Traceability of software artefacts has been recognised as an important factor for supporting various activities in the software system development process. In general, the objective of traceability is to improve the quality of software systems. More specifically, traceability information can be used to support the analysis of implications and integration of changes that occur software systems; the maintenance and evolution of software systems; the reuse of software system components by identifying and comparing requirements of new and existing systems; the testing of software system components; and system inspection, by indicating alternatives and compromises made during development. Traceability enables system acceptance by allowing users to better understand the system and contributes to clear and consistent system documentation. Over the last few years, the software and system engineering communities have developed a large number of approaches and techniques to address various aspects of traceability. Research into software traceability has been mainly concerned with the study and definition of different types of traceability relations; support for the generation of traceability relations; development of architectures, tools, and environments for the representation and maintenance of traceability relations; and empirical investigations into organisational practices regarding the establishment and deployment of traceability relations in the software development life cycle. However, despite its importance and the work resulted from numerous years of research, empirical studies of traceability needs and practices in industrial organisations have indicated that traceability support is not always satisfactory. As a result, traceability is rarely established in existing industrial settings. In this article, we present a roadmap of research and practices related to software traceability and identify issues that are still open for further research. Our roadmap is organised according to the main topics that have been the focus of software traceability research.

The paper summarizes the authors ’ current research on supporting model-to-model traceability. The authors present a graph theoretic definition of what they mean by models and traceability links. This definition is realized by the use of XML technologies to represent the models and traceability links. Practical means to represent different types of system models (e.g., source code and design) using XML are discussed. Traceability links are also implemented using XML technologies in an efficient and scalable manner. The evolution of the system, along with the traceability links, is supported by a fine-grained versioning technique. This allows for versioning and differencing of specific elements of the models versus just lines or whole files.

Software traceability is a recognized challenge in software development that can be ameliorated with requirements management tools. Traceability information can be used in a number of different software engineering activities such as software change impact analysis and testing One main challenge in the automation of software testing is mapping modeling concepts to code concepts. The level of granularity and the semantics supported by available requirements management tools do not, however, fully support such mapping, or more sophisticated requirement change impact analysis. Scenarios have been used as an alternative (and sometimes complementary) way to express requirements and system behavior throughout the phases of software development. Scenarios are used with different representation and semantics across software phases, and these can be related. This paper argues for exploring scenarios as one means for tracing requirements to code, and using this information to leverage automation of activities that benefit from traceability such as change impact analysis and software testing.

Maintaining traceability links among software artifacts is particularly important for many software engineering tasks. Even though automatic traceability link recovery tools are successful in identifying the semantic connections among software artifacts produced during software development, no existing traceability link management approach can effectively and automatically deal with software evolution. We propose a technique to automatically manage traceability link evolution and update the links in evolving software. Our novel technique, called incremental Latent Semantic Indexing (iLSI), allows the fast and low-cost LSI computation for the update of traceability links by analyzing the changes to software artifacts and by re-using the results from previous LSI computation before the changes. We present our iLSI technique, and describe a complete automatic traceability link evolution management tool, TLEM, that is capable of interactively and quickly updating traceability links in the presence of evolving software artifacts. We report on our empirical evaluation with various experimental studies to assess the performance and usefulness of our approach.

An approach to support the sustained evolution of traceability links is proposed and outlined. A fine-grained differencing approach on the link endpoints is used to maintain the links in a scalable manner. Here scalable refers to large software systems with thousands of links. Details of the link model and representation are given followed by the process used to evolve traceability links.

Model Driven Architecture adopts a visual approach to software development. The main development activities are the construction of visual (typically Unified Modelling Language (UML)) models and the transformation of source models into target models, including application code generation. The use of visual models to produce application code often starts at the design (Platform Independent Model) level, and whereas business processes (Computation Independent Models (CIM)) have lately been considered, they are not used in MDA to either derive design models or application code. This paper enhances the MDA process by considering the early stages of software development that pertain to problem domain analysis. We argue that problem domain analysis and modelling can form valuable input to the more formal MDA phases at the CIM and PIM levels. We propose the use of a visual notation that allows informal modelling of domain-based concepts. Modelling at this stage using the proposed notation is geared to support involvement of non-technical business stakeholders whilst feeding into business process modelling at the CIM phase.

A query language for traceability is proposed and presented. The language, TQL, is based in XML and supports queries across multiple artifacts and multiple traceability link types. A number of primitives are defined to allow complex queries to be constructed and executed. Example queries are presented in the context of traceability questions. The technical details of the language and issues of implementation are discussed. 1