kent0304 / semantic-scoring

Automatic Scoring for Picture Description
0 stars 0 forks source link

Automatic Scoring for Picture Description

Automatic grading system for language learners.

About

In language learning, training output skill such as speaking and writing is vital in order to retain the learned knowledge. However, scoring descriptive questions by humans would be costly, and this is why automatic scoring systems attract attention. In this research, we try to realize an automatic scoring system for picture description. Concretely, (i) we first analyze the trends of errors that English learners would make, (ii) then create a pseudo dataset by artificially mimicking the errors, and (iii) finally consider a model that judges whether a given pair of a picture and a sentence is valid or not. In experiments, we trained the model with the created pseudo data and evaluate it with the answers provided by actual learners. From experimental results, we found that our model outperforms a random agent.

Picture Description

In this task, language learners are asked to describe a picture.

Model

Judge the correctness of the answer based on the semantic features of the picture and answer.