A1 Req Trace (autorecovered).docx

Advanced Requirement Engineering Tehreem Shabbir 17-1076 MSSE Submitted To: Mam Irum Inayat Submitted on: 23 March 2018

Ref Focus

Method Proposed

Evaluation technique

Results

[1]

It Focuses on the common text around linking terms or words in order to find related textual documents.

Swarm intelligence

The result obtained for the Pine dataset, had recall of 78% with precision of 40% at a 0.2 threshold and slightly outperforming TF-IDF which had 72% recall and 48% precision at a 0.1

[2]

Focused on methods like classical vector space model algorithm, the vanilla vector algorithm, for improving candidate link generation by applying IR techniques.

Information Retrieval of Requirement Tracing

The techniques have been validated using two real-world datasets from two problem domains. The swarm agents mimic and borrow useful behavioral features from communal insects so that it can identify candidate links between two sets of documents. The simple swarm approach, the heuristic to select a term or traversal path in the search space is based on term or document attributes. Evaluation of the algorithms against a comparable keyword-based tool and analysts showed that the retrieval with thesaurus algorithm outperforms all in

We found that this algorithm does not outperform analysts or existing tools in terms of recall & precision, but it does

Limitation

IR techniques are not able to predict the coverage or satisfaction of traced requirements by their lower level requirements.

terms of recall and sometimes - in terms of precision.

[3]

Machine learning approach is presented for generating requirements traceability relations. It is based on learning algorithm that generated traceability rules which capture traceability relations between requirement statements and object models.

Rule-based traceability

[4]

The focus in on effectiveness of information retrieval methods in automating the tracing of textual requirements

RETRO

[5]

An approach to automatically create the RTM based on fuzzy logic, called RTM-Fuzzy is presented, it

RTM-Fuzzy

A series of experiments were conducted to validate the proposed approach. Experiments were based on an industrial case study that includes the specification of commercial requirements and use-cases for a family of software intensive TV products. Experiments have been conducted using two datasets: a small MODIS dataset and a large CM dataset.

For evaluation of these approaches, an experimental study was conducted where the RTMs which

perform faster and with no keyword assignment required of analysts. The retrieval with keyphrases algorithm resulted in improved recall but with decreased precision. The creation of traceability rules is provided by the user and is based on the generalization of existing traceability rules.

We found evidence that RETRO does satisfy the goals of Accuracy and Scalability, Discernibility and Endurability. The effectiveness achieved in relation to the reference RTM: RTM-E

More experiments will be conducted in future.

-

combines two other approaches, one based on functional requirements' entry data called “RTME” and the other based on natural language processing – called “RTM-NLP”

were created automatically were compared to the reference RTM (oracle) created manually based on stakeholder knowledge.

[6]

It presented A heuristic based approach that uses frequent pattern mining. This work constitutes of the existing approaches in uncovering traceability between requests in bug repositories and source code, it expands the horizons of traceability research via mining software repositories.

Heuristic based approach in which a tool, named Sqminer.

KDE (K Desktop Environment) was used to evaluate this approach so that it recovers traceability links.

[7]

The Reasoning Component of LESD aims at helping the engineer to build traceability links between requirements. the engineer asks a question about an entity or an activity using a language based on concepts and conceptual relations. The LESD traceability tool will search for requirements involving these

LESD system

-

achieved 78%, RTM-NLP 76% and the RTMFuzzy 83% effectiveness. It provides enhanced effectiveness in requirement Traceability. The results showed high precision predictions of various types of software artifacts.

The reasoning tool defined for traceability is considered the first step towards the identification of linguistic characteristic of quality criteria like consistency, complete, modifiability.

It does not integrate our traceability tools directly into a versioncontrol system.

Experimentation will be conducted in Future Work

entities. The answer to the question is a list of requirements.

Reference

[1] “Use of Artificial Intelligence in Software Development Life Cycle: A state of the Art Review (PDF Download Available).” [Online]. Available: https://www.researchgate.net/publication/274254538_Use_of_Artificial_Intelligence_in_Software_ Development_Life_Cycle_A_state_of_the_Art_Review. [Accessed: 23-Mar-2018]. [2] “Improving requirements tracing via information retrieval - IEEE Conference Publication.” [Online]. Available: http://ieeexplore.ieee.org/document/1232745/. [Accessed: 23-Mar-2018]. [3] G. Spanoudakis, A. d’Avila-Garces, and A. Zisman, Revising Rules to Capture Requirements Traceability Relations: A Machine Learning Approach. 2003. [4] “Advancing candidate link generation for requirements tracing: the study of methods - IEEE Journals & Magazine.” [Online]. Available: http://ieeexplore.ieee.org/document/1583599/. [Accessed: 23Mar-2018]. [5] A. D. Thommazo, T. Ribeiro, G. Olivatto, V. Werneck, and S. Fabbri, “An Automatic Approach to Detect Traceability Links Using Fuzzy Logic,” in Proceedings of the 2013 27th Brazilian Symposium on Software Engineering, Washington, DC, USA, 2013, pp. 21–30. [6] H. Kagdi, J. I. Maletic, and B. Sharif, “Mining software repositories for traceability links,” in 15th IEEE International Conference on Program Comprehension (ICPC ’07), 2007, pp. 145–154. [7] M. Bras and Y. Toussaint, “Artificial Intelligence Tools For Software Engineering: Processing Natural Language Requirements,” vol. 2.