Deep learning-based pedagogical word recommendation system for predicting and improving vocabulary skills of foreign language learners
A method in which a server recommends a word to a user according to the present specification, includes receiving training data from a network and training an AI model by using the training data; inputting (1) a user vector and (2) a word vector to the AI model, and generating (1) a user embedding vector and (2) a word embedding vector for determining whether the user knows a word related to the word vector, on the basis of the trained AI model; inputting (1) the user embedding vector and (2) the word embedding vector to a function for determining whether the user knows a word related to the word vector; and outputting a result value for predicting whether the user knows a word related to the word vector from the function.
1. A word recommendation method in which a server recommends a word to a user, comprising:
a step of receiving training data from a network and training an AI model by using the training data;
a step of inputting (1) a user vector and (2) a word vector to the AI model, and generating (1) a user embedding vector and (2) a word embedding vector for determining whether the user knows a word related to the word vector, on the basis of the trained AI model;
a step of inputting (1) the user embedding vector and (2) the word embedding vector to a function for determining whether the user knows a word related to the word vector; and
a step of outputting a result value for predicting whether the user knows a word related to the word vector from the function,
wherein the AI model includes (1) the user embedding model for generating the user embedding vector, (2) the word embedding model for generating the word embedding vector of a word related to word information, and (3) the function, and
wherein the function is an arbitrary similarity scoring function for determining similarity between word embedding vectors acquired from the word embedding model.
2. The word recommendation method according to claim 1 , wherein (1) the user embedding model and (2) the word embedding model are optimized to encode (1) the user vector and (2) the word vector closest to each other.
3. The word recommendation method according to claim 2 , wherein the function outputs the result value on the basis of proximity between (1) the user embedding vector and (2) the word embedding vector.
4. The word recommendation method according to claim 3 , wherein the function outputs the result value on the basis of the following equation:
y ij =σ(ƒ( u,v )),
wherein the u, the v, and the ŷ ij denote the user vector, the word vector, and the result value, respectively.
5. The word recommendation method according to claim 1 , further comprising a step of transmitting recommended word information to a terminal of the user on the basis of the result value.
6. The word recommendation method according to claim 5 , wherein the training data includes information of a word added to a vocabulary list for learning by one or more users.
7. A server which recommends a word to a user, comprising:
a communication module;
a memory; and
a processor,
wherein the processor receives training data from a network through the communication module, trains an AI model by using the training data, inputs (1) a user vector and (2) a word vector to the AI model, generates (1) a user embedding vector and (2) a word embedding vector for determining whether the user knows a word related to the word vector, on the basis of the trained AI model, inputs (1) the user embedding vector and (2) the word embedding vector to a function for determining whether the user knows a word related to the word vector, and outputs a result value for predicting whether the user knows a word related to the word vector from the function,
wherein the AI model includes (1) the user embedding model for generating the user embedding vector, (2) the word embedding model for generating the word embedding vector of a word related to word information, and (3) the function, and
wherein the function is an arbitrary similarity scoring function for determining similarity between word embedding vectors acquired from the word embedding model.
8. The server according to claim 7 , wherein the processor transmits recommended word information to a terminal of the user on the basis of the result value through the communication module.