Information processing method, computer-readable recording medium storing information processing program, and information processing device

A non-transitory computer-readable recording medium storing an information processing program causing a computer to execute a processing, the processing comprising:

• acquiring an unprocessed sentence;
• assigning a relationship candidate to each pair of a name of a compound and a numerical value without a unit in the acquired sentence;
• acquiring the numerical value without the unit in the pair using an unlearned model;
• predicting the first unit for the numerical value by using a prediction model that predicts a word that appears immediately before or immediately after an input word, wherein to predict the unit for the numerical value the input word is the numerical value;
• acquiring a second unit corresponding to the relationship candidate with reference to a storage;
• determining whether the relationship exists in the relationship database;
• determining whether the first unit matches the second unit if the relationship exists;
• generating learning data regarding the relationship candidate based on the acquired sentence when the first unit matches the second unit;
• and
• training a discrimination model that discriminates a relationship between words using the learning data.

An information processing apparatus comprising:

• a memory;
• and
• a processor coupled to the memory and configured to:
• acquire an unprocessed sentence;
• assign a relationship candidate to each pair of a name of a compound and a numerical value without a unit in the acquired sentence;
• acquire the numerical value without the unit in the pair using an unlearned model;
• predict the first unit for the numerical value by using a prediction model that predicts a word that appears immediately before or immediately after an input word, wherein to predict the unit for the numerical value the input word is the numerical value;
• acquire a second unit corresponding to the relationship candidate with reference to a storage;
• determine whether the relationship exists in the relationship database;
• determine whether the first unit matches the second unit if the relationship exists;
• generate learning data regarding the relationship candidate based on the acquired sentence when the first unit matches the second unit;
• and
• train a discrimination model that discriminates a relationship between words using the learning data.