As a result, measures of scientific  production (which includes these socio-technical features) are a more relevant  indication of likelihood to adopt than measures of scientific progress,  although they could also indicate a potentially contentious or controversial  topic that is generating lots of different opinions.  However, controversy does not  prevent adoption, and in some cases may accelerate adoption

Methodology

Statistical comparisons of time series

Problem:

Comparison of dissimilar bibliometric technology profiles to  investigate whether literature based technology substitution  groupings may be determined from separate measures of science and technology

Dissimilarity  between technology time series:

  1. Time  series based on different number of observations  (e.g. different timespans)
  2. Time  series may be out of phase
  3. Time  series at different stages through the Technology Life Cycle
  4. Time  series may be representative of dissimilar industries
  5. Time  series may repeatedly move between different stages of the Technology Life  Cycle
  6. Time  series may be subject to long-term and shorter term cyclic trends

Potential  time series classification methods:

  1. Supervised learning (i.e. training  examples all have known classification labels) \cite{lin2012pattern}
  2. Semi-supervised learning (i.e.  training examples have both known and unknown classification labels) \cite{lin2012pattern}
  3. Unsupervised learning (i.e. data has unknown classification labels, e.g. clustering) \cite{lin2012pattern}

Pre-processing  of time series:

Statistical  significance testing:

Time series pattern recognition techniques

Examples  of common time series pattern recognition techniques (not exhaustive):