Information processing apparatus configured to perform debug analysis based on a distributed representation vector for lines of a debug log, method, and non-transitory recording medium

1. An information processing apparatus comprising: circuitry configured to,

acquire an operation log of a target device over a network, the target device being a machine tool connected to the information processing apparatus via the network, the target device including a plurality of modules therein,

generate a first vector of a distributed representation indicating the operation log such that, for each line of the operation log, the first vector is multi-dimensional vector uniquely representing a respective one of the lines of the operation log,

calculate similarity the first vector and a second vector of a distributed representation based on another operation log to identify the first vector,

detect that the target device includes a defective module in response to a determination that an identification result of the first vector matches past defect information, and

disable operation of the defective module of the target device.

2. The information processing apparatus of claim 1 , wherein the circuitry is further configured to:

learn an embedded matrix that predicts surrounding lines from each line of the operation log; and

generate the first vector based on the embedded matrix.

3. The information processing apparatus of claim 1 , wherein the circuitry is further configured to:

apply preprocessing to the operation log that is acquired, the preprocessing contributing to generation of the first vector of the distributed representation from the operation log; and

generate the first vector from the operation log having been applied with the preprocessing.

4. The information processing apparatus of claim 1 , wherein the circuitry is further configured to

calculate similarity between the first vector, and the second vector prepared based on dictionary data.

5. The information processing apparatus of claim 4 , wherein the circuitry is further configured to

display on a display, the identification result of the first vector,

wherein, when the similarity is greater than a first threshold, the circuitry displays on the display the similarity as the identification result, and

when the similarity is no greater than the first threshold and is greater than a second threshold, the circuitry identifies sub-classification using fixed logic corresponding to each classification of the similarity, and displays the identified sub-classification as the identification result.

6. The information processing apparatus of claim 4 , wherein the circuitry is further configured to

calculate the similarity between the first vector, and the second vector that is based on the dictionary data and stored in an external device.

7. The information processing apparatus of claim 1 , wherein the circuitry is further configured to

accumulate the operation log as the dictionary data after identifying the first vector.

8. The information processing apparatus of claim 4 , wherein the circuitry is further configured to:

divide the operation log into a plurality of log portions according to a state of the operation log; and

generate the first vector for each log portion.

9. The information processing apparatus of claim 8 , wherein the circuitry is further configured to:

set division condition of the operation log; and

divide the operation log according to the division condition.

10. The information processing apparatus of claim 8 , wherein the circuitry is further configured to:

perform identification sequentially on each log portion of the first vector; and

cancel the identification on the first vector when the similarity between the first vector and the second vector is larger than a third threshold.

11. The information processing apparatus of claim 8 , wherein the circuitry is further configured to

integrate weak discriminators each being an identification result of the log portion of the first vector, into an integrated weak discriminator, and

perform identification on the integrated weak discriminator.

12. The information processing apparatus of claim 1 , wherein the circuitry is further configured to:

acquire the operation log for each of a plurality of modules;

generate the first vector for the operation log corresponding to each module;

perform identification on the first vector of each module; and

perform identification on an integrated weak discriminator, the integrated weak discriminator being generated by integrating weak discriminators each being the identification result of the first vector of each module.

13. The information processing apparatus of claim 1 , wherein the circuitry is further configured to

perform identification on an integrated weak discriminator, the integrated weak discriminator being generated by integrating weak discriminators each being the identification result of the first vector and data other than the operation log.

14. The information processing apparatus of claim 1 , wherein the circuitry is further configured to

when software that solves the defect is available, update software of the defective module with the software that solves the defect.

15. The information processing apparatus of claim 1 , wherein the circuitry is further configured to

determine whether the identification result matches past failure information, and

automatically issues an order of replacement parts according to the past failure information when the identification result matches the past failure information.

16. A method comprising;

acquiring an operation log of a target device over a network, the target device being a machine tool connected to the information processing apparatus via the network, the target device including a plurality of modules therein;

generating a first vector of a distributed representation indicating the acquired operation log such that, for each line of the operation log, the first vector is multi-dimensional vector uniquely representing a respective one of the lines of the operation log;

identifying the first vector based on a calculation of the first vector and a second vector of a distributed representation based on another operation log;

detecting that the target device includes a defective module in response to a determination that an identification result of the first vector matches past defect information; and

disabling operation of the defective module of the target device.

17. A non-transitory recording medium storing a plurality of instructions which, when executed by one or more processors, causes the processors to perform a method comprising:

acquiring an operation log of a target device over a network, the target device being a machine tool connected to the information processing apparatus via the network, the target device including a plurality of modules therein;

generating a first vector of a distributed representation indicating the acquired operation log such that, for each line of the operation log, the first vector is multi-dimensional vector uniquely representing a respective one of the lines of the operation log; and

identifying the first vector based on a calculation of the first vector and a second vector of a distributed representation based on another operation log;

detecting that the target device includes a defective module in response to a determination that an identification result of the first vector matches past defect information; and

disabling operation of the defective module of the target device.