Searching and Analyzing Intellectual Property on the Internet

This thesis concerns the usage of intelligent agents to search the Internet to identify possible infringements of and legal impediments to the procurement of intellectual property, such as existing or proposed patents, trademarks, and copyrights. The agent can utilize any of the known artificial intelligence models, such as expert systems, case-based reasoning, and fuzzy logic, and can return a ranked list of located data relevant to the searched intellectual property. A trademark search agent model is proposed that will account for differences between an input trademark, expand the search terms in an expansive heuristic manner, locate relevant data based upon the input and expanded terms, and return the most relevant terms located in a hierarchical format of relevancy. A patent intelligent search agent is proposed which can simulate the combinatory logic used by patent examiners to reject patent applications as being merely a collection of known elements.