摘要 |
A context-based keyword recommending apparatus for online advertisements and a method thereof are provided to analyze online contents to analyze the contents according to each section, and to extract preliminary keywords as considering weighted values of each section, frequency of the preliminary keywords and others. A contents analyzer(110) analyzes an input data stream(DATA), and extracts information contents(CONT) including at least one section. A recommending keyword output unit(130) extracts at least one preliminary keyword(PRE_KW) included in the information contents, and responds to distributions and frequency of the preliminary keywords in each section to determine a recommending keyword(RE_KW), then outputs the determined keyword. |