Whatever it is called by the librarians (traditional, modern, virtual, etc.), there is chain process, that starts with the user need, need analysis, need negotiation, consulting appropriate sources, customizing search strategy, search, retrieval, evaluation of search results, modification if required, and delivery or suggesting alternative sources...
"Nearly forty years ago now, Shera (1964) not only foresaw the use of computing to take some of the menial labor out of library reference work, but he also proposed a method for achieving that goal. Shera wrote that “the really great promise of automation is to be sought in… the opportunity it affords to analyze the reference process and re-define reference service” (p. 203). He saw this redefinition as removing the “fetch and carry” aspects, and the potential to raise the intellectual level of reference work. To achieve this end, Shera proposed an agenda for action: first, analyze the processes involved in reference service, and how questions are handled by humans in those processes, and second, create algorithms to represent these processes. Forty years later, Shera’s agenda still provides a sound course of action." Jeffrey Pomerantz, Question Types in Digital Reference: An Evaluation of Question Taxonomies. Unpublished doctoral dissertation, Syracuse University, Syracuse, NY.-- Chapter One
Interesting anecdotes about Reference Transactions by Swiss Army Librarian
On the same shelf:
Course Readings at University of Texas, Austin