ANALYTICS FOR APPLICATION PROGRAMMING INTERFACES

• US 20150128156A1
• Filed: 11/07/2013
• Published: 05/07/2015
• Est. Priority Date: 11/07/2013
• Status: Active Grant

• Current Assignee: Accenture Global Services Limited (Accenture PLC)
• Original Assignee: Accenture Global Services Limited (Accenture PLC)
• Inventors: Zhu, Qian, Tung, Teresa, Grimmer, Benjamin
• Granted Patent: US 9,146,787 B2
• Time in Patent Office: Days
• US Class Current: 719/328
• CPC Class Codes: G06F 11/302 where the computing system component is a software system
• CPC Class Codes: G06F 11/3072 where the reporting involves data filtering, e.g. pattern matching, time or event triggered, adaptive or policy-based reporting
• CPC Class Codes: G06F 11/3409 for performance assessment
• CPC Class Codes: G06F 21/552 involving long-term monitoring or reporting
• CPC Class Codes: G06F 8/74 Reverse engineering; Extracting design information from source code
• CPC Class Codes: G06F 8/75 Structural analysis for program understanding
• CPC Class Codes: G06F 9/541 via adapters, e.g. between incompatible applications
• CPC Class Codes: G06N 5/045 Explanation of inference; Explainable artificial intelligence [XAI]; Interpretable artificial intelligence
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: H04L 67/01 Protocols
• CPC Class Codes: H04L 67/60 Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources admission control or resource allocation H04L47/70

Abstract

API analytics systems and methods are provided. Frequently occurring API usage patterns may be determined from API call data collected from across an API ecosystem. Alternatively or in addition, a classification structure, such as a decision tree, may be generated from the API usage patterns. A type of activity that resulted in a set of API calls being invoked may be determined from the classification structure. A similarity or difference between the set of API calls and the frequently occurring API usage patterns may also be determined and/or identified.

Claims

1. An application programming interface (API) analytics system comprising:

a memory comprising

API call data that identifies a set of API calls invoked in response to an unknown activity, and

predetermined API usage patterns that each identify a series of API calls performed as a result of a corresponding use case; and

a usage identification module configured to form truncated API call data in which duplicated API calls in the set of API calls are consolidated in the truncated set of API calls, the usage identification module further configured to determine a type of the unknown activity based on the truncated API call data and on a classification structure determined from the predetermined API usage patterns.

2. The system of claim 1, wherein each API call in the set of API calls is a respective programmatic procedure.

3. The system of claim 1, wherein the usage identification module is configured to determine that one of the predetermined API usage patterns matches the truncated API call data, and to identify any differences between the truncated API call data and one of the predetermined API usage patterns that matches the truncated API call data.

4. The system of claim 1 further comprising a pattern classification module configured to generate the classification structure from the predetermined API usage patterns, wherein the usage identification module is configured to determine the type of the unknown activity through an application of the truncated API call data to the classification structure.

5. The system of claim 1 wherein an API call repeated n times in the API call data is represented as a transition from the API call to the API call in the truncated API call data when n is greater than a repetition threshold.

6. The system of claim 1 further comprising a pattern recognition module configured to determine the predetermined API usage patterns from API calls that are detected when the corresponding use case is performed.

7. The system of claim 1 further comprising an API monitor that monitors an API ecosystem and collects the API call data.

8. A non-transitory computer readable storage medium comprising computer executable instructions, the computer executable instructions executable by a processor, the computer executable instructions comprising:

instructions executable to retrieve application programming interface (API) call data that identifies sets of API calls invoked as a result of performance of a use case, wherein API requests that invoked API calls in each one of the sets of API calls share a corresponding common context;

instructions executable to generate a first truncated API call data from the API call data, wherein duplicated API calls in each of the sets of API calls are consolidated in the first truncated API call data;

instructions executable to identify API usage patterns in the first truncated API call data that occur above a threshold frequency; and

instructions executable to identify a similarity between the identified API usage patterns and a second truncated API call data that identifies API calls invoked during an unknown activity.

9. The computer readable storage medium of claim 8 further comprising instructions executable to convert the first truncated API call data to directed graphs.

10. The computer readable storage medium of claim 9 further comprising instructions executable to identify the API usage patterns that occur above the threshold frequency in the first truncated API call data with a frequent subgraph mining algorithm applied to the directed graphs.

11. The computer readable storage medium of claim 8, wherein each API call in the set of API calls is for a corresponding programmatic procedure that was invoked over a network in response to a corresponding one of the API requests.

12. The computer readable storage medium of claim 8 further comprising instructions executable to monitor an API ecosystem and collect the API call data in memory.

13. The computer readable storage medium of claim 8 further comprising instructions executable to group the API call data into the sets of API calls based on a context included in the API requests that invoked the API calls in each one of the sets of API calls.

14. The computer readable storage medium of claim 8 further comprising instructions executable to group the API call data into the sets of API calls based on a context identified by an API management gateway.

15. A method to determine an activity type that resulted in an API (Application Programming Interface) workflow pattern, the method comprising:

retrieving a first API call data that identifies sets of API calls detected during performance of a plurality of use cases, wherein each one of the sets of API calls results from performance of a respective one of the use cases;

generating a truncated API call data from the first API call data by consolidating duplicated API calls in each of the sets of API calls with a processor;

detecting an API usage pattern in the truncated API call data with the processor, the API usage pattern identifying a series of truncated API calls that is repeated within and/or across at least one of the sets of API calls;

retrieving a second API call data that identifies a set of API calls that are invoked during an unknown activity; and

determining, with the processor based on the second API call data and the API usage pattern, a type of activity that resulted in invocation of the set of API calls identified in the second API call data.

16. The method of claim 15, further comprising determining a difference between the API usage pattern and an extracted API usage pattern that is extracted from the set of API calls invoked during the unknown activity.

17. The method of claim 15, wherein determining the type of activity comprises:

generating a classification decision tree from the API usage pattern; and

determining the type of activity by applying attributes of the set of API calls, which is represented by truncated API call data, to the classification decision tree.

18. The method of claim 15 further comprising generating the second API call data by monitoring an API ecosystem.

19. The method of claim 18, wherein the API ecosystem is monitored and the type of activity that resulted in the set of API calls is determined in real-time.

20. The method of claim 15 further comprising producing an alert in response to a determination that the type of activity that resulted in the set of API calls is a predetermined type of activity.