Abstract

LEZCANO, Luis A; GUZMAN, Jaime A  and  VELEZ, Carlos A. A Model for the Identification of KAOS Elements (Knowledge Acquisition Automated Specification) from the Specification of Requirements in Natural Language. Inf. tecnol. [online]. 2015, vol.26, n.6, pp.129-138. ISSN 0718-0764.  http://dx.doi.org/10.4067/S0718-07642015000600015.

In this work, a model to automatically obtain the basic elements of the goal diagram known as KAOS (goal, agent, and entity) and to minimize the polysemic semantic ambiguity in getting these elements is proposed. Furthermore, the proposed model is validated by using case studies. The KAOS goal diagram (Knowledge Acquisition Automated Specification) is often used in the definition and analysis phases of software engineering to determine the processes that are carried out in the organization that request the software. In different works of the scientific literature that use the KAOS diagram some problems caused by the manual elaboration of the diagram by the software analyst are present (confusions, subjectivity, and ambiguity, among others). With the proposed methodology in this work the basic elements of the goal diagram KASOS are obtained in automatic form avoiding the errors mentioned above.

Keywords : software engineering; natural language; validation of requirements; semantic disambiguation; KAOS.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )