DETECTING DEVICE, DETECTING METHOD, AND DETECTING PROGRAM
A detection device ( 10 ) learns a machine learning model (typing characteristic model) so as to minimize an abnormality degree of a user himself/herself with respect to typing by machine learning (for example, unsupervised machine learning) using free typing of the user himself/herself. Thereafter, the detection device ( 10 ) calculates a typing abnormality degree of a detection target by using the machine learning model after learning, judges that the typing is performed by a person other than the user himself/herself, and detects abnormality when the calculated abnormality degree exceeds a predetermined threshold value.
1 . A detection device comprising to:
acquire keystroke information of free typing of a user himself/herself;
learn a machine learning model so as to minimize an abnormality degree with respect to the keystroke information of free typing of the user himself/herself by machine learning with respect to the keystroke information of free typing of the user himself/herself;
acquire keystroke information of typing of a detection target;
calculate a typing abnormality degree indicated by the keystroke information of the detection target by using the learned machine learning model;
determine that the typing of the detection target is performed by a person other than the user himself/herself and detect abnormality when the calculated abnormality degree exceeds a predetermined threshold value; and
output a detection result of the abnormality.
2 . The detection device according to claim 1 , wherein
the machine learning is unsupervised machine learning with respect to the keystroke information of free typing of the user himself/herself.
3 . The detection device according to claim 2 , wherein
the machine learning model is a model using Variational AutoEncoder (VAE).
4 . The detection device according to claim 1 , wherein
the keystroke information is information indicating a series of keys pressed in typing, a time point when each of the keys is pressed, and a time point when each of the keys is released.
5 . A detection method which is executed by a detection device, the detection method comprising:
acquiring keystroke information of free typing of a user himself/herself;
learning a machine learning model so as to minimize an abnormality degree with respect to the keystroke information of free typing of the user himself/herself by unsupervised machine learning with respect to the keystroke information of free typing of the user himself/herself;
acquiring keystroke information of typing of a detection target;
calculating a typing abnormality degree indicated by the keystroke information of the detection target by using the learned machine learning model;
judging that the typing of the detection target is performed by a person other than the user himself/herself and detecting abnormality when the calculated abnormality degree exceeds a predetermined threshold value; and
outputting a detection result of the abnormality.
6 . A computer-readable non-transitory recording medium storing computer-executable program instructions that when executed by a processor cause a computer to execute a detection program comprising:
acquiring keystroke information of free typing of a user himself/herself;
learning a machine learning model so as to minimize an abnormality degree with respect to the keystroke information of free typing of the user himself/herself by unsupervised machine learning with respect to the keystroke information of free typing of the user himself/herself;
acquiring keystroke information of typing of a detection target;
calculating a typing abnormality degree indicated by the keystroke information of the detection target by using the learned machine learning model;
judging that the typing of the detection target is performed by a person other than the user himself/herself and detecting abnormality when the calculated abnormality degree exceeds a predetermined threshold value; and
outputting a detection result of the abnormality.
7 . The detection method according to claim 5 , wherein
the machine learning is unsupervised machine learning with respect to the keystroke information of free typing of the user himself/herself.
8 . The detection method according to claim 7 , wherein
the machine learning model is a model using Variational AutoEncoder (VAE).
9 . The detection method according to claim 5 , wherein
the keystroke information is information indicating a series of keys pressed in typing, a time point when each of the keys is pressed, and a time point when each of the keys is released.
10 . The computer-readable non-transitory recording medium according to claim 6 wherein the detection method further comprises:
the machine learning is unsupervised machine learning with respect to the keystroke information of free typing of the user himself/herself.
11 . The computer-readable non-transitory recording medium according to claim 10 wherein the detection method further comprises:
the machine learning model is a model using Variational AutoEncoder (VAE).
12 . The computer-readable non-transitory recording medium according to claim 6 wherein the detection method further comprises:
the keystroke information is information indicating a series of keys pressed in typing, a time point when each of the keys is pressed, and a time point when each of the keys is released.