MetaReq: A Metadata Driven Strategic Semantics Oriented Model for Recommendation of Software Requirements

There's a need for a semantics-oriented software requirement recom- mendation that encompasses knowledge and a strong classification ecosystem. The model uses preprocessing of the query words unto which entities from wiki data are harvested, which comprise World Wide Web knowledge. The model en- compasses strategies like classification of the dataset using decision tree classifiers and classification of the metadata using LSTM classifier. The encompassment of CoSim rank and Pielou's evenness index along with normalized pointwise mutual information (NPMI) measure create a semantics-oriented computation mechanism. The overall framework has the maximum mean precision, accuracy, recall, F-measure lowest value of FDR making it the best-in-class strategy for soft- ware requirement recommendation.