Construction of an Evaluation Corpus for Opinion Extraction

Opinion retrieval aims to tell if a document is positive, neutral or negative on a given topic. Opinion extraction further identifies the document’s supportive and the non-supportive evidence. This paper defines the annotation of opinionated material. The algorithm employs opinion holders, a topic’s conceptual words, sentiment words, opinion operators, and negation operators to recognize opinions. An opinion extraction system is developed and then reflects the major views of selected information sources. The text-based evidence extracted is ready for opinion summarization and opinionated question answering.