Jonathan I. Maletic

Recovering documentation-to-source-code traceability links using latent semantic indexing

An information retrieval technique, latent semantic indexing, is used to automatically identify traceability links from system documentation to program source code. The results of two experiments to identify links in existing software systems (i.e., the LEDA library, and Albergate) are presented. These results are compared with other similar type experimental results of traceability link identification using different types of information retrieval techniques. The method presented proves to give good results by comparison and additionally it is a low cost, highly flexible method to apply with regards to preprocessing and/or parsing of the source code and documentation.

An information retrieval approach to concept location in source code

Concept location identifies parts of a software system that implement a specific concept that originates from the problem or the solution domain. Concept location is a very common software engineering activity that directly supports software maintenance and evolution tasks such as incremental change and reverse engineering. This work addresses the problem of concept location using an advanced information retrieval method, Latent Semantic Indexing (LSI). LSI is used to map concepts expressed in natural language by the programmer to the relevant parts of the source code. Results of a case study on NCSA Mosaic are presented and compared with previously published results of other static methods for concept location.

A survey and taxonomy of approaches for mining software repositories in the context of software evolution

A comprehensive literature survey on approaches for mining software repositories (MSR) in the context of software evolution is presented. In particular, this survey deals with those investigations that examine multiple versions of software artifacts or other temporal information. A taxonomy is derived from the analysis of this literature and presents the work via four dimensions: the type of software repositories mined (what), the purpose (why), the adopted/invented methodology used (how), and the evaluation method (quality). The taxonomy is demonstrated to be expressive (i.e., capable of representing a wide spectrum of MSR investigations) and effective (i.e., facilitates similarities and comparisons of MSR investigations). Lastly, a number of open research issues in MSR that require further investigation are identified.

3D representations for software visualization

The paper presents a new 3D representation for visualizing large software systems. The origins of this representation can be directly traced to the SeeSoft metaphor. This work extends these visualization mechanisms by utilizing the third dimension, texture, abstraction mechanism, and by supporting new manipulation techniques and user interfaces. By utilizing a 3D representation we can better represent higher dimensional data than previous 2D views. An overview of our prototype tool and its basic functionality is given. Applications of this method to particular software engineering tasks are also discussed.

RECOVERY OF TRACEABILITY LINKS BETWEEN SOFTWARE DOCUMENTATION AND SOURCE CODE

An approach for the semi-automated recovery of traceability links between software documentation and source code is presented. The methodology is based on the application of information retrieval techniques to extract and analyze the semantic information from the source code and associated documentation. A semi-automatic process is defined based on the proposed methodology. The paper advocates the use of latent semantic indexing (LSI) as the supporting information retrieval technique. Two case studies using existing software are presented comparing this approach with others. The case studies show positive results for the proposed approach, especially considering the flexibility of the methods used.

An XML-based lightweight C++ fact extractor

A lightweight fact extractor is presented that utilizes XML tools, such as XPath and XSLT to extract static information from C++ source code programs. The source code is first converted into an XML representation, srcML, to facilitate the use of a wide variety of XML tools. The method is deemed lightweight because only a partial parsing of the source is done. Additionally, the technique is quite robust and can be applied to incomplete and noncompilable source code. The trade off to this approach is that queries on some low level details cannot be directly addressed. This approach is applied to a fact extractor benchmark as comparison with other, heavier weight, fact extractors. Fact extractors are widely used to support understanding tasks associated with maintenance, reverse engineering and various other software engineering tasks.

A task oriented view of software visualization

A number of taxonomies to classify and categorize software visualization systems have been proposed in the past. Most notable are those presented by Price (1993) and Roman (1993). While these taxonomies are an accurate representation of software visualization issues, they are somewhat skewed with respect to current research areas on software visualization. We revisit this important work and propose a number of re-alignments with respect to addressing the software engineering tasks of large-scale development and maintenance. We propose a framework to emphasize the general tasks of understanding and analysis during development and maintenance of large-scale software systems. Five dimensions relating to the what, where, how, who, and why of software visualization make up this framework. The focus of this work is not so much as to classify software visualization system, but to point out the need for matching the method with the task. Finally, a number of software visualization systems are examined under our framework to highlight the particular problems each addresses.

Assessing the Comprehension of UML Class Diagrams via Eye Tracking

Eye-tracking equipment is used to assess how well a subject comprehends UML class diagrams. The results of a study are presented in which eye movements are captured in a non-obtrusive manner as users performed various comprehension tasks on UML class diagrams. The goal of the study is to identify specific characteristics of UML class diagrams, such as layout, color, and stereotype usage that are most effective for supporting a given task. Results indicate subjects have a variation in the eye movements (i.e., how the subjects navigate the diagram) depending on their UML expertise and software-design ability to solve the given task. Layouts with additional semantic information about the design were found to be most effective and the use of class stereotypes seems to play a substantial role in comprehension of these diagrams.

Source code files as structured documents

A means to add explicit structure to program source code is presented. XML is used to augment source code with syntactic information from the parse tree. More importantly, comments and formatting are preserved and identified for future use by development environments and program comprehension tools. The focus is to construct a document representation in XML instead of a more traditional data representation of the source code. This type of representation supports a programmer centric view of the source rather than a compiler centric view. Our representation is made relevant with respect to other research on XML representations of parse trees and program code. The highlights of the representation are presented and the use of queries and transformations discussed.

Using latent semantic analysis to identify similarities in source code to support program understanding

The paper describes the results of applying Latent Semantic Analysis (LSA), an advanced information retrieval method, to program source code and associated documentation. Latent semantic analysis is a corpus based statistical method for inducing and representing aspects of the meanings of words and passages (of natural language) reflective in their usage. This methodology is assessed for application to the domain of software components (i.e., source code and its accompanying documentation). Here LSA is used as the basis to cluster software components. This clustering is used to assist in the understanding of a nontrivial software system, namely a version of Mosaic. Applying latent semantic analysis to the domain of source code and internal documentation for the support of program understanding is a new application of this method and a departure from the normal application domain of natural language.