Hey City Zen!

The temporal analysis of the word 'Novel' in PubMed by Paulo Pereira in Conta-Natura led me to try a few things... With Andreas Rechtsteiner I performed a rough sketch of an analysis of the words that tend to co-occur with 'Novel' to see if we can think about what is causing this surge in 'novelty' in the biomedical literature....

The image below is a graph of some words with a large co-occurrence probability with 'Novel' in PubMed. Edge thickness denotes larger co-occurrence probability (proximity):



We also computed a co-occurrence probability with the years 1990 to 2004, followed by computing the correlation of the pattern of occurrence of each term for these years (see details). The figure below depicts the terms most correlated in time.



From the figure we can see that the term 'Novel' has a similar temporal pattern as that of 'gene', 'genome', 'identify', and 'proteome', the term 'New' has a similar pattern as 'protein', and the term 'report' is correlated with 'immune'.A plausible hypothesis is that whatever process is causing the increase in use of the term 'Novel' is the same process that is causing the increase in use of 'gene', 'genome' and 'proteome' -- hardly news for anyone familiar with this literature! ;-)