Semantic Similarity Between Word using Images

Measurement the linguistics similarity between words is a vital part in varied tasks on the online like relation extraction, community mining, document agglomeration, and automatic data extraction. Despite the utility of linguistics similarity measures in these applications, accurately measurement linguistics similarity between 2 words (or entities) remains a difficult task. A method is proposed to estimate semantic similarity using page counts and text snippets retrieved from a web search engine for two words. Specifically, various word co-occurrence measures are defined using page counts and integrate those with lexical patterns extracted from text snippets. To identify the numerous semantic relations that exist between two given words, pattern extraction algorithm and pattern clustering algorithm are proposed. The best combination of page counts-based co-occurrence measures and lexical pattern clusters is learned mistreatment support vector machines. The projected methodology outperforms varied baselines and antecedently projected web-based linguistics similarity measures on 3 benchmark knowledge sets showing a high correlation with human ratings. Moreover, the projected methodology considerably improves the accuracy in a very community mining task.