Improving IV&V Using Information Retrieval and Text Mining

  • Hayes, Jane (PI)
  • Dekhtyar, Alexander (CoI)

Grants and Contracts Details


The fundamental purpose of IV&V is to insure that the right processes have been used to build the right system. To that end, we must verifY that the approved processes and artifacts are guiding development during each lifecycle phase as well as validate that all requirements have been implemented at the end of the lifecycle. A requirements traceability matrix (RTM) is a prerequisite for both of these. Current approaches to after-the-fact requirements tracing include manual searching for key phrases in high and low level documents, keyword hierarchy development and assignment to all elements in high and low level documents, as well as keyword matching. These approaches are time consuming and still require the IV&V analyst to evaluate large lists of poor quality candidate links (many are not links). Our ultimate goal is to significantly improve the quality and speed of the requirements tracing process and alleviate its tedium for the IV&V analysts. To that extent, we propose to continue our research on applicability of Information Retrieval (IR) methods to requirements tracing. We are planning to concentrate on more sophisticated techniques that better reflect the nature of the requirements tracing problem. We will develop new IR-based methods for candidate link generation that take better advantage of our knowledge about requirements tracing problem. We will implement more complex methods for processing analyst feedback during the requirements tracing process. We will conduct a series of experiments on significantly larger datasets then before to ensure the scalability and robustness of proposed techniques. We will explore the problems of tracing non-textual artifacts. The methods developed within this project will be incorporated into RETRO (REquirements TRacing On-target), a special-purpose, standalone requirements tracing tool we are developing.
Effective start/end date9/1/048/31/05


Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.