A Novel Approach of Automatic Judgment for Subjective Questions

Most subjective questions in many traditional examination systems are still scored manually by teachers with low efficiency. This paper proposes a novel approach that develops a similarity calculation model, which combines the Word2Vec model and the keyword weight TF-IDF algorithm to automatically score subjective questions. The paper utilizes TF-IDF method to obtain keywords and their weights, to calculate word vectors and sum of weights and get a vector representation of each text; the model proposes the text-similarity vectors via the cosine similarity formulas. The experimental results show that the proposed approach can be utilized in various fields without labeled training sets in advance and takes less calculation time. The relative errors of the model and manual scores are maintained less than 10%, the average accuracy is as high as 95%.