Exploring an automated method for the analysis of virtual reference interactions

Abstract

PurposeThis study aims to determine if automated coding with regular expression is a strong methodology to identify themes in virtual reference chat.Design/methodology/approachThe authors used a combination of manual and automated coding of chat transcripts for a period of two years to identify the categories of questions related to the new library system. This methodology enabled them to determine if regular expression accurately identified the topics of chat transcripts.FindingsThey discovered that regular expression is an appropriate method to identify themes in virtual reference interactions. This method enabled them to establish that patrons asked questions related to system changes in the weeks following their implementations.Originality/valueThis study highlights a new methodology for transcript analysis.