Prosecution Insights
Last updated: November 20, 2026
Application No. 18/954,239

APPLICATION PROGRAMING INTERFACE IDENTIFICATION METHOD AND APPARATUS, MEDIUM, AND ELECTRONIC DEVICE

Non-Final OA §101§103
Filed
Nov 20, 2024
Priority
Nov 21, 2023 — CN 202311560814.6
Examiner
MILLS, FRANK D
Art Unit
2194
Tech Center
2100 — Computer Architecture & Software
Assignee
Beijing Zitiao Network Technology Co., Ltd.
OA Round
1 (Non-Final)
70%
Grant Probability
Favorable
1-2
OA Rounds
1y 4m
Est. Remaining
92%
With Interview

Examiner Intelligence

Grants 70% — above average
70%
Career Allowance Rate
431 granted / 618 resolved
+14.7% vs TC avg
Strong +23% interview lift
Without
With
+22.7%
Interview Lift
resolved cases with interview
Typical timeline
3y 4m
Avg Prosecution
23 currently pending
Career history
653
Total Applications
across all art units

Statute-Specific Performance

§101
16.5%
-23.5% vs TC avg
§103
52.4%
+12.4% vs TC avg
§102
12.1%
-27.9% vs TC avg
§112
12.7%
-27.3% vs TC avg
Black line = Tech Center average estimate • Based on career data from 618 resolved cases

Office Action

§101 §103
DETAILED ACTION Claims 1-20 are rejected under 35 USC § 101, as directed to abstract idea. Claims 9-12 are rejected under 35 USC § 101, as directed to non-statutory matter. Claims 1-20 are rejected under 35 USC § 103. Notice of Pre-AIA or AIA Status The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . Claim Rejections - 35 USC § 101 35 U.S.C. 101 reads as follows: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title. Claims 1-20 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Claims 1, 9, and 13 Step 2A Prong One. Claims 1, 9, and 13 recite matching an Api request path against a history updated multiway tree having wildcard nodes for dynamic parameters, checking whether the matching target is listed, and adding it when absent. A person can compare /aaa/123/aaax with /aaa/*/aaac in a written hierarchy, check the list, and record the pattern once. The claims impose no required data volume, path complexity, or processing rate that would preclude this practical evaluation. These observations and evaluations are mental processes. See MPEP § 2106.04(a)(2), III. Step 2A Prong Two. This judicial exception is not integrated into a practical application. Path extraction is input gathering recited by the information obtained, without a particular acquisition or parsing technique. The tree, historical updating and wildcard matching classify that information; conditional addition records the category without duplication. This improves the list’s contents, not computer functionality. The medium, storage and processing apparatus of claims 9 and 13 are recited only to store and execute these operations. The API context limits the field of use. Together, these limitations implement the classification without a further technical application. See MPEP §§ 2106.05(a), (f), (g), and (h). Step 2B. The claims lack additional elements amounting to significantly more. Maintaining entries performs “Electronic recordkeeping;” retaining and accessing records and programs performs “Storing and retrieving information in memory.” At the claimed level of generality, these are the conventional functions identified in MPEP § 2106.05(d), II, items iii and iv. In the ordered combination, extraction supplies a path, wildcard matching and membership testing evaluate it, and recordkeeping preserves the result. That relationship implements the classification scheme itself; the additional elements supply no further application or inventive concept. See MPEP §§ 2106.05(d), (f), and (g). Claims 2, 10, and 14 These claims add an unmatched path with adjacent components in parent and child relationships. The adjacency specified logical order, without a particular node storage or access mechanism. A person can make the same addition to a written hierarchy. The complete combination maintains classification records without practical integration or significantly more. See MPEP §§ 2106.05(d) and (f). Claims 3, 11, and 15 These claims add periodic traversal to identify dynamic parameter nodes and replace them with preset characters. No traversal technique or required review speed is specified. The period schedules evaluation; replacement records its category for later matching. The complete combination updates classification information through memory access and recordkeeping, without practical integration or significantly more. See MPEP §§ 2106.05(a), (d), and (f). Claims 4 and 16 These claims use a pretrained deep learning model for periodic traversal, without specifying a model architecture or inference procedure. The model is defined by the classification function it performs for character replacement and later matching. The complete combination thus uses the model as a tool for evaluation without practical integration or significantly more. See MPEP §§ 2106.05(a) and (f). Claims 5 and 17 These claims combine all sibling content before model input and classify all children as dynamic when the group includes a dynamic parameter. This specifies the input grouping and classification rule, while leaving the model’s inference technique unspecified. With the inherited replacement and matching operations, the combination implements that information evaluation without a further technical application or inventive concept. See MPEP §§ 2106.05(a) and (f). Claims 6 and 18 These claims update the model from the API list at a second preset period. The update is specified by its data source and schedule, without a training algorithm or parameter update rule. The complete feedback cycle applies the classifier to accumulated task information without requiring improved learning operation, practical integration, or significantly more. See MPEP §§ 2106.05(a) and (f). Claims 7 and 19 These claims restore original characters in the current API list, determine dynamic parameter labels for selected nodes, and train the model using corresponding node and label pairs. Restoration and labeling specify the examples supplied; training remains functionality recited without a learning procedure beyond using those pairs. The complete sequence reconstructs and classifies information for the model without requiring improved learning or computer operation. it therefore supplies neither practical integration nor significantly more. See MPEP §§ 2106.05(a), (d), and (f). Claims 8, 12, and 20 These claims organize paths into site subtrees, identify the request URL’s domain, and match within the corresponding subtree. Domain selection narrows the records examined, as with a categorized written index; no particular computer storage or access mechanism is required. The efficiency follows from categorical selection. Together with inherited matching and listing, this information lookup supplies neither practical integration nor significantly more. See MPEP §§ 2106.05(a), (f), and (g). Claims 9-12 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. Claims 9-12 Claims 9-12 rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The present specification defines a “computer-readable medium” as including “a computer-readable signal medium.” A computer-readable signal medium “may include a data signal propagated in a baseband or as part of a carrier, the data signal carrying computer-readable program code.” Specification, ¶ 0122. Neither the stored program language nor the added program operations exclude that embodiment. A claim encompassing a propagated signal falls outside the statutory categories. See MPEP § 2106.03, II. Claim Rejections - 35 USC § 103 The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. Claims 1-3, 9-11, and 13-15 are rejected under 35 U.S.C. 103 as being unpatentable over Treadway, U.S. PG-Publication No. 2021/0173728 A1 (hereinafter TREADWAY), in view of Sonawane, U.S. Patent No. 9,667,704 B1 (hereinafter SONAWANE), further in view of Savenkov, U.S. PG-Publication No. 2020/0192727 A1 (hereinafter SAVENKOV). Claim 1 TREADWAY discloses an application programing interface identification method. ¶ 0072: The API route modeling procedure builds a tree from observed URLs and identifies qualifying paths as models for APIs. ¶ 0076: When candidate thresholds are met, the procedure “adds the full path of the candidate to the list of models of APIs.” TREADWAY discloses extracting a current application programing interface (API) request path from current network traffic. ¶ 0018: Traffic may be observed in real time, and URL paths are parsed into segments. ¶ 0073: The procedure processes each observed URL and parses its path into segments. TREADWAY discloses wherein the request path tree is a multiway tree that is dynamically updated based on a historical API request path, and a node belonging to a dynamic parameter in the request path tree is a wildcard node. ¶ 0073: Each path segment becomes a tree node whose parent represents the preceding segment. Observed paths share prefixes and branch by segment values. ¶ 0086: Additional traffic updates existing counters and introduces new segment nodes; previously compacted wildcard nodes may retain a wildcard. ¶ 0085: Variable sibling values are combined into a wildcard node named “*”. TREADWAY does not expressly disclose in response to a target path matching the current API request path existing in a request path tree. SONAWANE discloses in response to a target path matching the current API request path existing in a request path tree. col. 9, ll. 35-50: After the final requested segment is matched, the current node is the “final exact matched result node.” col. 9, ll. 1-11: Traversal follows a wildcard child when a literal match is unavailable; otherwise processing may terminate with only a parent match. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the request path processing of TREADWAY to incorporate the complete literal and wildcard tree lookup taught by SONAWANE. One of ordinary skill in the art would be motivated to integrate that lookup into TREADWAY, with a reasonable expectation of success, in order to reduce request matching work by following stored segment mappings instead of searching API group lists for each request, as taught SONAWANE col. 11, l. 54-col.12 l. 9. TREADWAY-SONAWANE does not expressly disclose in response to an API list not including the target path, adding the target path to the API list. SAVENKOV discloses in response to an API list not including the target path, adding the target path to the API list. ¶ 0280: An API schema lookup module receives a detected schema and checks whether a corresponding API schema exists in the library. ¶ 0281: An existing API entry is updated; “Otherwise a new API Schema is created in the API schema library.” ¶ 0277: The library stores all known API schema definitions. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the API inventory admission operation of TREADWAY-SONAWANE to incorporate the library presence check and conditional creation of a new API entry taught by SAVENKOV. One of ordinary skill in the art would be motivated to integrate that conditional creation operation into TREADWAY-SONAWANE, with a reasonable expectation of success, in order to avoid duplicate records and redundant entry creation by checking for a corresponding API entry before creating one, as inferred from SAVENKOV ¶¶ 0280-0281. Claim 2 TREADWAY discloses in response to the target path not existing in the request path tree, adding the current API request path to the request path tree to obtain a new request path tree, wherein in the new request path tree, any path node on the current API request path is a parent node of an adjacent downstream path node. ¶ 0074: If a segment node does not exist, the module “creates a new node with that name.” ¶ 0073: Each URL is parsed into path segments and added to the tree; the “parent of a node represents the preceding path segment.” ¶ 0086: Newly observed path segments are added when they do not exist. Claim 3 TREADWAY discloses traversing the request path tree at a first preset period to identify the node belonging to the dynamic parameter in the request path tree. ¶ 0075: The module “periodically runs a process” that compacts sibling nodes having high variability. ¶ 0085: The procedure compares the number of sibling nodes below a parent with a variance threshold to decide whether to replace them with a wildcard node. TREADWAY disclose replacing the node belonging to the dynamic parameter in the request path tree with a preset character. ¶ 0085: Sibling nodes meeting the variance test become a wildcard node named “*”. Claims 9-11 Claims 9-11 are rejected utilizing the rationale for claims 1-3; the claims are directed to a medium storing instructions corresponding to the method. Claims 13-15 Claims 13-15 are rejected utilizing the rationale for claims 1-3; the claims are directed to a system performing the method. Claims 4-5 and 16-17 are rejected under 35 U.S.C. 103 as being unpatentable over TREADWAY, in view of SONAWANE, further in view of SAVENKOV, further in view of Mittal et al., U.S. Patent No. 10,873,618 B1 (hereinafter MITTAL). Claim 4 MITTAL discloses traversing the request path tree at the first preset period by using a pre-trained deep learning model. col. 11, ll. 18-35: The LSTM string classifier is a deep neural technique trained on samples labeled “dynamic” or “not dynamic,” then applied to child component strings. col. 12, ll. 49-67: Component scores determine whether the children are dynamic, then control returns to evaluate the next tree node. col. 13, ll. 9-23: Dynamic children are collapsed and assigned a generic designator; the resulting tree represents learned API endpoints. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the periodic parameter identification of TREADWAY-SONAWANE-SAVENKOV to incorporate the trained LSTM classifier and associated parameter scoring taught by MITTAL. One of ordinary skill in the art would be motivated to integrate that classification procedure into TREADWAY-SONAWANE-SAVENKOV, with a reasonable expectation of success, in order to distinguish dynamic parameters from static path components more reliably by supplementing sibling counts with learned string evidence and contextual scores, as taught by MITTAL at col. 13, ll.1-8. Claim 5 MITTAL discloses for each node in the request path tree, after all child nodes of the node are spliced, inputting all the child nodes of the node to the deep learning model to determine whether all child nodes of the node include the dynamic parameter. ¶ 0066: The trained deep RNN LSTM classifier processes the “string of characters that makes up a child component (and all of its siblings) to determine whether they are dynamic. ¶ 0056: The procedure traverses the URL component tree and, at each node having children, determines whether those children are dynamic using the Figure 5 procedure. MITTAL discloses in response to the child nodes of the node including the dynamic parameter, determining that all the child nodes of the node belong to the dynamic parameter. ¶ 0073: When the overall group score exceeds its threshold, the group of child components is deemed dynamic. Processing then returns to evaluate the next tree node. ¶ 0075: When a parent node’s children are determined to be dynamic, all child subtrees merge into a single dynamic subtree with a generic designator. Claims 16-17 Claims 16-17 are rejected utilizing the rationale for claims 4-5; the claims are directed to a system performing the method. Claims 6-7 and 18-19 are rejected under 35 U.S.C. 103 as being unpatentable over TREADWAY, in view of SONAWANE, further in view of SAVENKOV, further in view of MITTAL, further in view of Levin et al., U.S. PG-Publication No. 2023/0025896 A1 (hereinafter LEVIN). Claim 6 MITTAL discloses updating the deep learning model. col. 11, ll. 19-36: The RNN LSTM classifier is trained on a large corpus of labeled training samples, using the classes “dynamic” and “not dynamic” for URL component strings. TREADWAY-SONAWANE-SAVENKOV-MITTAL does not expressly disclose the updating the deep learning model according to the API list at a second preset period. LEVIN discloses updating the deep learning model according to the API list at a second preset period. ¶ 0046: Update criteria include elapsed time, with scheduled checks such as every hour. ¶ 0019: Traffic batches may be collected in “periodic intervals according to a schedule,” including every hour. ¶ 0016: A representative node contains underlying nodes and their corresponding labels and content. ¶ 0035: Each representative node stores multiple nodes and their corresponding content. It would have been obvious to a person having ordinary skill in the art before the effective filling date of the claimed invention to modify the training sample preparation of TREADWAY-SONAWANE-SAVENKOV-MITTAL to incorporate the retention of original node values within representative nodes and scheduled maintenance taught by LEVIN. One of ordinary skill in the art would be motivated to integrate that selection into TREADWAY-SONAWANT-SAVENKOV-MITTAL, with a reasonable expectation of success, in order to reuse concrete observed path values for recurring training without collecting the same observations again, based on LEVIN ¶¶ 0016, 0033, 0035, and 0046. Claim 7 MITTAL discloses for each reference path of at least part of reference paths in the reference API list, determining label information of at least part of path nodes on the reference path, wherein the label information is used for representing whether a corresponding path node belongs to the dynamic parameter. col. 11, ll. 19-36: Training samples carry “dynamic” or “not dynamic” labels. MITTAL discloses training the deep learning model by using each piece of the label information and a path node corresponding to each piece of the label information as training samples. col. 11, ll. 19-36: The deep classifier trains on labeled character string samples and classifies URL components. MITTAL does not expressly disclose obtaining a current API list at the second preset period; and replacing preset characters in the current API list with original characters before replacement to obtain a reference API list. LEVIN discloses obtaining a current API list at the second preset period. ¶ 0019: Traffic collection occurs at scheduled intervals, including every hour. ¶ 0046: Tree update criteria are checked on a recurring schedule. LEVIN discloses replacing preset characters in the current API list with original characters before replacement to obtain a reference API list. ¶ 0016: Representative nodes retain underlying nodes and their corresponding labels and content. ¶ 0033: Original user1, user2, and user3 nodes can be compacted into representative node (user). ¶ 0035: Representative nodes store the constituent nodes and their corresponding content. Claims 18-19 Claims 18-19 are rejected utilizing the rationale for claims 6-7; the claims are directed to a system performing the method. Claims 8, 12, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over TREADWAY, in view of SONAWANE, further in view of SAVENKOV, further in view of MITTAL, further in view of Burckart et al., U.S. PG-Publication No. 2005/0055437 A1 (hereinafter BURCKART). Claim 8 BURCKART discloses wherein the request path tree comprises a plurality of request path subtrees respectively corresponding to a plurality of sites, wherein each of the plurality of sites corresponds to at least one domain name. ¶ 0032: Each tree node represents a URL clause. The example separates “www.ibm.com,” “welcome,” and the final page into successive clauses. ¶ 0053: Different clause values select different child nodes; subsequent clauses select further descendants. BURCKART discloses after a current API request is extracted, determining a target domain name in a uniform resource locator where the current API request path is located. ¶ 0043: The received URL is parsed and progressively hashed while its clauses are identified. BURCKART discloses in response to a request path subtree corresponding to a target site including a path matching the current API request path, determining that the target path exists in the request path tree, wherein the target site is a site that corresponds to the target domain name in the plurality of sites. ¶ 0041: The hash entry for a URL clause identifies the subtree to which traversal proceeds. ¶ 0046: The selected subtree root becomes the current node. Processing repeats until the entire URL is processed or a required subtree cannot be identified. ¶ 0047: After complete URL processing, targets associated with the terminal node are returned as matches. It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the organization and lookup of the request path tree of TREADWAY-SONAWANE-SAVENKOV to incorporate the domain values subtree selection implemented with progressive hashing taught by BURCKART. One of ordinary skill in the art would be motivated to integrate that subtree selection into TREADWAY-SONAWANT-SAVENKOV, with a reasonable expectation of success, in order to reduce URL lookup work by selecting the subtree through a hash table that avoids collision handling and scans of candidate buckets, as taught by BURCKART ¶ 0040. Claims 12 Claim 12 is rejected utilizing the rationale for claim 8; the claim is directed to a medium storing instructions corresponding to the method. Claim 20 Claims 20 is rejected utilizing the rationale for claim 8; the claim is directed to a system performing the method. Conclusion The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. See Azaria et al., U.S. PG-Publication No. 2022/0350686 A1, teaching discovering API endpoints from observed request traffic, organizing URL segments into a tree, merging variable parameter nodes, and periodically updating an endpoint profile with normalized URL patterns (¶¶ 0036, 0040-0041, 0047-0054, 0063). Any inquiry concerning this communication or earlier communications from the examiner should be directed to FRANK D MILLS whose telephone number is (571)270-3194. The examiner can normally be reached M-F 9-5:30 CT. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, KEVIN YOUNG can be reached at (571)270-3180. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000. /FRANK D MILLS/Primary Examiner, Art Unit 2194 October 17, 2026
Read full office action

Prosecution Timeline

Nov 20, 2024
Application Filed
Oct 20, 2026
Non-Final Rejection mailed — §101, §103 (current)

Precedent Cases

Applications granted by this same examiner with similar technology

Patent 12801428
APPLICATION PERFORMANCE ENHANCEMENT SYSTEM AND METHOD BASED ON A USER'S MODE OF OPERATION
4y 5m to grant Granted Nov 17, 2026
Patent 12801958
SERVICE FOR MANAGING QUANTUM COMPUTING RESOURCES
3y 2m to grant Granted Nov 17, 2026
Patent 12792384
REBALANCING UNDER-UTILIZED VIRTUAL MACHINES IN HYPERSCALER ENVIRONMENTS
2y 11m to grant Granted Nov 10, 2026
Patent 12770006
METHOD AND USER INTERFACE FOR TRAINING ARTIFICIAL NEURAL NETWORK MODELS IN ENVIRONMENT INCLUDING MULTIPLE COMPUTING NODES
1y 3m to grant Granted Oct 20, 2026
Patent 12743305
METHOD AND APPARATUS FOR COLLABORATIVE TASK PLANNING FOR ARTIFICIAL INTELLIGENCE AGENTS
3y 4m to grant Granted Sep 22, 2026
Study what changed to get past this examiner. Based on 5 most recent grants.

Strategy Recommendation AI-generated — please review before filing

Get a prosecution strategy drawn from examiner precedents, rejection analysis, and claim mapping.
Typically takes 5-10 seconds — AI-generated, attorney review required before filing

Prosecution Projections

1-2
Expected OA Rounds
70%
Grant Probability
92%
With Interview (+22.7%)
3y 4m (~1y 4m remaining)
Median Time to Grant
Low
PTA Risk
Based on 618 resolved cases by this examiner. Grant probability derived from career allowance rate.

Sign in with your work email

Enter your email to receive a magic link. No password needed.

Personal email addresses (Gmail, Yahoo, etc.) are not accepted.

Free tier: 3 strategy analyses per month