This research relates to application development and specification documents (SRS) in software development. The SRS document contains essential functional and non-functional requirements in application development. Each element in the SRS must correspond to the results of the elicitation carried out. However, opinions sometimes differ when analyzing the document. This research aims to develop an application to validate the correspondence between elicitation results and artifacts in SRS semantically using the word2vec technique. The results showed that this application gave the highest score for the validation of functional requirements (0.533) and non-functional requirements (0.719). In addition, the value of questionnaire reached 0.71. The Cohen Kappa index will assess these results. If the score is below 0.21, the artifact will be repaired to achieve a score above that threshold. This research has created a text validity application that can read text files in Txt format, extract input through text pre-processing, and perform semantic suitability assessments using the Word2Vec method.
Keywords — Functional Requirement, Non-Functional Requirement, Requirement Elicitation, Text Pre-Processing