Description

One of the key limitations about research involving big data is the lack of a sound methodological process that drives the conceptual and analytical questions posed to the data. In this study, we adapt the popular CRISP-DM process to analyze large volumes of unstructured data to generate analytical insights. We add specificity to the CRISP-DM methodology. Specifically, we propose “Cross Industry Standard Process for Electronic Social Network Platforms (CRISP-eSNeP)”, as an extension to the CRISP-DM methodology. Our methodology focuses on efficient pre-processing of large and unstructured electronic social network data. We illustrate our arguments by applying this methodology to understand the relationship between user influence and information characteristics as depicted on the Twitter microblogging platform.

Share

COinS
 

Adapting CRISP-DM Process for Social Network Analytics: Application to Healthcare

One of the key limitations about research involving big data is the lack of a sound methodological process that drives the conceptual and analytical questions posed to the data. In this study, we adapt the popular CRISP-DM process to analyze large volumes of unstructured data to generate analytical insights. We add specificity to the CRISP-DM methodology. Specifically, we propose “Cross Industry Standard Process for Electronic Social Network Platforms (CRISP-eSNeP)”, as an extension to the CRISP-DM methodology. Our methodology focuses on efficient pre-processing of large and unstructured electronic social network data. We illustrate our arguments by applying this methodology to understand the relationship between user influence and information characteristics as depicted on the Twitter microblogging platform.