Merging Textual Knowledge Represented by Element Fuzzy Cognitive Maps

摘要

Importance degree and difference degree ofkeywords in different topics have been measured by theassociated weights in Element Fuzzy Cognitive Maps(E-FCMs) which can represent textual knowledge effectively.Logic “and” operation is introduced to roughly evaluate thesimilarities between the mass E-FCMs in order to form thesimilar sets of textual knowledge. Based on the associatedweight measuring and the logic operation, anE-FCMs-based knowledge merging algorithm is proposed toinspect the noisy and the redundancy information hidden inthe original E-FCMs belonging to one similar set. A formulaobtained through F-measure is employed as an indicator tomeasure the loss of textual information during the mergingprocess of E-FCMs. The merging algorithm and theindicator provide a concise representation of textualknowledge that can be used in understanding-basedautomatic text classification and clustering, as well asrelevant knowledge aggregation and integration. Theproposed algorithm will have very good applicationprospects in future.