Research And Application Of Subjective Test Based On Equipment Maintenance

Equipment maintenance training goal is to make maintenance person to master the equipment maintenance knowledge, with equipment maintenance skills, capable of equipment maintenance work. With the development of educational science and computer technology, the way of knowledge inspection gradually develops toward the paperless and electronic direction, and it promotes the development of education to the intellectualized direction. As one of the key technologies of smart education, automatic measurement of exam papers plays an important role in smart education.In this paper, the automatic evaluation of subjective questions in equipment maintenance field is studied and applied in the knowledge assessment of equipment maintenance training.In equipment maintenance field, automated examination and evaluation in the subject area of automatic marking technology has been very mature. This paper focuses on subjective questions automatic assessment in equipment maintenance field, firstly, this paper introduces the present situation of subjective questions automatic evaluation in the field of intelligent education. According to the calculation method of semantic similarity, this paper divides it into three categories: simple word matching, dependent on world knowledge , With the aid of large-scale corpus, introduce the typical evaluation model for all kinds of evaluation programs, summarize the advantages and disadvantages of various evaluation schemes, give the improvement plan for the shortcomings. Based on the previous research, this paper improves the simple word matching algorithm in text similarity calculation. Using word vector and segment vector to express semantic information, this paper proposes an automatic subjective evaluation model based on equipment maintenance. Firstly, fuzzy evaluation is carried out by knowledge tree, and then use the multi-factor semantic similarity algorithm to get accurate score. During the process of constructing the knowledge tree, the method of reference concept graph and the tree data structure are used to construct hierarchical knowledge tree according to the level and relation of knowledge points. In the multi-factor text semantic similarity algorithm, similarity degree is calculated by the method of words shape matching and text vector distance. Based on the theoretical derivation of the algorithm, the problems and shortcomings of the model are summarized and analyzed through the equipment maintenance exercise experimental data, and the areas that need to be improved and strengthened in the next work are given.As one of the key technologies of smart equipment maintenance,subjective question automatic measurement is the technical difficult of intelligent equipment maintenance education software use the subjective question to measure the degree of knowledge.In this paper, we carry out a detail study on the automatic evaluation of subjective questions with reference answers. On the basis of previous studies, word vector and neural network language model are introduced into the field of automatic measurement of subjective questions. A new automatic measurement model is proposed and applied to equipment maintenance training scene in order to promote the research and application of automatic measurement of subjective questions in equipment maintenance field.