Natural Language Processing: What are some efficient algorithms in semantic analysis?

What are some good algorithms(in terms of accuracy) out there to perform semantic analysis on two sentences. The task I am trying to achieve here is try to get the similarity of context between two sentences.