In this project, we perform the analysis of question answering site named StackOverflow. Here, we designed and developed a knowledge extraction system that predicts meaningful knowledge from dataset. The datasets are mined using algorithms to meet the prediction and interests of users. The task of system is to predict the queries and provide the visualization in the form of graphs, histogram, piecharts. We use classification methods to make a prediction considering various factors like Number of Answers, Number of comments, favourite count, Tags, Number of Views and so on. The system will predict and recommend more relevant and related questions to the user query. Based on confidence value, the system will provide number of association rules which are representative of the way a user will check other related relevant questions. The system will discover various association rules across different levels. Here, we make use of test dataset provided by stack overflow website. Analysis is performed using various methods like Logistics Regression algorithm, Apriori algorithm, Support Vector Machine algorithm.
- R Studio
- Node.js
- MySql
- HighCharts
- AWS