A non-hierarchical neural network approach for analyzing textual data. (Poster)
Battleson, Brenda L.
Woelfel, Joseph D.
MetadataShow full item record
Artificial neural networks excel at recognizing patterns in textual data. Its pattern recognition capability allows a neural network engine to assign weights representing the multiple connections among concepts. These weights can then be used to create dendograms or otherwise categorize concepts in a hierarchical manner. However this approach has its limitations. Words often have different meanings depending on the context in which they occur. For instance, the term "mustang" can refer to a horse, car or airplane and thus appear in up to 3 clusters. Yet a hierarchical clustering method is unable to fully describe multiple relationships because it is only able to show concepts connected in one way. Each concept is assigned to only one "best" cluster in the output suggesting that there is only one meaning of that concept in the data analyzed. The use of a non-hierarchical approach can address this limitation since it allows the researcher to interact with the neural network to explore all possible meanings of a concept. Thus, in the resulting output a concept may appear in as many clusters as are appropriate. In this study hierarchical and non-hierarchical procedures are compared. Two large datasets, one containing multiple newspaper articles, the other pre-processed data listing only terms of interest, are examined.