Implementation of the Requirement Elicitation Comparison Process with Step Performed Based on Text Weighting Using TF-IDF at SRS Onlinelaundry

Different perceptions of a Requirement Elicitation exist to form a Step Performed. This possibility can occur, so we need to compare the two artifacts. To address this problem, this study aims to conduct an analysis that compares text data on Requirement Elicitation and Step Performed through artifact comparison in the onlinelaundry application documentation. The study will accomplish this analysis by implementing TF-IDF weighting using the text preprocessing activity. Through text preprocessing preparation for data processing in the form of 5 groupings from requirement elicitation (q1 to q5) and 12 step performed (p1 to p12). In addition, the use case diagram and use case description (step performed) have the same number, namely 12. The text preprocessing stages in this research are case folding, tokenization, stop words, and stemming using the Natural Language Toolkit (NLTK Python). It has successfully compared requirement elicitation and step performed through text weighting using TF-IDF, visualized through a weighting matrix. The results of this weighting then carried out the validity of the results of the comparison process of 0.4089. These results classify the process under the “Fair Agreement” category. Based on identifying these categories, researchers can utilize them as a reference for making recommendations to enhance step performed artifacts, aiming to improve the cohen kappa index and approach a value of one on the scale. Based on identifying these categories, researchers can use it as a reference to make recommendations that can improve the step performed artifacts, aiming to increase the cohen kappa index and get closer to the value of one on the scale.