Automated Trace Retrieval

Software Traceability is defined by the Center of Excellence for Software Traceability as the "ability to interrelate any uniquely identifiable software engineering artifact to any other, maintain required links over time, and use the resulting network to answer questions of both the software product and its development process". It supports numerous critical software engineering tasks such as certification of software intensive products, compliance verification, change management, design allocation, and requirements validation. Unfortunately the actual practice of creating and maintaining traceability links is effort intensive and error prone. Researchers at SAREC engage in many facets of Software Traceability research including projects to automate the process of creating and maintaining traceability links using machine learning and data mining techniques.

Over the past 15 years, Software Traceability researchers have investigated a wide variety of information retrieval and machine learning techniques to automated the traceability process. For the most part, these techniques hit a glass ceiling, achieving high recall at relatively low precision. At the same time, human domain experts are well capable of reasoning about whether two artifacts are related or not. This work is our quest to mimic the human reasoning process during the trace creation process. To do so we need the ability to dynamically build knowledge bases of related concepts for highly-technical systems engineering domains (in itself a daunting task), and then reasoning over the facts in the knowledge base to determine when a trace link should be created.

Recent publications include:

• Mona Rahimi, Jane Cleland-Huang: Evolving software trace links between requirements and source code. Empirical Software Engineering 23(4): 2198-2231 (2018)
• Michael Rath, Jacob Rendall, Jin L. C. Guo, Jane Cleland-Huang, Patrick Mäder:Traceability in the wild: automatically augmenting incomplete trace links. ICSE 2018: 834-845
• Salome Maro, Jan-Philipp Steghöfer, Jane Huffman Hayes, Jane Cleland-Huang, Miroslaw Staron: Vetting Automatically Generated Trace Links: What Information is Useful to Human Analysts? RE 2018: 52-63
• Jin Guo, Jinghui Cheng, Jane Cleland-Huang: Semantically enhanced software traceability using deep learning techniques. ICSE 2017: 3-14
• Jin Guo, Natawut Monaikul, Cody Plepel, Jane Cleland-Huang: Towards an intelligent domain-specific traceability solution. ASE 2014: 755-766
• Jin Guo, Jane Cleland-Huang, Brian Berenbach: Foundations for an expert system in domain-specific traceability. RE 2013: 42-51
• Jin Guo, Natawut Monaikul, Jane Cleland-Huang: Trace links explained: An automated approach for generating rationales. RE 2015: 202-207

Researchers

• Jin Guo, PhD Student, Topic: Intelligent Domain-Specific Traceability (graduated)
• Mona Rahimi, PhD Student, Topic: Trace Link Evolution (graduated)
• Jinfeng Lin, Topic: High-Performance Traceability at Scale (current)
• Yalin Liu, Topic: Semantically Embued Software Project Queries (current)
• Jacob Rendall, Topic: Configurable, Multi-Objective Traceability (current)
• Dr. Jane Cleland-Huang, SAREC Director, Universith of Notre Dame (current)

Funded Projects

• $444,326 SHF: Medium: RUI: Collaborative Research: Advanced Traceability for Composing Product Line Safety Cases, 07/01/2015 (4 years) (Collaborator Robyn Lutz, Iowa State University). Total award $850,000.
• $249,973 CI-EN: RUI: Collaborative Research: TraceLab Community Infrastructure for Replication, Collaboration, and Innovation, $249,973, 06/01/2015. (3 years) Total award approx. $800,000.
• $529,383 SHF: Small: RUI: Generating High Quality Trace Links through Intelligent Composition of Tracing Features, 08/01/13
• $2,000,000, 2010-2013, National Science Foundation, MRI-R2: Development of a Software Traceability Instrument to Facilitate and Empower Traceability Research and Technology Transfer, PI: Jane Cleland-Huang ($1,500,000), Co-PI Denys Poshyvanyk, College of William and Mary ($250,000), Jonathan Maletic ($250,000), CNS: 0959924.
• $400,000, 2005-2010, CAREER: Goal Centric Traceability for Managing Systemic Requirements, Software and Hardware Foundation, Computing Processes and Artifacts, CPA 0447594.
• $270,000, 2003-2006, National Science Foundation, SEL, PI: Jane Cleland-Huang, Co-PI: Raffaella Settimi, SEL 00306303.
• $220,000, 2007-2010, Siemens Corporate Research, Automated Traceability. (Additional funding committed for fiscal year 2012 for a project on Automated Traceability in Mechatronics Systems).
• $250,000, 2008-2011, National Science Foundation, CPA-SEL-T: Collaborative Research: Traceability+: A Service Oriented Framework to Support Value-Added Software Traceability, PI: Jane Cleland-Huang ($250,000), Jane Huffman Hayes, University of Kentucky ($250,000), Jonathan Maletic, Kent State University ($250,000), CPA 0810924.