Building Knowledge Flow of Textual Topics for the e-Science Knowledge Grid

In order to build knowledge flow of textual topics, the enormous topics of scientific texts are needed to be managed and organized. Island is proposed to manage the topic of scientific text belonging to the same domain. Interest table is presented to store the textual topics that researcher is interested recently. KM-Chord is introduced to organize KM, which not only reflects the concepts (keywords) of textual topic, but also takes the relationships between concepts into account. So KM-Chord has more strong semantic information than traditional Chord. Knowledge map (KM) representing textual topic can be located quickly with the help of KM-Chord, and the semantic information between KMs can be found (eg. similar and subtype relationship) for building the knowledge flow of the textual topics. The knowledge flow of textual topics can support knowledge innovation, cooperative teamwork, problem-solving and decision-making in e-Science Knowledge Grid. Experiments show KM-Chord not only can locate and route the textual topic, but also build the knowledge flow of textual topic effectively. So the proposed KM-Chord is a good method to build the Knowledge Flow of textual topics for the e-science Knowledge Grid, Semantic Grid and current Web.