Abstract: A clinical trial main keyword prediction method executed by a clinical trial main keyword prediction device according to an embodiment of the present invention comprises the steps of: when question data is received from a user terminal, removing predetermined clinical non-use words from the question data; extracting words on the basis of blanks from clinical trial title data obtained by removing the predetermined clinical non-use words; executing a morpheme analysis of each word to generate tokens each including a pair of a word and a morpheme value; inputting the tokens in a pre-trained learning model, and extracting tokens matching the tokens of the question data from the learning model; arranging the tokens on the basis of a weight pre-assigned to each token and a frequency indicated by a label pre-assigned to each token; and converting the arranged tokens into text and providing the text to the user terminal.

Abstract: A method for providing similar clinical trial data, executed by a server for providing similar clinical trial data according to an embodiment of the present invention, comprises the steps of: when receiving clinical trial data from a user terminal, determining a type of the clinical trial data; generating a vector by using each of pieces of metadata of the clinical trial data or generating a vector by tokening words extracted from the clinical trial data according to the type of the clinical trial data; inputting the vector into a pretrained learning model, and calculating a distance between a prestored vector in the learning model and the vector; and measuring a similarity grade according to the distance between the vectors, and extracting and providing clinical trial data having a similarity grade that is less than or equal to a specific grade.

Abstract: A clinical trial main keyword prediction method executed by a clinical trial main keyword prediction device according to an embodiment of the present invention comprises the steps of: when question data is received from a user terminal, removing predetermined clinical non-use words from the question data; extracting words on the basis of blanks from clinical trial title data obtained by removing the predetermined clinical non-use words; executing a morpheme analysis of each word to generate tokens each including a pair of a word and a morpheme value; inputting the tokens in a pre-trained learning model, and extracting tokens matching the tokens of the question data from the learning model; arranging the tokens on the basis of a weight pre-assigned to each token and a frequency indicated by a label pre-assigned to each token; and converting the arranged tokens into text and providing the text to the user terminal.