University of Maryland

Cloud Computing

Ping Wang and Jia Sun

Why do some concepts come to be highly popular, significantly reshaping the IT landscape, while others do not? We address this question by exploring the communities of organizations that underlie IT innovations. In an illustrative investigation of the community for cloud computing, we integrate theories of organizational ecology and scale-free networks and apply their associated methods.

Based on organizational ecology theory and the scale-free network theory, we raised three hypotheses:

  • Hypothesis 1: Legitimation is positively associated with the entry rate of an IT innovation community.
  • Hypothesis 2: Competition is negatively associated with the entry rate of an IT innovation community.
  • Hypothesis 3: The extent to which competition reduces the entry rate of an IT innovation community declines with the degree of scale-freeness of the community.

To test these hypotheses, we search in the ProQuest National Newspapers database for the phrase “cloud computing” in the title, subject, abstract, and full text of news articles published during a ten-year period (2002 to 2011). We used Stanford NER, a Natural Language Processing (NLP) tool, to automatically identify the organization names from the paragraphs. Stanford NER is a Java implementation of a named entity recognizer. We wrote a PERL script to extract the organization names from the paragraphs. The precision of Stanford NER for organization identification was approximately 62% and thus the result we obtained had many errors. We relied on human judgment to correct the errors by employing CrowdFlower, a crowdsourcing service.

Regarding the relationships between the organizations involved (i.e. the structure of the community), organizations were identified as co-occurring when they were mentioned within the same paragraph in an article.

The figure below shows the number of articles on cloud computing and the monthly rate of new organizations that started appearing in these articles.


The figure below shows clusters of organizations that appeared in the Cloud Computing articles in 2008, as an example.


Following previous research on organizational ecology, we applied density-dependence model and the latest measure for scale-freeness. Using regression analysis, we have found support to all three hypotheses. Substantial support to the theories and the effectiveness of the methods have led us to embark on a promising research program focused on IT innovation concepts and communities. This program affords several opportunities to break new grounds in Information Systems research: (1) adding an ecological explanation to the theories of IT innovations; (2) fertilizing new ground for IT innovation research; (3) stimulating research on inter-organizational scale-free networks; and (4) demonstrating the utility of computational discourse analysis.

Relevant Publication: