Sentiment Analysis Framework Development

We ar looking for a social network sentiment analysis expert. The expert should have background in natural languge processing, data mining (text mining) and machine learning algorithms. The development platform can be selected among any programming language, but Java is preferred. The project requires to develop a novel hybrid sentiment analysis framework with machine learning algorithms. The implementations of some algorithms may be added from existing implementations from Weka or anyother open source software to the developed platform. The platform should be compatible to different datasets. The platform should allow user to change the configuration of the model easily. For example, changing the classification algorithm of Naive bayes with another classifier should be easily changed. The developed framework should outperform existing sentiment analysis classifiers in terms of accuracy, precision, etc. metrics. The framework will be tested on several (4-5 different) sentiment benchmark data sets, such as twitter, movie recommendation, etc. Sample plan for the project is given below: Other plan suggestions are welcome, but the total project deadline is 9 months max and budget is confined to 3000$. Please keep these considerations in mind. The coder is also expected to submit a "Progress report" at the end of each month. ------------------------------------------------------------------------------------------ Milestones Description Month 1 Analysis & Propose of Algorithm, Identification of Existing Algorithms, Ensemble Techniques, Identification of Datasets to be used. Month 2 Implementation and Results of Algorithm 1 Month 3 Implementation and Results of Algorithm 2 Month 4 Implementation and Results of Algorithm 3 Month 5 Implementation and Results of Ensemble Technique 1 Month 6 Implementation and Results of Ensemble Technique 2 Month 7 Implementation and Results of Ensemble Technique 3 Month 8 Comparison of Results of the Individual Algorithms and Ensemble Techniques ---------------------------------------------------------------------------------------------------