Software Requirements Specification Quality Research Articles

A Defect Detection Method for the Primary Stage of Software Development

In the early stage of software development, a software requirements specification (SRS) is essential, and whether the requirements are clear and explicit is the key. However, due to various reasons, there may be a large number of misunderstandings. To generate high-quality software requirements specifications, numerous researchers have developed a variety of ways to improve the quality of SRS. In this paper, we propose a questions extraction method based on SRS elements decomposition, which evaluates the quality of SRS in the form of numerical indicators. The proposed method not only evaluates the quality of SRSs but also helps in the detection of defects, especially the description problem and omission defects in SRSs. To verify the effectiveness of the proposed method, we conducted a controlled experiment to compare the ability of checklist-based review (CBR) and the proposed method in the SRS review. The CBR is a classic method of reviewing SRS defects. After a lot of practice and improvement for a long time, CBR has excellent review ability in improving the quality of software requirements specifications. The experimental results with 40 graduate students majoring in software engineering confirmed the effectiveness and advantages of the proposed method. However, the shortcomings and deficiencies of the proposed method are also observed through the experiment. Furthermore, the proposed method has been tried out by engineers with practical work experience in software development industry and received good feedback.

Structured SRS for e-Government Services With Boilerplate Design and Interface

There are many projects being carried out to develop e-Government applications. In order to develop an efficient application, a proper requirements specification is strictly required. However, in the case of improper requirements specifications, errors are unavoidable for the developed applications. Therefore, it is necessary to use a model to determine the quality of software requirements. In the literature, several studies have been presented to propose a quality model for software requirements. Yet, there is no study on the development of a quality model for software requirements in Turkish language. Thus, the purpose of this study is to propose a quality assessment model of Software Requirements Specification (SRS) in Turkish language for e-Government applications. The proposed model aims at defining common texts and confirming that the sentences used in writing SRS are developed within an assured structure in order to minimize and standardize the errors originating from natural language. For this purpose, a model based on the Rupp’s boilerplate is created that allows the analyst or requirements engineer to accurately enter the requirements statements. In this study, an interface (webpage) is also created so that the model may match requirements with common text templates, compute similarity values, and also may insert the requirements, regulatory documents, and user requirements into the template in a more convenient format. In order to evaluate the proposed model, the sentences in the 32 documents, including 843 requirements, were adapted to the model. Then, the usability of the model was validated by the requirement engineers serving in a Government service. According to the results, it is concluded that the proposed model is applicable, and is able to improve the quality of SRS in Turkish language for e-Government applications.

Metrics for software requirements specification quality quantification

The initial and crucial phase of the automation and software development is identifying requirements and documenting them in an appropriate format that denotes software requirement specification (SRS). The quality and productivity at different phases of the software development depend on requirements specification. The quality of the end product of software development is proportionate to the quality of SRS. Hence, estimating the quality of the SRS is essential. Though the numerous metrics have been defined in contemporary research, the IEEE standard metrics are considered authentic to scale SRS quality. This manuscript endeavored to redefine the IEEE standard metrics’ measuring approaches to improvise SRS quality assessment. The experimental study exhibiting the significance and robustness of the proposed approach that compared to the contemporary contributions of the recent literature

Ambi Detect: An Ambiguous Software Requirements Specification Detection Tool

Software Requirements Specification (SRS) is considered a highly critical artifact in the software development. All phases in software development are influenced by this artifact. Defects in software requirements may higher the risk of project overschedule that contributes to cost overrun of the project.Researchers have shown that finding defects in the initial software development phase is important becausethe cost of the bug is cheaper if it is fixed early. Hence, our main goal is to provide a platform for requirement engineers to produce better requirement specifications. We propose AmbiDetect, a (prototype) tool toautomatically classify ambiguous software requirements. AmbiDetect combines text mining and machine learning for ambiguous requirement specification detection. The text mining technique is used to extract classification features as well as generating the training set.AmbiDetect usesa machine learning technique to perform the ambiguous requirement specification detection. From an initial user study to validate the tool, the result indicates that the accuracy of detection is reasonably acceptable.Although AmbiDetect is an early experimental tool, we optimist that this tool can be a platform to improve SRS quality.

Domain Ontology for Requirements Classification in Requirements Engineering Context

Research in recent years has shown a merge between the areas of requirements engineering and semantic technologies. With the release of the semantic concept and the progress of semantic technologies, the opportunities for applying ontologies as a means to define information and knowledge semantics have become increasingly accepted in different domains. Concurrently, the implementation of most requirements classification techniques does not handle the semantic aspects of requirements. If the meaning of requirements and their relations can be handled, software developers can obtain more effective requirement classifications to produce requirements specifications of higher quality. In this study, a domain ontology is proposed to present a requirements classification technique that can be used to share and describe different classifications. The proposed ontology is built using a systematic method based on Methontology and it is implemented using Protege. The developed ontology was successfully evaluated using validation and verification tests. The validation test included the evaluation of content and competency questions, while the verification test included the evaluation of taxonomy and the implementation of the FOCA method. The proposed ontology may represent a significant contribution to ontology libraries. In addition, this ontology can be used in several ways to increase the quality of software requirements specification documents. It could also ensure consistency between requirements, and facilitate communication between requirements engineers owing to the use of same terminologies for various software applications.

A Fuzzy Case-Based Reasoning Model for Software Requirements Specifications Quality Assessment

Different software Quality Assurance (SQA) audit techniques are applied in the literature to determine whether the required standards and procedures within the Software Requirements Specification (SRS) phase are adhered to. The inspection of the Software Requirements Specification (iSRS) system is an analytical assurance tool which is proposed to strengthen the ability to scrutinize how to optimally create high-quality SRSs. The iSRS utilizes a Case-Based Reasoning (CBR) model in carrying out the SRS quality analysis based on the experience of the previously analyzed cases. This paper presents the contribution of integrating fuzzy Logic technique in the CBR steps to form a Fuzzy Case-Based Reasoning (FCBR) model for improving the reasoning and accuracy of the iSRS system. Additionally, for efficient cases retrieval in the CBR, relevant cases selection and nearest cases selection heuristic search algorithms are used in the system. Basically, the input to the relevant cases algorithm is the available cases in the system case base and the output is the relevant cases. The input to the nearest cases algorithm is the relevant cases and the output is the nearest cases. The fuzzy Logic technique works on the selected nearest cases and it utilizes similarity measurement methods to classify the cases into no-match, partial-match and complete-match cases. The features matching results assist the revised step of the CBR to generate a new solution. The implementation of the new FCBR model shows that converting numerical representation to qualitative terms simplifies the matching process and improves the decision-making of the system.

A Tool for Analyzing Software Requirements Document Quality

The quality of a Software Requirements Specification (SRS) is measured in terms of quality properties such as completeness, conciseness, consistency and understandability. In general, evaluation of the SRS quality is done manually during review sessions. The evaluation process, however, is hugely dependent on the expertise of human experts i.e. the reviewers. In fact, the judgment of the human experts could also be inconsistent due to various factors including experience, knowledge and domain. The objectives of this study are to (1) identify feasible rules to measure SRS quality; and (2) help requirements engineer to improve their SRS quality. In this study, we analyzed SRS quality properties from the literature and identified quality factors that are feasible to be automated. From here, we identified two types of properties that are (1) requirements sentence quality (RSQ) and (2) requirements document quality (RDQ). For each of the type, its relevant quality indicators were identified. From here, rules on how to identify the quality indicators were further investigated and documented. As a case study, we implemented SRS Quality-Checker tool concept for demonstrating how the rules were implemented to measure the SRS quality

Quality of software requirements specification in agile projects: A cross-case analysis of six companies

Agile Software Development (ASD) has several limitations concerning its requirements engineering activities. Improving the quality of Software Requirements Specifications (SRSs) in ASD may help to gain a competitive advantage in the software industry. Based on the findings of a Systematic Mapping study, six industrial case studies in different contexts were conducted to investigate and characterize the requirements specification activity in ASD. Data collected from documents, observations, and interviews with software engineers were triangulated, analyzed, and synthesized using Grounded Theory and Meta-Ethnography. The analysis and cross-synthesis of the six case studies resulted in a model describing the phenomenon. This model defines the simplicity and objectivity as essential quality factors of SRSs in ASD. The main factors that affect the SRSs quality in ASD projects are related to their customer-driven nature that leads to prolix SRSs, hindering its understanding from the developer perspective. The emerged model is supported by explanations and provides a deeper understanding of the requirements specification activity in ASD. This creates opportunities for further studies and improvements in SRSs for ASD in industry.

A Process Model for Standardization and Increase in the Requirements Quality

One of the biggest problems encountered in Requirements Engineering is the fact that the requirements are poorly specified, inconsistent with the stakeholder's needs or badly written. This study aims to present a model of Requirements Engineering Process allowing requirements description standardization, seeking to improve the software requirements specification quality. The proposed process will operate in the Requirements Specification and Validation assisting the Requirements Engineer on writing requirements in Natural Language. A comparative study between the proposed process and others who offer assistance to Requirements Engineering is also presented, so benefits of using the proposed process could be identified in software development.

Web-Based System for Software Requirements Quality Analysis Using Case-Based Reasoning and Neural Network

The overall success of a software project depends on the quality of its software requirements specifications (SRS). Hence, it is very important to get the requirements correct from the onset of the project. This research paper proposes a web-based system to perform SRS quality analysis using Case-Based Reasoning (CBR) and Artificial Neural Network (ANN). CBR is an AI technique that learns and deduces solutions based on past cases or experiences that are stored in a case base. However, when the case base becomes very large, the “Retrieve” phase of CBR becomes very tedious. So, in this research, we use ANN to improve the retrieval phase within the CBR. ANN measures the similarity of the new case against all existing cases in the case base. This results in a more efficient method of performing quality analysis for a given SRS document.

Implementing Case-Based Reasoning Technique to Software Requirements Specifications Quality Analysis

Implementing Case-Based Reasoning Technique to Software Requirements Specifications Quality Analysis