The dilemma and path of rural college students returning to their hometowns for employment and entrepreneurship in the post-epidemic period under the background of big data

The dilemma of exploring the children in institutions of higher learning to find jobs or start their own companies of rural origin and return to their hometown is paving the way for college students to find jobs of rural origin. In this paper, a decision tree optimized by a partial least squares regression algorithm is proposed to construct a DT-PLS data classification model in the case of cloud data, and the performance of the DT-PLS model is evaluated for the DT-PLS model. Using college students’ employment and entrepreneurship data from colleges and universities as examples, two factors that influence college students of rural origin to return to the registered permanent residence to find a job or start a company, namely subjective factors and contextual factors, are mined, and data analysis is carried out for the two influencing factors. From the 12 sub-indicators of subjective factors, the mean values of those considered very unconformable, unconformable, uncertain, conformable and very conformable were 17.87%, 15.38%, 10.82%, 16.09% and 39.84%, respectively. Regarding the 10 sub-indicators of contextual factors, the mean values of the percentages of those considered very non-conforming, non-conforming, uncertain, conforming, and very conforming were 6.56%, 20.95%, 13.69%, 26.13%, and 32.67%, respectively. The DT-PLS classification, the main method here, analyzes the current dilemmas of returning college students of rural origin to entrepreneurship and employment and prompts more college students to invest in the grassroots to provide fresh talent blood for rural revitalization.