Correction of reaction rules databases by active learning

1. A method of reaction rules database correction, comprising:

receiving a first user-input correction to a first reaction rule in a reaction rules database;

locating a second reaction rule in the reaction rules database that is similar to the first reaction rule;

calculating a correctness score for the located second reaction rule;

determining that the correctness score for the located second reaction rule is below a threshold correctness score;

presenting, in response to the determining that the correctness score for the located second reaction rule is below the threshold correctness score, the located second reaction rule to a user;

receiving a second user-input correction to the located second reaction rule;

updating the reaction rules database to include the second user-input correction and

learning, based on the updating, a function for predicting correctness scores for additional reaction rules in the reaction rules database.

2. The method of claim 1 , further comprising:

locating an additional reaction rule in the reaction rules database that is similar to the located second reaction rule;

calculating a correctness score for the additional reaction rule;

determining that the correctness score for the additional reaction rule is below the threshold correctness score;

presenting, in response to the determining that the correctness score for the additional reaction rule is below the threshold correctness score, the additional reaction rule to the user;

receiving a user-input correction to the additional reaction rule; and

updating the reaction rules database to include the user-input correction to the additional reaction rule.

3. The method of claim 1 , further comprising:

receiving a user-input confidence level for the second user-input correction;

determining that the user-input confidence level is below a threshold confidence level;

presenting, in response to the determining that the user-input confidence level is below the threshold confidence level, the located second reaction rule to a second user;

receiving a response from the second user confirming the second user-input correction; and

updating the reaction rules database to include the confirmation from the second user.

4. The method of claim 1 , wherein the presenting the located second reaction rule includes automatically opening a new window in a user interface of a computer-assisted synthetic design program.

5. The method of claim 1 , further comprising generating an alert upon determining that the correctness score for the located second reaction rule is below the threshold correctness score.

6. The method of claim 1 , wherein the first reaction rule and the located second reaction rule include incorrect bonds.

7. The method of claim 1 , wherein the correctness score is a measure of how similar the located second reaction rule is to the first reaction rule.

8. The method of claim 1 , wherein the reaction rules database includes reaction rules extracted from at least one source selected from a group consisting of at least one scientific article, at least one book, and at least one patent.

9. A system, comprising:

at least one processing component;

at least one memory component;

a user interface;

a reaction rules database; and

a reaction rules correction module, comprising:

a machine learning component, executing on the at least one processing component, configured to:

receive a first user-input correction to a first reaction rule in the reaction rules database;

locate a second reaction rule in the reaction rules database that is similar to the first reaction rule;

calculate a correctness score for the located second reaction rule;

determine that the correctness score for the located second reaction rule is below a threshold correctness score;

present, in response to the correctness score being below the threshold correctness score, the located second reaction rule to a user, wherein the located second reaction rule is displayed on the user interface;

receive a second user-input correction to the located second reaction rule;

update the reaction rules database to include the second user-input correction to the located second reaction rule; and

based on the update, learn a function for predicting correctness scores for additional reaction rules in the reaction rules database.

10. The system of claim 9 , wherein the machine learning component is further configured to:

receive a user-input confidence level for the second user-input correction;

determine that the user-input confidence level is below a threshold confidence level;

present, in response to the user-input confidence level being below the threshold confidence level, the located second reaction rule to a second user;

receive a response from the second user confirming the second user-input correction; and

update the reaction rules database to include the confirmation from the second user.

11. The system of claim 9 , wherein the machine learning component is further configured to generate an alert upon determining that the correctness score for the located second reaction rule being below the threshold correctness score.

12. The system of claim 9 , wherein the reaction rules correction module further comprises a retrosynthetic analysis component.

13. The system of claim 9 , wherein the first reaction rule and the located second reaction rule include incorrect bonds.

14. The system of claim 9 , wherein the correctness score for the located second reaction rule is a measure of how similar the located second reaction rule is to the first reaction rule.

15. A computer program product for data storage management, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the device to perform a method, the method comprising:

receiving a first user-input correction to a first reaction rule in a reaction rules database;

locating a second reaction rule in the reaction rules database that is similar to the first reaction rule;

calculating a correctness score for the located second reaction rule;

determining that the correctness score for the located second reaction rule is below a threshold correctness score;

presenting, in response to the determining that the correctness score for the located second reaction rule is below a threshold correctness score, the located second reaction rule to a user;

receiving a second user-input correction to the second reaction rule;

updating the reaction rules database to include the second user-input correction to the located second reaction rule; and

learning, based on the updating, a function for predicting correctness scores for additional reaction rules in the reaction rules database.

16. The computer program product of claim 15 , further comprising:

locating an additional reaction rule in the reaction rules database that is similar to the located second reaction rule;

calculating a correctness score for the additional reaction rule;

determining that the correctness score for the additional reaction rule is below the threshold correctness score;

presenting, in response to the determining that the correctness score for the additional reaction rule is below the threshold correctness score, the additional reaction rule to the user;

receiving a user-input correction to the additional reaction rule; and

updating the reaction rules database to include the user-input correction to the additional reaction rule.

17. The computer program product of claim 15 , further comprising:

receiving a user-input confidence level for the located second user-input correction;

determining that the user-input confidence level is below a threshold confidence level;

presenting, in response to the determining that the user-input confidence level is below the threshold confidence level, the located second reaction rule to a second user;

receiving a response from the second user confirming the second user-input correction; and

updating the reaction rules database to include the confirmation from the second user.

18. The computer program product of claim 15 , wherein the presenting the located second reaction rule includes automatically opening a new window in a user interface of a computer-assisted synthetic design program user interface.

19. The computer program product of claim 15 , further comprising generating an alert upon determining that the correctness score is below the threshold correctness score.

20. The computer program product of claim 15 , wherein the first reaction rule and the located second reaction rule include incorrect bonds.