Automatic formulation and optimization of linear problems from a structured paragraph

Though plenty of research work has been done on natural language processing and linear problem optimization, none of them focused on automatic linear problem formulation from passages and optimizing them at a time. In this research, we developed a problem formulation method using NLTK and developed a completely new linear optimization solver using C# GUI and Python pulp library. We classify the problem with Support Vector Machine(SVM) to identify that the problem is a linear or a non-linear problem. Support Vector Machine(SVM) has an accuracy of 70% due to the small dataset. We extract all the cardinal digits using NLTK Pos_Tagger from the passage. Using the proposed algorithm, we generate all the equations for objective functions and constraints. We also developed software for the solution of the formulated problems. We got an accuracy of 100% for equation generation with our software. Moreover, the proposed system shows an impressive result in problem formulation from our designed structured passage.