCA) that based on ontology component library, which depended on thesemantic environment description of the knowledge of domain ontology and thecomponent library components structure. Component information marked by the use ofrich semantic knowledge component ontologies, which not only shield the differencesamong the descriptions of component information by the component library, but alsomake it convenient for users to access to the components required in the softwaredevelopment process quickly and efficient.This article use comprehensive experimental sample data, and design completeexperimental process, as well as analysis the experimental results detailed, which showthat OBSCA algorithm not only guarantee the correctness, feasibility and robustness ofcomponents mark, but also the average recall rate, the average composite index and theaverage precision of the technical evaluation achieve the desired objectives in thepractical application of the algorithm.