2022, Vol.27, No.3

Journal of The Korea Society of Computer and Information 2023 KCI Impact Factor : 0.65

pISSN : 1598-849X / eISSN : 2383-9945
https://journal.kci.go.kr/jksci

In the event of a maritime distress accident, rapid search and rescue operations using rescue assets are very important to ensure the safety and life of drowning person’s at sea. In this paper, we analyzed the surface layer current in the northwest sea area of Ulleungdo by applying machine learning such as multiple linear regression, decision tree, support vector machine, vector autoregression, and LSTM to the meteorological information collected from the maritime observation buoy. And we predicted the drowning person’s route at sea based on the predicted current direction and speed information by constructing each prediction model. Comparing the various machine learning models applied in this paper through the performance evaluation measures of MAE and RMSE, the LSTM model is the best. In addition, LSTM model showed superior performance compared to the other models in the view of the difference distance between the actual and predicted movement point of drowning person.

In this study we leveraged a probing method to evaluate how a pre-trained L2 LSTM language model represents sentences with relative and coordinate clauses. The probing experiment employed adapted models based on the pre-trained L2 language models to trace the syntactic properties of sentence embedding vector representations. The dataset for probing was automatically generated using several templates related to different sentence structures. To classify the syntactic properties of sentences for each probing task, we measured the adaptation effects of the language models using syntactic priming. We performed linear mixed-effects model analyses to analyze the relation between adaptation effects in a complex statistical manner and reveal how the L2 language models represent syntactic features for English sentences. When the L2 language models were compared with the baseline L1 Gulordava language models, the analogous results were found for each probing task. In addition, it was confirmed that the L2 language models contain syntactic features of relative and coordinate clauses hierarchically in the sentence embedding representations.

The purpose of this study is to explore the implications for the systematic operation of the AI curriculum by analyzing the current status of the AI major curriculum in universities. To this end, This study analyzed the relevant curriculum of domestic universities(a total of 51 schools) and overseas QS Top 10 universities based on the industry demand-based standard of AI major curriculum developed through prior research. The main research results are as follows. First, in the case of domestic universities, Python-centered programming subjects were lacking. Second, there were few subjects for advanced learning such as AI application and convergence. Third, the subjects required to perform the AI developer job were insufficient. Fourth, in the case of colleges, the ratio of AI mathematics-related subjects was low. Based on these results, this study presented implications for the systematic operation of the AI major education.