| Popis: |
Bi-directional communication, which enhances the effectiveness of education through question-and-answer and/or discussion sessions, is important in asynchronous e-Learning systems. Assignments are also important for e-Learning programs to motivate students and track their learning. However, assigned reports in e-Learning programs are submitted online, and it is very easy for students to plagiarize by pasting text from other reports or web sites. This chapter introduces intelligent interaction support systems including an intelligent bulletin board system (iBBS) and a plagiarism-detecting assignment management system (pAMS). Both systems provide text-based intelligent functions that apply natural language processing and support smooth communication and a fair evaluation of e-Learning classes. First, for iBBS, a function for the automated detection of important posting notifications (IPNs) is introduced. The IPN function performs information filtering; however, it is not feasible to estimate the statistics of word occurrences, which are required for conventional information filtering techniques, prior to opening a discussion. Nevertheless, since the subjects of discussion in e-Learning classes are always related to the class topic, the ontology is consistent throughout the discussion. A concept-based vector space model, which can virtually capture ontological information from a collection of documents, is therefore effective for the IPN function. Second, pAMS, an assignment management system with plagiarism detection, is introduced. Students electronically submit their assigned reports using pAMS and the system stores the submitted documents. The system then compares each of submitted reports to other reports and web pages, and detects identical or similar parts of text in the reports. The system then makes the teacher aware of any potential plagiarism. This chapter describes the method for identifying the similar parts of two reports, even if one has been skillfully modified from the other. In addition, the method of retrieving original source materials is introduced. Finally, the evaluation results using a limited set of reports are described. Intelligent Interaction Support for e-Learning 95 |
