Abstract: 【Purpose/significance】With the explosive growth of information in today's society, the key to achieving breakthrough re⁃ search progress and improving the efficiency of academic research lies in how to extract valuable information from diverse technologi⁃ cal data sources. Furthermore, it involves the discovery of inter-disciplinary research opportunities and the combination of relation⁃ ships within valuable academic domains.【Method/process】Establishing a relational analysis framework among various disciplines, in⁃ corporating the perspective of evolutionary theory, involves extracting technologies and their impacts across disciplines through trigger words and SAO structure analysis to form knowledge genes. Constructing databases for knowledge genes in each discipline, integrating sentiment analysis results from social comments, and utilizing a heterogeneous graph model to build a network of relationships among elements. Finally, employing the results of demand factor analysis to naturally filter alternative knowledge genes, obtaining evolved knowledge genes in the context of interdisciplinary intersections, thus realizing the discovery of opportunities for cross-disciplinary collaborative research.【Result/conclusion】In the perspective of knowledge evolution, technologies, impacts, and societal demands across different disciplines are abstracted as individual genes with relationships to natural selection. The research findings indicate that the introduction of the knowledge evolution perspective can effectively enhance the efficiency of discovering opportunities for in⁃ terdisciplinary collaborative research.【Innovation/limitation】The introduction of the concept of knowledge evolution has achieved a precise interpretation of the multidimensional relationships among technological points, impacts, and practical demands across differ⁃ ent disciplines. This provides a reliable reference for research guidance and knowledge services. However, this method has limitations in aspects such as mapping knowledge gene networks and quantitative assessment methods.