摘要：In this study, a system developed to summarize by automatically evaluating comments about product with using text mining techniques will be described. The data has been primarily went through morphological analysis process, because they are texts written in natural language. Words and adjectives meaning positive or negative are determined. They show product features in texts. The tree structure is established according to Turkish grammar rules as subordinate and modified words are designated. The software which uses the depth-first search algorithm on the tree structure is developed. Data from result of software is stored in the SQL database. When any inquiry is made from these data depending on any property of product, numerical information which indicates the degree of satisfaction about this property is obtained.
其他摘要：Bu çalışmada, metin madenciliği tekniklerinden yararlanılarak bir ürün hakkındaki yorumları otomatik olarak değerlendirip özetlemek için geliştirilmiş bir sistem tanıtılmıştır. İncelenen veriler doğal dilde yazılmış metinler olduğundan öncelikle morfoloji
关键词：Text Mining;Information Extraction;Natural Language Processing
其他关键词：Metin madenciliği;Bilgi çıkarımı;Doğal dil işleme
