With the rapid development of science and technology, the demand for talents in the engineering field has gradually shifted from traditional technical skills to diversified and complex comprehensive abilities. As an interdisciplinary technology, data mining has become a key tool for modern engineering problem solving, and has put forward new requirements for the ability evaluation of engineering talents. In the era of big data, data mining technology plays a vital role in the field of engineering. It can help engineers find patterns and predict trends from massive data, so as to optimize engineering design and improve project efficiency. Traditional talent evaluation methods often rely on subjective evaluation and qualitative analysis, which limits the accuracy of evaluation to some extent. In recent years, with the development of machine learning and artificial intelligence technology, data mining algorithms began to be introduced into the field of talent evaluation, which provides the possibility for quantitative and objective evaluation of talent ability. These algorithms can dig out the deep information from the multi-dimensional and multi-level data, and provide a more scientific basis for the ability evaluation of engineering talents. MATLAB simulation shows that under certain evaluation criteria, the accuracy and reliability of the ability evaluation of new engineering talents by the data fusion algorithm is better than the dynamic programming algorithm.