Data-Driven LIS Methods Reveal Cyclical Evolution Over 32 Years
Study maps emergence-stability cycles in library science research methods from 1990 to 2022.
A team led by Chengzhi Zhang at Nanjing University conducted a fine-grained analysis of Library and Information Science (LIS) research methods from 1990 to 2022. Using automated extraction on academic papers, they identified four categories of data-driven method entities: algorithms and models, data resources, software and tools, and metrics. The study aimed to assess how the data-centric research paradigm has reshaped the LIS discipline over three decades.
The findings reveal that data resources are the most pivotal driver of methodological evolution in LIS, far more than algorithms or tools. The researchers observed a recurring cyclical pattern they term 'emergence-stability/practical application' – new methods appear, stabilize, then become practically applied. This pattern holds true across different research topics and method types. The work provides a quantitative lens for understanding how computational and data-driven approaches have gradually permeated a traditionally qualitative field, offering insights for librarians, information scientists, and research methodologists.
- Extracted four fine-grained method entity categories from LIS papers spanning 1990-2022
- Data resources identified as the primary driver of methodological evolution in the field
- Research methods exhibit a cyclical pattern of 'emergence-stability/practical application'
Why It Matters
Offers quantitative evidence of how data-centric approaches systematically reshape LIS methodology over decades.