<p>The translation of scientific knowledge into societal impact represents an increasingly complex process, in which knowledge brokering organizations (KBOs) serve as vital intermediaries connecting diverse stakeholders and facilitating knowledge exchange. This study investigates the participation strategies of three specific types of KBOs—libraries, museums, and archives—based on case data from the Research Excellence Framework. Employing computational methods including fuzzy matching and the Perception Agent tool, we systematically extracted and analyzed descriptive segments documenting institutional engagement. Our multimethod approach examines participation detail from both semantic and quantitative perspectives, utilizing BERTopic for thematic modeling and constructing an Institutional Topic Affiliation Network (ITAN) through semantic similarity analysis to identify potential collaborative domains. The results demonstrate significant divergence in participation distribution and pattern across the three organizational types. While each institution maintains distinct functional characteristics, observed overlaps in specific cases suggest <i>complementary roles within the broader knowledge ecosystem</i>. The analysis reveals specialized organizational profiles: museums operate as experiential learning environments, libraries function as public knowledge hubs, and archives serve as foundations of social trust. For enhanced impact, we propose tripartite collaboration in domains including regional cultural development, public issues, and performing arts preservation, alongside targeted bilateral partnerships addressing themes such as conflict memory studies and gender and community history. These integrated approaches offer robust mechanisms for advancing science communication effectiveness.</p>