Automated identification of website errors

1. A system, comprising:

at least one computing device comprising a processor and a memory; and

machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:

receive a website navigation sequence of a series of web page interactions between a client device and a website;

determine an expected completion time for a next measurement of the client device executing the website navigation sequence based at least in part on a first machine learning model being trained with historical navigation sequence data;

determine an actual completion time for the next measurement of the client device executing the website navigation sequence;

determine an anomaly website event in the website navigation sequence based at least in part on the actual completion time failing to meet a boundary threshold associated with the expected completion time; and

determine the anomaly website event is a website error based at least in part on a second machine learning model being trained with a plurality of previous website errors identified from a plurality of previous website navigation sequences.

2. The system of claim 1 , wherein the first machine learning model is a time series regression model, and the boundary threshold is a range between a lower bound and an upper bound determined based at least in part on the expected completion time.

3. The system of claim 2 , wherein at least one of the upper bound or the lower bound are determined based at least in part on a threshold confidence level and the expected completion time.

4. The system of claim 1 , wherein the first machine learning model is a time series regression model, and the machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:

generate an additive function for the first machine learning model based at least in part on the historical navigation sequence data, the additive function is used to generate the expected completion time for the next measurement.

5. The system of claim 1 , wherein the machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:

display a user interface that includes the anomaly website event for the website navigation sequence;

receive a specification of an anomaly type for the anomaly website event from the user interface; and

update the first machine learning model based at least in part on the anomaly type for the anomaly website event.

6. The system of claim 5 , wherein the machine-readable instructions, when executed by the processor, cause the computing device to at least:

determine that a respective confidence level for one of a plurality of anomaly types fails to meet an accuracy threshold; and

update the second machine learning model based at least in part on the anomaly type for the anomaly website event based at least in part on the respective confidence level failing to meet the accuracy threshold.

7. The system of claim 1 , wherein determining the anomaly website event is the website error further comprises machine-readable instructions that, when executed by the processor, cause the computing device to at least:

determine a respective probability level for each of a plurality of anomaly types for the anomaly website event based at least in part on the second machine learning model, wherein the plurality of anomaly types comprise the website error; and

select the website error from the plurality of anomaly types based at least in part on the website error having a highest probability level among the plurality of anomaly types.

8. A method, comprising

receiving, by a computing device, a website navigation sequence of a client device interacting with a website;

determining, by the computing device, a predicted value for a next measurement of the client device executing the website navigation sequence based at least in part on a first machine learning model being trained with historical navigation sequence data;

determining, by the computing device, an actual value for the next measurement of the client device executing the website navigation sequence;

determining, by the computing device, an anomaly website event in the website navigation sequence based at least in part on the actual value failing to meet a boundary threshold associated with the predicted value; and

determining, by the computing device, an anomaly type for the anomaly website event based at least in part on a second machine learning model being trained with a plurality of previous anomaly website events from a plurality of previous website navigation sequences being classified as a plurality of previous anomaly types.

9. The method of claim 8 , wherein the anomaly type comprises at least one of: a website error, a promotional offer, or a new tag.

10. The method of claim 8 , wherein the first machine learning model is a time series regression model, and the boundary threshold is a range between a lower bound and an upper bound determined based at least in part on the predicted value.

11. The method of claim 8 , wherein the first machine learning model is a time series regression model, and further comprising:

generating an additive function for the first machine learning model based at least in part on the historical navigation sequence data, the additive function being used to generate the predicted value for the next measurement.

12. The method of claim 8 , further comprising:

displaying, by the computing device, a user interface that includes the anomaly website event for the website navigation sequence;

receiving, by the computing device, a specification of an updated anomaly type for the anomaly website event from the user interface; and

updating, by the computing device, the first machine learning model based at least in part on the updated anomaly type for the anomaly website event, wherein the updated anomaly type replaces the anomaly type.

13. The method of claim 12 , further comprising:

determining, by the computing device, that a respective confidence level for one of a plurality of anomaly types fails to meet an accuracy threshold; and

updating, by the computing device, the second machine learning model based at least in part on the anomaly type for the anomaly website event based at least in part on the respective confidence level failing to meet the accuracy threshold.

14. The method of claim 8 , further comprising:

determining, by the computing device, a respective probability level for each of a plurality of anomaly types for the anomaly website event based at least in part on the second machine learning model; and

selecting, by the computing device, that a website error from the plurality of anomaly types for the anomaly website event based at least in part on the website error having a highest probability level among the plurality of anomaly types.

15. The method of claim 8 , wherein the boundary threshold is a range between a lower bound and an upper bound, at least one of the lower bound and the upper bound are determined based at least in part on a percentage of a total of the website navigation sequences.

16. A system, comprising:

at least one computing device comprising a processor and a memory; and

machine-readable instructions stored in the memory that, when executed by the processor, cause the computing device to at least:

receive clickstream data that comprises a plurality of device identifiers, a plurality of device events, and a plurality of session identifiers, the plurality of device identifiers representing a plurality of client devices that have interacted with a website, the plurality of session identifiers representing a period of time in which one of the plurality of client devices generated at least one of the plurality of device events, the plurality of device events representing a plurality of interactions with the website;

identify a period of time associated with a respective session identifier and a respective device identifier for a respective client device from the clickstream data;

identify a subset of the plurality of device events that occurred within the period of time of the respective session identifier for the respective client device, the subset of the plurality of device events are identified based at least in part on an event time stamp associated with individual ones of the plurality of device events being within the period of time; and

generate a website navigation sequence of the subset of the plurality of device events for the respective client device based at least in part on the event time stamp for each of the subset of the plurality of device events.

17. The system of claim 16 , wherein the website navigation sequence comprises a sequence of a plurality of web pages associated with the website that have been displayed on the respective client device.

18. The system of claim 16 , wherein the period of time of the respective session identifier comprises a completion time for the respective client device to complete a task over a plurality of web pages.

19. The system of claim 16 , wherein the plurality of device events comprise at least one of: the respective client device entering data into a particular data field, the respective client device interacting with a user interface component on a web page, or an amount of time spent viewing a portion of a web page.

20. The system of claim 16 , wherein the machine-readable instructions, when executed by the processor, cause the computing device to at least:

determine a plurality of website navigation sequences for a remaining portion of the plurality of device identifiers based at least in part on the clickstream data; and

determine a device subset of the plurality of device identifiers that have a particular website navigation sequence among the plurality of website navigation sequences that corresponds to the website navigation sequence.