Fuzzy Analysis of historical texts for Prediction of gender roles using Decision Making Process and Machine Learning Techniques

Gender roles prediction is widely explored using machine learning techniques for historical texts that are helpful in decision making process based on fuzzy analysis based on access control system. Gender related cues and granulation of information are carried out with the proposed framework and categorized with three attributes like lexical features, context windows and word patterns. Conventional kind of machine learning techniques predicts the gender roles as classification task for discrimination that aims for the fuzzy model that strictly distinguishes masculine and feminine gender without any bias. Fuzzy logic model is highly suitable for gender prediction in historical texts by processing the texts dealing with fuzziness. Bias can be reduced with the classification problem by the machine learning techniques based on female and male classes. The continuous attributes are treated with analysis based on fuzzy intervals. The proposed work addresses the limitations that are faced by the existing literature works and concentrates on gender prediction using historical texts where the proposed fuzzy model was evaluated against machine learning techniques such as K-nearest neighbour, logistic regression and random forest technique. The results of the study indicated that the fuzzy model has higher accuracy in gender role prediction and decision making using historical text dataset with 86%, 92% and 89% for various attributes in the fuzzy model and it is superior over other machine learning techniques.

Перевод пока недоступен