Co-occurrence Matrices and their Applications in Information Science: Extending ACA to the Web Environment - Computer Science > Information RetrievalReport as inadecuate




Co-occurrence Matrices and their Applications in Information Science: Extending ACA to the Web Environment - Computer Science > Information Retrieval - Download this document for free, or read online. Document in PDF available to download.

Abstract: Co-occurrence matrices, such as co-citation, co-word, and co-link matrices,have been used widely in the information sciences. However, confusion andcontroversy have hindered the proper statistical analysis of this data. Theunderlying problem, in our opinion, involved understanding the nature ofvarious types of matrices. This paper discusses the difference between asymmetrical co-citation matrix and an asymmetrical citation matrix as well asthe appropriate statistical techniques that can be applied to each of thesematrices, respectively. Similarity measures like the Pearson correlationcoefficient or the cosine should not be applied to the symmetrical co-citationmatrix, but can be applied to the asymmetrical citation matrix to derive theproximity matrix. The argument is illustrated with examples. The study thenextends the application of co-occurrence matrices to the Web environment wherethe nature of the available data and thus data collection methods are differentfrom those of traditional databases such as the Science Citation Index. A setof data collected with the Google Scholar search engine is analyzed using boththe traditional methods of multivariate analysis and the new visualizationsoftware Pajek that is based on social network analysis and graph theory.



Author: Loet Leydesdorff, Liwen Vaughan

Source: https://arxiv.org/







Related documents