Exploring the Evolution of Academic Collaboration Networks in Design Research

Ignacio Toledo

School of Design, Universidad del Desarrollo

Paulina Contreras

School of Design, Universidad del Desarrollo

The Motivation of This Study

  • The School of Design celebrates its 25th anniversary
  • We are thinking on the next 25 years of Design
  • Our own curiosity
  • Chisocnet Conference

What is Design?

“Design is to design a design to produce a design.”
— John Heskett

“Everyone designs who devises courses of action aimed at changing existing situations into preferred ones.”
— Herbert Simon

“Design is the conscious and intuitive effort to impose meaningful order.”
— Victor Papanek

The Evolution of Design Research

  • Origin of Design as a creative discipline
    • The Bauhaus (1919)
    • The School of ULM (1953)
  • From creativity-oriented to academic research
    • Foundation of the Design Research Society (DRS) (1966)
    • First Design PhD Programmes (early 90’s)

Design Research Society

“The Design Research Society is a learned society committed to promoting and developing design research. Founded in 1966, It is the longest established, multi-disciplinary worldwide society for the design research community. Our international Conference series, Special Interest Groups, and online presence draws together a community from around the world in all areas of design research.”
— The Design Research Society website

Research Proposal

This study aims to analyze the evolution of academic collaboration networks in the field of design research, focusing on conferences organized by the Design Research Society (DRS).

  • Is there evidence of a emerging research community in design?
  • Is there evidence of a emerging research program in design?

Methods & Data

  • Data was collected conference papers in DRS Digital Library
  • Data covers from 2002 - 2024 and includes:
    • Title
    • Authors
    • Institutions
    • Abstracts
    • Keywords
    • DOI

Methods & Data

  • Network structure analysis (Newman, 2001)
    • Biennial Conference network
    • Accumulated network
  • Structural Topic Model (Roberts, Stewart, and Airoldi 2013)
    • Top Topics
    • Topic Evolution
    • Topic Correlation

Network Structure - Single Conferences

Naturally, since scholar participate in a limited number of works presented to the conference.

Consequently, we observe very sparse networks dominated by the cliques associated to single publications.

Network Structure - Agreggated

The aggregated network adds up all nodes and edges from previous years

Allows to visualize the structure of network. Does not account for dissolving edges dynamics.

Network Grow

Density & Diameter

Distance & Clustering

Components

Structural Topic Model

The title and abstract for each article was combine into a single text.

We estimate the model for K = 20. And set a prevalence model using the year as metadata.

data_prep <- data      %>% stm::textProcessor(.$paper_text, ., 
                                              customstopwords = c("design"))
data_prep <- data_prep %$% stm::prepDocuments(documents, vocab, meta)
model_stm_k20 <- data_prep %$% stm::stm(documents, vocab, K = 20, 
                                        prevalence =~ year, data = meta, 
                                        init.type = "Spectral", seed = 314)

Next Steps

  • Include Author Affiliation
  • Community Detection
  • Test for co-evolution
    • Topic Assorativity
    • Network as metadata
  • Improve Visualization