Thematic Persistence
P. Stavropoulos 
Description
Thematic Persistence captures the ability of a research topic to remain present and influential in the scientific record over extended periods of time. It reflects the continuity, longevity, and stability of themes, distinguishing enduring areas of inquiry from those that are short-lived or sporadic.
Persistent topics often indicate fields with strong conceptual foundations, long-term societal or technological relevance, or strategic importance for research policy. Conversely, non-persistent topics may point to fleeting interests or speculative research directions.
Measuring thematic persistence helps assess the structural evolution of science, identify durable agendas, and guide funding, policy, and curriculum decisions.
Thematic Persistence Score (TPS)
One way to operationalize persistence is through composite indicators such as the Thematic Persistence Score (TPS). TPS combines multiple aspects of a topic’s evolution (continuity across years, growth, impact, and recency) into a single measure.
Other methodologies apply different approaches, such as linking clusters across time periods, defining continuity typologies, or evaluating the survival of citation-based topics.
Measurement
Thematic persistence can be measured through a combination of:
- Temporal continuity: duration and uninterrupted presence of topics across consecutive years or periods.
- Growth dynamics: how the volume of publications on a topic changes over time.
- Impact measures: the influence of topic publications relative to their fields.
- Structural stability: whether a topic maintains coherence in its conceptual or citation network.
- Recency: whether a topic remains active in the most recent period.
The precise operationalization depends on the chosen methodology, as outlined below.
Datasources
OpenAIRE Research Graph
The OpenAIRE Research Graph offers extensive metadata on publications, including:
- Publication year, which is crucial for identifying consecutive topic appearances.
- Citation metadata, enabling FWCI computation via connected sources.
Semantic Scholar
Semantic Scholar offers full-text access and machine-readable metadata, including:
- Year of publication
- Citation counts
Existing Methodologies
SciNoBo Toolkit
The SciNoBo Toolkit provides essential functionalities for TPS:
- Field of Science (FoS) classification: Uses a hierarchical taxonomy (6 levels) to assign topics to publications, capturing both broad disciplines and fine-grained emerging themes. This allows robust tracking of how topics evolve across scientific fields.
- Citation Analysis: Aggregates citation metrics across publications, facilitating computation of Field-Weighted Citation Impact (FWCI) for each topic-year combination.
These tools make it feasible to apply the TPS metric across large bibliographic datasets with a rich contextual understanding of scientific domains.
TPS Formula:
For each topic, sequences of consecutive years are identified where the topic appears in publications. For each such sequence s, the score is computed as:
\[ \text{Score}_s = (\text{Length}_s)^{1.5} \times \text{Count}_s \times \text{Growth}_s \times \text{FWCI}_s \times \text{Recency}_s \]
Where:
- \(\text{Length}_s\): Length of the sequence (in years)
- \(\text{Count}_s\): Number of publications in the sequence -
- \(\text{Growth}_s = \frac{\text{LastYearCount}}{\text{FirstYearCount}}\) (capped at 3)
- \(\text{FWCI}_s\): Mean Field-Weighted Citation Impact for publications in the sequence
- \(\text{Recency}_s = 1 + \frac{w (\text{LastYear}_s - \text{MaxYear} + 10)}{10}\), with \(w\) as a recency weight (e.g. 0.2)
The final TPS is the sum of the scores for all sequences of the topic:
\[ \text{TPS}_{\text{topic}} = \sum_s \text{Score}_s \]
This approach emphasizes continuity, while integrating growth, impact, and recency.
Longitudinal Co-word Analysis (SciMAT)
The SciMAT framework (Cobo et al. 2011) measures persistence by detecting continuing themes across consecutive time periods.
- Topics identified via keyword co-occurrence networks.
- Continuity measured using the Inclusion Index:
\[ \text{Inclusion}(U,V) = \frac{|U \cap V|}{\min(|U|,|V|)} \]
- Topics linked across periods with high inclusion are continuing; absence of links indicates new or discontinued themes.
This approach emphasizes structural continuity of thematic vocabularies.
Direct-Citation Topic Survival
The CWTS publication-level classification system (Waltman and Eck 2012) enables persistence analysis based on citation-linked topic clusters.
- Topics are defined via direct citation clustering.
- Persistence is measured through indicators such as:
- Survival length: number of years a cluster remains active.
- Activity stability: whether publication volume is maintained or growing.
- Survival length: number of years a cluster remains active.
This approach measures persistence at the topic-cluster level, grounded in citation networks.
Continuity Typologies
The continuity framework (Yan 2014) quantifies persistence by categorizing topics into distinct evolutionary types:
- Steady: stable over time
- Concentrating: narrowing focus while persisting
- Diluting: broadening and dispersing
- Sporadic: intermittent appearance
- Emerging: new and growing
Continuity is evaluated by the strength of inter-year linkages among topic clusters.
This allows distinguishing different modes of persistence and topic evolution.
References
Reuse
Citation
@online{apartis2024,
author = {Apartis, S. and Catalano, G. and Consiglio, G. and Costas,
R. and Delugas, E. and Dulong de Rosnay, M. and Grypari, I. and
Karasz, I. and Klebel, Thomas and Kormann, E. and Manola, N. and
Papageorgiou, H. and Seminaroti, E. and Stavropoulos, P. and Stoy,
L. and Traag, V.A. and van Leeuwen, T. and Venturini, T. and
Vignetti, S. and Waltman, L. and Willemse, T.},
title = {Open {Science} {Impact} {Indicator} {Handbook}},
date = {2024},
url = {https://handbook.pathos-project.eu/sections/2_academic_impact/thematic_persistence.html},
doi = {10.5281/zenodo.14538442},
langid = {en}
}