This project attempts build a classification and ranking system with algorithms mainly with Python Scikit[6]. We plan to identify the pros and cons of system architecture and chosen algorithm, and evaluate the algorithm with experiments. The primary goal is to design and implement a system supporting SVM classification and then SVM ranking algorithm that incorporates and measure algorithmic level performance.
Considering the numbers of heterogeneous infrastructure available in the open source framework, and the given definition/scope of the recommendation engine powered by SVM, the sub-system design will incorporate several popular technology to better use the power of these tools, to achieve an recommender solution given the scope. The sub-system architecture is composed of four layers - Data Ingestion Layer - Data Batch Layer - Data Analytics Engine Layer (Logistic Regression, And SVM)