Language is a hallmark of intelligence, and endowing computers with the ability to analyze and generate language as a field of research is known as Natural Language Processing (NLP) - has been the dream of Artificial Intelligence. Software requirements are typically captured in natural languages (NL) such as English and then analyzed by software engineers to generate a formal software design/model. However, English is syntactically ambiguous and semantically inconsistent. Hence, English specifications of software requirements cannot only result in erroneous and absurd software designs and implementations but, the informal nature of English is also a main obstacle in machine processing of English complex specification of the software requirements. To tackle this key dispute, there is need to introduce a controlled NL representation for software requirements, to generate perfect and consistent software models. Proposed framework aims to model complex software requirements expressed in natural language and represent them with a new methodology that captures the natural language understanding(NLU) of events and models them using Stochastic Petri Nets (SPN) instead of only intermediate graph based structure using techniques of Natural Language Processing (NLP), this helps in removing ambiguity and corrects interpretation of requirements. To eliminate ambiguity, work combines all the different meanings (SPN graphs) of each ambiguous sentence into colored SPN graph. SPNs are state machines that help us to visualize better, the combined SPN graph. It can also represent knowledge about the requirement, which can be used to derive test case in early development phase. Hence aim of proposed work is twofold that overcomes the problem of ambiguity and knowledge representation. Stakeholder�s document is input to framework, pre-processed by some pre-filter with certain functionality to improve the parsing. This parsed output gets converted into simple graph which in turn is converted into SPN graph with color representation to improve ambiguity. Pre-filter may be designed with self-learning capabilities to perk up output without human involvement.

NLP, SPN, SRS

[1] K.R. Chowdhary �Natural Language Processing� April 29, 2012
[2] Bures, T., Hnetynka et al. �Requirement Specifications Using Natural Languages� Technical Report D3S-TR-2012-05 December 2012.
[3] Agung Fatwanto � Software Requirements Specification Analysis Using Natural language Processing Technique � IEEE Quality in Research 2013
[4] A. Reed and B. Kellogg, Higher Lessons in English., 1877. (In Wikipedia: Sentence Diagram, en.wikipedia.org/wiki/Sentence_Diagram, last accessed: April 21st 2013).
[5] OMG. Semantics of Business voc bulary and Rules. (SBVR) Standard v1.0. Object Management
[6] Ashfa Umber & I. S. Bajwa � Minimizing Ambiguity in Natural Language Software Requirements Specification� 2011 IEEE.
[7] Annervaz K.M., Vikrant Kaulgud & et. al �Natural Language Requirements Quality Analysis Based on Business Domain Models� ASE 2013, Palo Alto, USA New Ideas Track, 2013 IEEE
[8] M. Ilieva and O. Ormandjieva, �Automatic transition of natural language software requirements specification into formal presentation,� in Natural Language Processing and Information Systems. Springer, 2005, pp.392�397
[9] Bourbakis, N., Manaris, R �An SPN based Methodology for Document Understanding� IEEE nternational Conference on Tools for Artificial Intelligence, Tapei, Taiwan, 1998, pages 10-15.
[10] Haas, Peter J., Stochastic Petri Nets � Modelling, Stability,Simulation. Springer-Verlag New York, Inc 2002, ISBN 0-387-95445-7
[11] Yeh, J-Y., Ke, H-R., Y, W-P, Meng, I-H., Text summarization using a trainable summarizer and latent semantic analysis, Information Processing and Management, Vol. 41 (2005)pages 75-95.
[12] Ko, Y., Seo, J., An effective sentence-extraction technique using contextual information and statistical approaches for text summarization, Pattern Recognition Letters 29 (2008) pages1366-1371.
13] Moens, M.F, Angheluta, R., Dumortier J., Generic technologies for single- and multidocuments summarization, Information Processing and Management, Vol. 41 (2005) pages 569-586.