A restricted set of full-text articles from a sub-specialty
In this study, our approach of co-word analysis will of biochemistry was analyzed and compared in terms of be ''bottom-up'' in order to analyze the emergence of co-occurrences and co-absences of words. By using the structural elements at higher levels of textual organizadistribution of words over the sections, a clear distinction. The properties at the next-higher level can be anation among ''theoretical'' ''observational,'' and ''methodlyzed as the latent dimensions of the network of relations ological'' terminology can be made in individual articles. However, at the level of the set this structure is no longer among the lower-level units. Therefore, we first attribute retrievable: Words change both in terms of frequencies words to sentences as units of analysis. From this matrix of relations with other words, and in terms of positional of sentences vs. words, new matrices can be generated meaning from one text to another. These results accord by repetitive aggregation. 1 Figure 1 illustrates the process: with Hesse's (1980) thesis about the sciences as fluid At each higher level one can analyze the subsequent manetworks. The fluidity of networks in which nodes and links may change positions is expected to destabilize trices with reference to the question of whether new strucrepresentations of developments of the sciences on the tural properties are indicated. The structural properties basis of co-occurrences and co-absences of words. The (''eigenvectors'') of the matrix represent the specific orconsequences for the lexicographical approach to genganization of the textual structures at the corresponding erating artificial intelligence from scientific texts are dislevel of aggregation.
cussed.
In a previous study, I analyzed the latent structure of the full-text of a single biochemistry article 2 using this ''Plasma-Membrane Transport of Alanine Is Rate-Limiting for Its Metabolism in Rat-Liver Parenchymal Cells,'' published in the October