Abstract Urban and rural grassroots governance is the cornerstone of the modernization of national governance, and strengthening effective public participation is also the continuation and sublimation of the strategy of poverty alleviation. This paper utilizes gene structure for expression, constructs the RMUGS model, and then constructs a three-level retrieval algorithm based on the various levels of the model to analyze big data on the results achieved by public participation in urban-rural grassroots governance in Z town of Guangdong province. Town Z has been able to achieve good results in urban-rural grassroots governance, as shown in the data analysis results. Significant results have been achieved in the improvement of infrastructure, with the overall degree of improvement reaching more than 75%, and the highest improvement of public lighting reaching 90% of the degree of improvement. The number of people actively participating has gone up from 60.33% in 2020 to 77.95% in 2022, with general residents without positions experiencing the greatest increase. The above data analyzed by the RMUGS model clearly shows that public participation in urban and rural grassroots governance has a very high degree of effectiveness, which provides effective directions and ideas for future grassroots governance.