An ELM-based model for affective analogical reasoning

摘要: Between the dawn of Internet through year 2003, there were just a few dozens exabytes information on Web. Today, that much is created weekly. The opportunity to capture opinions general public about social events, political movements, company strategies, marketing campaigns, and product preferences has raised increasing interest both in scientific community, for exciting open challenges, business world, remarkable fallouts financial prediction. Keeping up with ever-growing amount unstructured Web, however, formidable task requires fast efficient models opinion mining. In this paper, we explore how high generalization performance, low computational complexity, learning speed extreme machines can be exploited perform analogical reasoning vector space model affective common-sense knowledge. particular, by enabling reconfiguration such space, allow polarity associated natural language concepts calculated more dynamic accurate way and, hence, better concept-level sentiment analysis.