The development trend of research discipline has always been an important topic of concern to researchers, but the existing analysis methods cannot explore the changes of hotspots while excavating the hotspots of discipline. Therefore, by building a model of keyword frequency evolution based on word frequency analysis and Co-word Analysis fusion can reveal the changes of hot topics and their class structure in a panoramic way. Based on the three journals of library science and information science and the national social science fund, the frequency evolution network model based on word frequency change is constructed to reveal the research hotspots and the theme structure changes. By setting the threshold of word frequency growth to screen keywords, not only the development trend of research is effectively excavated, but also the similarities and differences of research hotspots under different classification conditions are compared. Compared with the g-index method, it can effectively excavate the emergence time of subject hotspots in advance, which provides an important reference for the research of scientific researchers.