Similarity Evaluation Based on Contextual Modelling

Measuring Text similarity problem still one of opened fields for research area in natural language processing and text related research such as text mining, Web page retrieval, information retrieval and textual entailment. Several measures have been developed for measuring similarity between two texts: such as Wu and Palmer, Leacock and Chodorow measure and others . But these measures do not take into consideration the contextual information of the text .This paper introduces new model for measuring semantic similarity between two text segments. This model is based on building new contextual structure for extracting semantic similarity. This approach can contribute in solving many NLP problems such as te xt entailment and information retrieval fields.