Facilitating Cross-Disciplinary Online Communication by Rich Paraphrasing Using Internet Agents

Online Paraphrasing Using Internet Agents (OPIA), is a technique to facilitate on-line communications, such as discussions on message boards across disciplines. If the reader of a document is not in the same discipline as the author, the material will be difficult to understand. Our initial interest is in facilitating discussions among computer scientists and biologists. However, communication can span to numerous crossdisciplinary teams. With members of a specific team being experts in a respective field, we paraphrase the documents of the other discipline to catalyze their understanding of the material. The end result of paraphrasing takes advantage of resources that are now freely available on-line. Specifically, it is a hypertext document that not only contains adaptively paraphrased text, but also pointers to items such as on-line dictionary/encyclopedia entries, explanatory/additional reading material, and illustrative images. Significantly, the richly paraphrased hypertext document exclusively accommodates the needs of the individual reading it. The steps in the paraphrasing algorithm include text extraction, syntactic and semantic analysis, dictionary lookup, selective substitution of text, and generation of hypertext links. Using intelligent Internet agents, search engines, natural language processing, and concepts from information theory, the proposed system retrieves the original document, performs keyword searches, parses the file, and replaces unknown words with synonyms or appositives while maintaining the syntactical structure and semantics of the author’s paper. General Terms Paraphrasing, On-line dictionaries, Theory, Internet, Information, Communication