Abstract Semantic Textual Similarity (STS) measures the meaning similarity of sentences. Applications of this task include machine translation, summarization, text generation, question answering, short answer grading, semantic search, dialogue and conversational systems.we have use unsupervised method CCA(Canonical correlation Analysis)using cosine similarity and WMD(word movers distance).