DETAILED ACTION
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 .
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Examiner Notes
(1) In the case of amending the Claimed invention, Applicant is respectfully requested to indicate the portion(s) of the specification which dictate(s) the structure relied on for proper interpretation and also to verify and ascertain the metes and bounds of the claimed invention. This will assist in expediting compact prosecution. MPEP 714.02 recites: “Applicant should also specifically point out the support for any amendments made to the disclosure. See MPEP § 2163.06. An amendment which does not comply with the provisions of 37 CFR 1.121 (b), (c), (d), and (h) may be held not fully responsive. See MPEP § 714.” Amendments not pointing to specific support in the disclosure may be deemed as not complying with provisions of 37 C.F.R. 1.131 (b), (c), (d), and (h) and therefore held not fully responsive. Generic statements such as "Applicants believe no new matter has been introduced" may be deemed insufficient.
(2) Examiner cites particular columns, paragraphs, figures and line numbers in the references as applied to the claims below for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested that, in preparing responses, the applicant fully consider the references in their entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the Examiner.
Remarks
Receipt of Applicant’s Amendment file on 10/31/2025 is acknowledged.
Response to Arguments
Regarding 35 U.S.C. 101 rejection, applicant's arguments filed 10/31/2025 have been fully considered but they are not persuasive.
Applicant argues that ““building, by the processing device, a hash table by hashing, using a hash function, permutations of correctly spelled tokens” and “searching the hash table to match the variations of the misspelled token with the permutations of the correctly spelled tokens,” as recited in amended claim 1, cannot practically be performed in the human mind, or by a human using pen and paper.” (last paragraph, page 12). Applicant further argues “Indeed, Applicant’s specification explains that “the permutation index is a hash table in which the permutations of the respective token are hashed using a hashing function” and that “a search time to search for permutations in permutation index is independent of the number of permutations in the permutation index” (Applicant's Specification, [0035] and [0046]). A human mind cannot practically implement a hash function to hash permutations of tokens into a hash table structure, nor can it perform hash table lookups that provide constant-time search performance independent of the data size. The technical implementation of hashing algorithms and hash table data structures involves computational processes that cannot be performed mentally by a human” (last paragraph , page 12). Respectfully, it is noted that one can mentally and/or with aid of pen/paper to creating a list/table of variation of corrected spell word/phrase/token by simply make a slight change of corrected spelled word/token with editing to one of the character of the corrected spelled word/token. To any extent as well that ‘hashed’ is a math algorithm applied to a function – essentially building the permutations is a mental process and hashing them is also mentally performable or math. Noted there are neither additional details of the hashing algorithm, nor require any number of permutations. Thus, this can be done mentally given those simple parameters under the Broadest Reasonable Interpretation.
Applicant also argues “Applicant submits that the recited features of claim 1 improve the functioning of computing devices within the technical field of spell correction for in-application searches. Indeed, Applicant’s Specification notes that “conventional spell correction techniques implemented for in-application searches often lead to user frustration due to slow search (i.e. high latency times)” (Applicant's Specification, [0001]). The specification further explains that “conventional spell correction techniques typically utilize end-to-end neural models to correct spelling errors” which “are slower than rule-based spellchecking techniques” (Applicant's Specification, [0014]). This highlights the technical problem of computational inefficiency in existing spell correction systems.” (page 13, first paragraph). Respectfully, it is noted that Applicant argues the claims provide a technical solution by “utilize a hybrid spellchecking architecture…”(page 13) but any speed improvement would occur in the mental process or by hand, as well as on the computer. That is not an improvement to the computer or technology, rather is an improvement applying to the abstract idea.
Applicant further argues “the claimed solution implements a specific hash table architecture where “the permutation index is a hash table in which the permutations of the respective token are hashed using a hashing function” and provides the technical benefit that “a search time to search for permutations in permutation index is independent of the number of permutations in the permutation index” (Applicant's Specification, [0035] and [0046]). Thus, by “building, by the processing device, a hash table by hashing, using a hash function, permutations of correctly spelled tokens” and “searching the hash table to match the variations of the misspelled token with the permutations of the correctly spelled tokens,” as recited in claim 1, the claimed solution achieves constant-time lookup performance regardless of the size of the collection of tokens” (page 15) but the claims have no such structure of description of the “hash table” at all; merely use hashing, which is a math concept, and/or storing that representation of the data.
The rejections of claims 12 and 17 are maintained for similar reasons as shown above.
Regarding to 35 U.S.C. 103 rejection, applicant’s arguments with respect to claims 1, 12 and 17 have been considered but are moot in view of the new ground(s) of rejection (See new reference of Gupta) (Explanation provided below).
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.
When considering subject matter eligibility under 35 USC 101, it must be determined whether the claim is directed to one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter (Step 1). If the claim does fall within one of the statutory categories, it must then be determined whether the claim is directed to a judicial exception (i.e., law of nature, natural phenomenon, and abstract idea) (Step 2A), and if so, it must additionally be determined whether the claim is a patent-eligible application of the exception. If an abstract idea is present in the claim, any element or combination of elements in the claim must be sufficient to ensure that the claim amounts to significantly more than the abstract idea itself (Step 2B). Examples of abstract ideas include fundamental economic practices; certain methods of organizing human activities; an idea itself; and mathematical relationships/formulas.
Analysis
At Step 1: The claim is directed to a "method" and thus directed to a statutory category.
At Step 2A, Prong One: The claim recites the following limitations directed to an abstract idea:
"identifying ... a misspelled token in the user query" as drafted recites a mental process as form of evaluation or judgement. One can mentally analyze a user query and determine a token/word is misspelled, for instance by comparison to mentally held correct spellings.
“building, by the processing device, a hash table by hashing, using a hash function, permutations of correctly spelled tokens, the permutations of a correctly spelled token including one-delete variations of the correctly spelled token having the correctly spelled token prepended" as drafted recites a mental process with an aid of pen/paper as form of evaluation or judgement. One can mentally create a list/table contained variation of corrected spelled word/phrase/token by delete one character from the corrected spelled token/word/phrase.
“generating,…, variations of the misspelled token by performing edit…” simply illustrate a mental process of changing the misspelled token to another token by remove/delete one character from the misspelled token. For instance, consistent with the specification in [0045]-[00467] for a misspelled token "tedt" one can mentally generate variations such as "edt" "tdt" "tet" and "ted".
identifying ... candidate tokens to replace the misspelled token from a collection of correctly spelled tokens by searching the hash table to matching the variations of the misspelled token with the permutations of the correctly spelled tokens" as drafted recites a mental process as form of evaluation or judgement. One can mentally identify candidate correct spellings by variations of the misspelled token and matching to known correct spellings using comparison method with the simplified list of variations of corrected spelled token using pen/paper. For instance, consistent with the specification in [0045]-[00467] for a misspelled token "tedt" one can mentally generate variations such as "edt" "tdt" "tet" and "ted" and match this with permutations of correct spellings such as "tent" and "test" since they both have a misspelling “tet.” Also noted that searching a hash table can be done mentally, that is one can compare hashes for matches as a form of evaluation or judgement that is practically performed mentally.
"generating, by a... model ... a ranking of the candidate tokens, the ... model trained on a training dataset including user queries entered via the search feature of the application having injected errors” as drafted recites a mental process as a form of evaluation or judgement. One can mentally rank candidate tokens, including by using a mental model that is trained (or based on) entered queries with synthetic errors. This limitation broadly recites "ranking" and “training” with no specifics required as to how this is performed. Consistent with the specification at [0068] such training dataset is one that includes common spelling error types such as "remove or add a vowel.” Accordingly, this recites a mental process as mentally ranking based on a model of common spelling errors.
"selecting ... a token from the candidate tokens based on the ranking" as drafted recites a mental process as a form of evaluation or judgement. One can mentally select or judge a best token based on ranking. Consistent with the specification at [0020] this can be selecting the "highest-ranked" candidate token and one can mentally make such an evaluation or judgement of the "highest-ranked.”
At Step 2A, Prong Two: The claim recites the following additional elements:
That the method steps are performed "by a processing device" which is a high level recitation of a generic computer component and represents mere instructions to apply on a computer as in MPEP 2106.05(f), which does not provide integration into a practical application.
"receiving ... a user query entered via a search feature of an application" recites insignificant extra-solution activity as mere data gathering such as ‘obtaining information’ as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
That the model for ranking is determined "by a machine learning" model is no more than generally linking the abstract idea to the particular field of use or technological environment of machine learning. (see MPEP 2106.05(h). As reflected in the claim, the step is akin to using machine learning as a mere tool (2106.05(f)). No specific type of machine learning processing or techniques are recited in the claim itself. The claimed ‘training dataset’ is applicable for training any model (even a mental model) and there is no indication of the claim as to how the machine learning model is trained using this data. Rather the limitation is entirely results oriented and merely links the abstract idea to being performed in the technological environment of machine learning. Accordingly, as drafted, this limitation does not provide integration into a practical application.
"outputting ... the selected token" recites insignificant extra-solution activity as mere data outputting as identified in MPEP 2106.05(g) and does not provide integration into a practical application. Adding a final step of outputting a result of the misspelling correction suggestion does not meaningfully limit the abstract idea of identifying the misspelling and selected candidate correction.
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
At Step 2B: The conclusions for the mere implementation using a computer and mere field of user are carried over and do not provide significantly more.
With respect to the "receiving ... a user query entered via a search feature of an application" identified as insignificant extra-solution activity above when re- evaluated this element is well-understood, routine, and conventional as evidenced by the court cases in MPEP 2106.05(d)CD, "i, Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); ... OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network);" and thus remains insignificant extra- solution activity that does not provide significantly more. With respect to the "outputting ... the selected token" identified as insignificant extra -solution activity above when re-evaluated this element is well-understood, routine, and conventional as evidenced by the court cases in MPEP 2106.05(d)UID, "1. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); ... OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network);" and thus remains insignificant extra-solution activity that does not provide significantly more. Note that the claim has no details as to how the outputting is performed, such that it encompasses general data transmission and sending of the information over a network. Looking at the claim as a whole does not change this conclusion. Therefore, the claim is ineligible under U.S.C. 101.
Claims 12 and 17 are rejected for similar reasons as shown in the rejection of claim 1.
Claims 2, 4-6, 8-10, 14-16, 18-19, 21-23 and 25-26 are dependent on their respective parent claims 1, 12 and 17 respectively and include all the limitations of claims 1, 12 and 17. Since these claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception, thus the claims are direct to abstract idea.
Claims 1-2, 4-6, 8-10, 12, 14-19, 21-23 and 25-26 are therefore not drawn to eligible subject matter as they are directed to an abstract idea without significantly more.
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 of this title, 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 1-2, 12, 17-18, 21 and 23 were not rejected under 35 U.S.C.103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al), further in view of Gupta (U.S. Patent No. 10,936,813 B1).
Regarding claim 1, Finch et al teach a method comprising:
receiving, by a processing device, a user query entered via a search feature of an application [“the user of electronic device 115 has entered a partial query 204” [0031]];
identifying, by the processing device, a misspelled token in the user query [“partial query 204 is a misspelled partial query intended to be a portion of a complete search query” [0032]];
identifying, by the processing device, candidate tokens to replace the misspelled token from a collection of correctly spelled tokens [“List 212 may be a ranked list of spelling- corrected completion suggestions” [0033]];
generating, by a machine learning model implemented by the processing device, a ranking of the candidate tokens, the machine learning model trained on a training dataset including user queries entered via the search feature of the application having injected errors [ “a previous user query may include an incorrectly spelled word such as “ecit” and a previously selected spelling correction that may have been provided to the user and selected by the user may include the correctly spelled word “exit” [0048] – [0049] . “List 212 may be a ranked list of spelling-corrected” [0035], “Moreover, intrinsically learned error types such as intrinsically machine-learned keyboard layout based errors and/or morpho-phonetic errors that have been learned by the machine-learning model can be used to rank suggested completions and/or suggested corrections from lookup table 500” [0066]];
selecting, by the processing device, a token from the candidate tokens based on the ranking [“The search query may be a selected one of the spelling-corrected versions of the partial query” [0041]]; and
outputting, by the processing device, the selected token [“the search query provided from electronic device 115 may be provided to search engine 302.” [0042]].
Finch teaches a system for correcting misspelling. However, Finch et al does not teach:
building, by the processing device, a hash table by hashing, using a hash function, permutations of correctly spelled tokens, the permutations of a correctly spelled token including one-delete variations of the correctly spelled token having the correctly spelled token prepended ;generating, by the processing device, variations of the misspelled token by performing edits on individual characters of the misspelled token; and identifying, by the processing device, candidate tokens to replace the misspelled token from a collection of correctly spelled tokens by searching the hash table to match the variations of the misspelled token with the permutations of the correctly spelled tokens.
Hantler teaches another misspelling correction system. Specifically, Hantler teaches: building, by the processing device, a hash table by hashing, using a hash function, permutations of correctly spelled tokens, the permutations of a correctly spelled token including one-delete variations of the correctly spelled token having the correctly spelled token prepended (For the candidate word WXYZ, … The deletions would be XYZ, WYZ, WXZ, and WXY [0023]Deletion (d) hash: OAT, CAT, COT, COA…. Referring again to FIG. 3, with the above hash tables pre-created, a test is performed for a distance one misspelling (as discussed above in conjunction with FIG. 1); also see [0021]);
generating, by the processing device, variations of the misspelled token by performing edits on individual characters of the misspelled token [“For the candidate word WXYZ, … The deletions would be XYZ, WYZ, WXZ, and WXY [0023]];
identifying, by the processing device, candidate tokens to replace the misspelled token from a collection of correctly spelled tokens by searching the hash table to match the variations of the misspelled token with the permutations of the correctly spelled tokens [“The contents of each hash are again illustrated by considering the sample dictionary word COAT… Deletion (d) hash: OAT, CAT, COT, COA…. Referring again to FIG. 3, with the above hash tables pre-created, a test is performed for a distance one misspelling (as discussed above in conjunction with FIG. 1). [0025] – [0034] “The spelling of at least one candidate word is corrected by obtaining at least one variant dictionary hash table based on variants of a set of known correctly spelled words, wherein the variants are obtained by applying … a deletion … on the correctly spelled words [abstract]].
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to combine the teachings of Finch et al and Hantler et al because they both directed to a system for correcting misspelling and Hantler et al teach a specific way to generate the candidate tokens to improve the operation of Finch et al system with a predictable result.
Finch as modified by Hantler do not explicitly disclose: the edits including deletes and not inserts, transposes, and replaces.
Gupta (U.S. Patent No. 10,936,813 B1) teaches: the edits including deletes and not inserts, transposes, and replaces (col. 7, line 37-50, the spell checker component 102 may utilize a symmetric delete algorithm (SDA). In some instances, the SDA may reduce the complexity of candidate suggestion by only using deletions instead of deletions, transpositions, replacements, and insertions. Using SDA may result in a significance performance gain, is orders of magnitude faster than other string-search algorithms, and is language independent (i.e., the alphabet is not required for deleting characters).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include the edits including deletes and not inserts, transposes, and replaces into spelling correction of Finch.
Motivation to do so would be to include the edits including deletes and not inserts, transposes, and replaces that may result in a significant performance gain (Gupta, col. 7, line 48-49).
Regarding claim 2, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, further teach wherein the collection of tokens includes: tokens corresponding to features of the application; tokens having at least a threshold frequency of occurrence in the user queries entered via the search feature of the application; or tokens of multiple languages (Finch, paragraph [0035], list 212 may be a ranked list of spelling-corrected and/or completion suggestions; the ranking shown in list 212 may be a natural output of a machine learning model that has been trained using training data in which the website corresponding to term “mediasite” is often searched for or selected in spelling corrections or completions; also see paragraph [0066], the trained machine-learning model includes parameters (weights such as weights for connections between units of the deep neural network, biases, threshold)…).
As per claims 12 and 17, these claims are rejected on grounds corresponding to the same rationales given above for rejected claim 1 and are similarly rejected.
As per claim 18, this claim is rejected on grounds corresponding to the same rationales given above for rejected claim 2 and is similarly rejected.
Regarding claim 21, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, further teach: wherein the one-delete variation corresponds to the respective token having one character deleted from the respective token (Hantler, teaches “For the candidate word WXYZ, … The deletions would be XYZ, WYZ, WXZ, and WXY [0023]).
Regarding claim 23, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 12, further teach: the operations further comprising generating the training dataset including user queries entered via the search feature of the application, and error queries by injecting errors of different types at different frequencies into the user queries (Finch, paragraph [0022], the machine-learning model is trained using a training data set of past queries (e.g., correctly and incorrectly spell partial queries) and user selected or user-engage for each past partial query; also see [0048] – [0049], “a previous user query may include an incorrectly spelled word such as “ecit” and a previously selected spelling correction that may have been provided to the user and selected by the user may include the correctly spelled word “exit”; paragraph [0035], list 212 may be a ranked list of spelling-corrected and/or completion suggestions; the ranking shown in list 212 may be a natural output of a machine learning model that has been trained using training data in which the website corresponding to term “mediasite” is often searched for or selected in spelling corrections or completions; also see paragraph [0066], the trained machine-learning model includes parameters (weights such as weights for connections between units of the deep neural network, biases, threshold)… “Moreover, intrinsically learned error types such as intrinsically machine-learned keyboard layout based errors and/or morpho-phonetic errors that have been learned by the machine-learning model can be used to rank suggested completions and/or suggested corrections from lookup table 500”).
Claims 4-6, and 14-15 are rejected under 35 U.S.C. 103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al) and Gupta (U.S. Patent No. 10,936,813 B1), further in view of Song et al. (U.S. Patent No. 9,081,821 B2).
Regarding claim 4, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, but do not explicitly disclose: wherein the generating of variations includes generating variations of the misspelled token that are less than a threshold number of edit distances from the misspelled token, and the identifying the candidate tokens includes searching the hash table to match the permutations of the correctly spelled tokens with the variations of the misspelled token.
Song teaches: wherein the generating of variations includes generating variations of the misspelled token that are less than a threshold number of edit distances from the misspelled token, and the identifying the candidate tokens includes searching the hash table to match the permutations of the correctly spelled tokens with the variations of the misspelled token (Song, col. 8, line 6-33, set of tokens can be retrieved from the DAWG storage; determining how much the characters from tok(i) need to be changed in order to arrive at each of the alternative tokens (for example, tok {i’}); also see col. 10, line 8-23, if a misspelling is not suspected, the process may end; if one is suspected, the tokens are compared against spell storage table contain alternative tokens; a spell cursor is launched to calculate scores for permutations of tokens and alternative tokens…; “high” may be considered to have less of an edit distance from “hi” than the edit distance between “Hu” and “hi”, noted, “high” is interpreted corrected spelled tokens; also see col. 7, line 6-24; in combination with the utilized of hash table taught by Hantler as follow: “The contents of each hash are again illustrated by considering the sample dictionary word COAT… Deletion (d) hash: OAT, CAT, COT, COA…. Referring again to FIG. 3, with the above hash tables pre-created, a test is performed for a distance one misspelling (as discussed above in conjunction with FIG. 1). [0025] – [0034] “The spelling of at least one candidate word is corrected by obtaining at least one variant dictionary hash table based on variants of a set of known correctly spelled words, wherein the variants are obtained by applying … a deletion … on the correctly spelled words [abstract], it reads on as claimed).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include wherein the identifying the candidate tokens includes generating variations of the misspelled token that are less than a threshold number of edit distances from the misspelled token, and identifying the correctly spelled tokens in the collection of tokens by matching permutations of the correctly spelled tokens with the variations of the misspelled token into spelling correction of Finch.
Motivation to do so would be to include wherein the identifying the candidate tokens includes generating variations of the misspelled token that are less than a threshold number of edit distances from the misspelled token, and identifying the correctly spelled tokens in the collection of tokens by matching permutations of the correctly spelled tokens with the variations of the misspelled token to address issue that there are many situations where the site is unable to match the query to items, which can commonly cause by spelling errors either through typographical errors, lack of knowledge of proper spelling of terms and/or lack of knowledge about proper formatting for particular terms (Song, col. 1, line 23-28).
Regarding claim 5, Finch as modified by Hantler, Gupta and Song teach all claimed limitations as set forth in rejection of claim 4, further teach wherein the generating the variations includes generating additional variations of the misspelled token that are greater than or equal to the threshold number of edit distances from the misspelled token based on less than a threshold number of the correctly spelled tokens being identified from the variations (Song, col. 8, line 6-33, set of tokens can be retrieved from the DAWG storage; determining how much the characters from tok(i) need to be changed in order to arrive at each of the alternative tokens (for example, tok {i’}); also see col. 10, line 8-23, if a misspelling is not suspected, the process may end; if one is suspected, the tokens are compared against spell storage table contain alternative tokens; a spell cursor is launched to calculate scores for permutations of tokens and alternative tokens…; “high” may be considered to have less of an edit distance from “hi” than the edit distance between “Hu” and “hi”, noted, “high” is interpreted corrected spelled tokens; also see col. 7, line 6-24; in combining with the generating variations taught by Hantler, it reads on as claimed).
Regarding claim 6, Finch as modified by Hantler, Gupta and Song teach all claimed limitations as set forth in rejection of claim 5, further teach wherein the identifying the candidate tokens includes searching the hash table to match the permutations of the correctly spelled tokens with the second variations of the misspelled token (Hantler teaches “The contents of each hash are again illustrated by considering the sample dictionary word COAT… Deletion (d) hash: OAT, CAT, COT, COA…. Referring again to FIG. 3, with the above hash tables pre-created, a test is performed for a distance one misspelling (as discussed above in conjunction with FIG. 1). [0025] – [0034] “The spelling of at least one candidate word is corrected by obtaining at least one variant dictionary hash table based on variants of a set of known correctly spelled words, wherein the variants are obtained by applying … a deletion … on the correctly spelled words [abstract] while Song, col. 8, line 6-33, teaches set of tokens can be retrieved from the DAWG storage; determining how much the characters from tok(i) need to be changed in order to arrive at each of the alternative tokens (for example, tok {i’}); also see col. 10, line 8-23, if a misspelling is not suspected, the process may end; if one is suspected, the tokens are compared against spell storage table contain alternative tokens; a spell cursor is launched to calculate scores for permutations of tokens and alternative tokens…; “high” may be considered to have less of an edit distance from “hi” than the edit distance between “Hu” and “hi”, noted, “high” is interpreted corrected spelled tokens; also see col. 7, line 6-24).
As per claims 14-15, these claims are rejected on grounds corresponding to the same rationales given above for rejected claims 4-5 respectively and are similarly rejected.
claims 8, 19 and 26 are rejected under 35 U.S.C.103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al) and Gupta (U.S. Patent No. 10,936,813 B1), further in view of Nayak et al. (U.S. Patent No. 9,317,606 B1).
Regarding claim 8, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, but do not explicitly disclose: wherein the ranking is further based on: a quantity of search results of the application that the candidate tokens produce; a click rate associated with the search results of the application that the candidate tokens produce; and linguistic features associated with the candidate tokens.
Nayak teaches: wherein the ranking is further based on: a quantity of search results of the application that the candidate tokens produce; a click rate associated with the search results of the application that the candidate tokens produce; and linguistic features associated with the candidate tokens (Nayak, col. 5, line 58-67, col. 6, line 1-5, generating candidate corrections using rule-based methods (such as based on inflections or based on commonly misspelled words) or similarity key techniques; the candidate correction can be ranked based on statistical probability such as transition probability or confusion probability; the words and scores determined by language model and the character-gram model can be returned to the spell check and spell correct module were the spell correct system can analyze words having an assigned score higher than a predetermined threshold in order to determine suggested correct spelling words; also see col. 6, line 14-19, models that look at the probability that a word would appear in a query based on the frequency of the word found in a collection and the statistical probability of the sequence of characters occurring in the language…).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include wherein the ranking is further based on: a quantity of search results of the application that the candidate tokens produce; a click rate associated with the search results of the application that the candidate tokens produce; and linguistic features associated with the candidate tokens into spelling correction of Finch.
Motivation to do so would be to include wherein the ranking is further based on: a quantity of search results of the application that the candidate tokens produce; a click rate associated with the search results of the application that the candidate tokens produce; and linguistic features associated with the candidate tokens for the language model term score, and will provide to the spell correct system a predetermined number N of search terms (Nayak, col. 1, line 50-60).
As per claim 19, this claim is rejected on grounds corresponding to the same rationales given above for rejected claim 8 and is similarly rejected.
As per claim 26, this claim is rejected on grounds corresponding to the same rationales given above for rejected claim 8 and is similarly rejected.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al) and Gupta (U.S. Patent No. 10,936,813 B1), further in view of Lai (U.S. Patent No. 10,409,803 B1).
Regarding claim 9, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, but do not explicitly disclose: receiving, by the processing device, application usage data including a plurality of user queries entered via the search feature of the application; and updating, by the processing device, the collection of tokens to include additional tokens from the plurality of user queries.
Lai teaches: receiving, by the processing device, application usage data including a plurality of user queries entered via the search feature of the application; and updating, by the processing device, the collection of tokens to include additional tokens from the plurality of user queries (Fig. 4, col. 13, line 8-21, receiving name search logs…; these domain name search logs may include interchangeable search tokens/terms identified from any combination of stored consecutive search session queries… server may then update the data associated in database with word pair data to reflect that the identified misspell either a typo or a phonetic spin).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include receiving, by the processing device, application usage data including a plurality of user queries entered via the search feature of the application; and updating, by the processing device, the collection of tokens to include additional tokens from the plurality of user queries into spelling correction of Finch.
Motivation to do so would be to include receiving, by the processing device, application usage data including a plurality of user queries entered via the search feature of the application; and updating, by the processing device, the collection of tokens to include additional tokens from the plurality of user queries for identifying misspelled tokens and acronyms and recommending domain names according to corrected misspellings, related acronyms or full forms of the acronyms, and alternative domain names for unigram-based domain name queries (Lai, col. 1, line 7-11).
Claims 10, 22 and 25 are rejected under 35 U.S.C. 103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al) and Gupta (U.S. Patent No. 10,936,813 B1), further in view of Yee et al. (U.S. Pub. No. 2021/0326526 A1).
Regarding claim 10, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 1, but do not explicitly disclose: identifying, by the processing device, a corrected token corresponding to the misspelled token in a database of overrides; bypassing, by the processing device, the identifying the candidate tokens and the generating based on the misspelled token being included in the database; and outputting, by the processing device, the corrected token.
Yee teaches: identifying, by the processing device, a corrected token corresponding to the misspelled token in a database of overrides (paragraph [0056]-[0057], for each entry in the commonly misspelled table 108, the misspelling is on the left of each line; tab delimited on the right is the correct, canonical gene term; the system 100 will check within the commonly misspelled table to determine if the word matches any of the misspellings in this commonly misspelled table, and if so, the correct canonical gene term is written into the output).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include identifying, by the processing device, a corrected token corresponding to the misspelled token in a database of overrides into spelling correction of Finch.
Motivation to do so would be to include identifying, by the processing device, a corrected token corresponding to the misspelled token in a database of overrides for address a method and apparatus for genome spelling correction and acronym standardization (Yee, paragraph [0002]).
Finch as modified by Hantler and Yee further teach: bypassing, by the processing device, the identifying the candidate tokens and the generating based on the misspelled token being included in the database; and outputting, by the processing device, the corrected token (Yee, paragraph [0056]-[0057], for each entry in the commonly misspelled table 108, the misspelling is on the left of each line; tab delimited on the right is the correct, canonical gene term; the system 100 will check within the commonly misspelled table to determine if the word matches any of the misspellings in this commonly misspelled table, and if so, the correct canonical gene term is written into the output).
As per claim 22, this claim is rejected on grounds corresponding to the same rationales given above for rejected claim 10 and is similarly rejected.
Regarding claim 25, Finch as modified by Hantler, Gupta and Yee teach all claimed limitations as set forth in rejection of claim 10, further teach: adding the additional misspelled token to the database of overrides based on based on the additional misspelled token having at least a threshold frequency of occurrence in user queries entered via the search feature of the application (Yee, paragraph [0055]-[0056], the frequency of each spelling variant, including the canonical terms, is also calculated; once the user has determined that the similar term (potential misspelled) is a misspelling of the canonical term, that misspelled is added into commonly misspelled table; in combination with the teaching of Finch, paragraph [0035], list 212 may be a ranked list of spelling-corrected and/or completion suggestions; the ranking shown in list 212 may be a natural output of a machine learning model that has been trained using training data in which the website corresponding to term “mediasite” is often searched for or selected in spelling corrections or completions; also see paragraph [0066], the trained machine-learning model includes parameters (weights such as weights for connections between units of the deep neural network, biases, threshold)…), it reads on as claimed.
Claim 16 is rejected under 35 U.S.C. 103 as being unpatentable over Finch et al. (U.S. PG Pub. No. 2019/0370393 Al) in view of Hantler et al. (U.S. PG Pub. No. 2008/0059876 Al) and Gupta (U.S. Patent No. 10,936,813 B1), further in view of Chandrasekar et al. (U.S. Patent No. 7,296,019 B1).
Regarding claim 16, Finch as modified by Hantler and Gupta teach all claimed limitations as set forth in rejection of claim 12, but do not explicitly disclose: wherein the dataset of user queries includes user queries of multiple languages.
Chandrasekar teaches: wherein the dataset of user queries includes user queries of multiple languages (col. 4, line 1-10, the words used for spell checking and correction may be collected through multiple channels or from multiple sources, including word commonly found on the web, and in users’ queries, as well as words from a standard language lexicon, all of which may be in one or more languages…).
It would have been obvious to one of ordinary skill in art before the effective filing date of the claim invention to include wherein the dataset of user queries includes user queries of multiple languages into spelling correction of Finch.
Motivation to do so would be to include wherein the dataset of user queries includes user queries of multiple languages to address a need for more comprehensive implementation of a dynamic lexicon for spellchecking of Web queries (Chandrasekar, col. 2, line 67, col. 3, line 1).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KEN HOANG whose telephone number is (571)272-8401. The examiner can normally be reached M-F 7:30am-5:00pm.
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, Charles Rones can be reached at (571)272-4085. 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.
/KEN HOANG/Examiner, Art Unit 2168
/CHARLES RONES/Supervisory Patent Examiner, Art Unit 2168