Machine learning for sentiment analysis

Closed - This job posting has been filled and work has been completed.

Job Description

I have 1000 news English articles rated positive, negative or neutral by a pool of raters. I need someone to use a machine learning algorithm to learn classification models, based on this data, that predict whether a given news article is positive, negative or neutral.

This should be relatively straightforward for someone who is familiar with supervised classification algorithms and their application. However, some experimentation with the parameters of the chosen classification algorithm will be needed; as well as some experimentation with different ways of turning the news articles into feature vectors.