Bøger af C Srinivas

Software reuse may be defined as the process of developing new software systems by making use of existing software components. Software reusability eliminates the need for developing a software system from the scratch. It also reduces the productivity cost. For example, the software component chosen for reuse may be a software requirement document or software test document or software code modules or programs. Retrieval of appropriate and right software component from the software component repository is an important task. This task has gained importance from the academia, researchers and also from industry perspective in order to reduce the costs. In principle, the retrieval of any software component from the software repository requires a search algorithm. This search algorithm is expected to retrieve the software component with the help of features specified by the user in his query. But, searching requires various parameters to be specified well ahead as part of user query. Therefore, searching task has its own limitations. An alternative approach could be the use of data mining and/or machine learning principles in retrieving appropriate software component with utmost similarity. Clustering of software components means it can group all software components with similar features into one group. From perspective of software engineering, all the components within the same cluster must have high cohesion and low coupling. Clustering reduces the time complexity for searching as all similar components are placed into one group. Clustering is not any one specific algorithm that we can stick firm to, but it must be viewed as the general task to be solved. For example, document clustering is one of the main themes in text mining. It refers to the process of grouping documents with similar contents or topics into clusters. This improves both availability and reliability of text mining applications such as information retrieval, classifying text, summarizing document sets. In similar lines, classification task can also facilitate identification of right and fitting component for reuse from among the existing components available in the repository.