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 .
DETAILED ACTION
This Office Action is sent in response to Application’s Communication received on 10/10/2023 for application number 18/378491. The Office hereby acknowledges receipt of the following and placed of record in file: Specification, Drawing, Abstract, Oath/Declaration, and Claims.
Claims (1-18), 19 and 20 are presented for examination.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 10/10/2023 was filed prior to current Office Action. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
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 1-20 are rejected under AIA 35 U.S.C. 103(a) as being unpatentable over of WHANG JOYCE JIYOUNG et al. Foreign Patent Application Publication KR 20240088529 A (hereinafter Whang) in view of Jiao et al. US Patent Application Publication US 20220067030 (hereinafter Jiao).
Regarding claim 1, Whang teaches A method for embedding a knowledge graph, the method being performed by at least one processor and comprising: obtaining a basic knowledge instance and a qualifier instance associated with the basic knowledge instance, wherein at least one of the basic knowledge instance or the qualifier instance comprises a mask element (FIG. 3, PAGE. 3, ¶ 5-6 wherein Wangs constructs a multi-layer knowledge graph including base knowledge and higher-level knowledge in step 310. Base knowledge includes a plurality of knowledge instances implemented from a plurality of entities, and in these knowledge instances, relationships between the plurality of entities may be expressed as a plurality of triplets. Each of the triplets consists of two entities and a relation representing the relationship between the two entities, and can be expressed as {head entity, relation, tail entity}. Higher-order knowledge includes a plurality of knowledge instances implemented from a plurality of triplets of base knowledge, and in these knowledge instances, the relationship between the plurality of triplets may be expressed by at least one upper-order triplet. Each of the upper triplets consists of two triplets and a higher relation representing the relationship between the two triplets, and can be expressed as [head triplet, upper relation, tail triplet)
Whang does not teach generating, through an embedder, a first instance embedding corresponding to the basic knowledge instance and a second instance embedding corresponding to the qualifier instance; and performing a prediction of the mask element by inputting, to a predictor, an instance embedding associated with the mask element, wherein the instance embedding corresponds to the first instance embedding or the second instance embedding.
However in analogous art of knowledge graph embedding, Jiao teaches generating, through an embedder, a first instance embedding corresponding to the basic knowledge instance and a second instance embedding corresponding to the qualifier instance (FIGS. 2-4, [0041-0050] wherein Jiao predict a link which is the target node, refer to as a link prediction or missing link or node or link missing from an incomplete triplet by for knowledge graph, wherein Jiao incorporates a transformer encoder to detecting the link using knowledge graph embeddings that includes multiple entity embeddings and the graph relationship and the relation type embeddings including the link) and performing a prediction of the mask element by inputting, to a predictor, an instance embedding associated with the mask element, wherein the instance embedding corresponds to the first instance embedding or the second instance embedding ([0006], [0028], [0041-0063] wherein Jiao discloses operations for detecting the link prediction by ranking and comparing the link predictions throughout the instances)
It would have been obvious to a person in the ordinary skill in the art before the effective filing date of the claimed invention to combine Jiao with Whang and Kurniawan by incorporating the method of generating, through an embedder, a first instance embedding corresponding to the basic knowledge instance and a second instance embedding corresponding to the qualifier instance; and performing a prediction of the mask element by inputting, to a predictor, an instance embedding associated with the mask element, wherein the instance embedding corresponds to the first instance embedding or the second instance embedding of Jiao into the method of obtaining a basic knowledge instance and a qualifier instance associated with the basic knowledge instance, wherein at least one of the basic knowledge instance or the qualifier instance comprises a mask element of Whang for the purpose of using vector representations or intermediate embeddings to enhance the transformer (Jiao: [0045]).
Regarding claim 2, Whang as modified by Jiao teach updating the embedder and the predictor, based on a difference between a result of the prediction and a correct answer about the mask element. (FIGS,2-4. 6, [0028], [0041-0050], [0063] wherein incorporates a loss function that reflect the chance that correct answer is predicted incorrectly).
Regarding claim 3, Whang as modified by Jiao teach perform an analysis of an association between a first embedding of the basic knowledge instance and a second embedding of the qualifier instance; and generate the first instance embedding and the second instance embedding based on a result of the analysis of the association between the first embedding of the basic knowledge instance and the second embedding of the qualifier instance (FIGS. 2-4, [0041-0050] wherein Jiao predict a link which is the target node, refer to as a link prediction or missing link or node or link missing from an incomplete triplet by for knowledge graph, wherein Jiao incorporates a transformer encoder to detecting the link using knowledge graph embeddings that includes multiple entity embeddings and the graph relationship and the relation type embeddings including the link), ([0006], [0028], [0041-0063] wherein Jiao discloses operations for detecting the link prediction by ranking and comparing the link predictions throughout the instances).
Regarding claim 4, Whang as modified by Jiao teach wherein the context encoder is implemented based on a self-attention module (Abstract, [0025], [0029-0032] wherein Jiao it applies a self-attention mechanism which directly models relationships).
Regarding claim 5, Whang as modified by Jiao teach wherein the generating of the first instance embedding and the second instance embedding comprises: reflecting type information for distinguishing the basic knowledge instance and the qualifier instance from each other in the first embedding and the second embedding; and analyzing the association between the first embedding and the second embedding in which the type information is reflected through the context encoder (FIG. 3, [0045] wherein Jiao learns interactions between an entity and its associated relation type and encodes information from relationship context).
Regarding claim 6, Whang as modified by Jiao teach wherein the embedder further comprises: a basic aggregator configured to aggregate first element embeddings of the basic knowledge instance to generate the first embedding; and an auxiliary aggregator configured to aggregate second element embeddings of the qualifier instance to generate the second embedding (Claim 4 text, [0042-0063], [0075] wherein Jiao outputs link predictions further comprises: outputting a token for each link prediction, wherein the token comprises an aggregation of the source embedding and the predicate embedding; and using the outputted token to determine the plausibility score for the link prediction).
Regarding claim 7, Whang as modified by Jiao teach wherein the qualifier instance is a first qualifier instance, and wherein the generating of the first instance embedding and the second instance embedding comprises: obtaining a second qualifier instance associated with the basic knowledge instance; performing, through the embedder, an analysis of an association among embeddings of the basic knowledge instance, the first qualifier instance, and the second qualifier instance; and generating the first instance embedding and the second instance embedding based on a result of the analysis of the association (FIG. 3, PAGE. 3, ¶ 5-6 wherein Wangs constructs a multi-layer knowledge graph including base knowledge and higher-level knowledge in step 310. Base knowledge includes a plurality of knowledge instances implemented from a plurality of entities, and in these knowledge instances, relationships between the plurality of entities may be expressed as a plurality of triplets. Each of the triplets consists of two entities and a relation representing the relationship between the two entities, and can be expressed as {head entity, relation, tail entity}. Higher-order knowledge includes a plurality of knowledge instances implemented from a plurality of triplets of base knowledge, and in these knowledge instances, the relationship between the plurality of triplets may be expressed by at least one upper-order triplet. Each of the upper triplets consists of two triplets and a higher relation representing the relationship between the two triplets, and can be expressed as [head triplet, upper relation, tail triplet), (FIG. 3, [0045] wherein Jiao learns interactions between an entity and its associated relation type and encodes information from relationship context), FIGS. 2-4, [0041-0050] wherein Jiao predict a link which is the target node, refer to as a link prediction or missing link or node or link missing from an incomplete triplet by for knowledge graph, wherein Jiao incorporates a transformer encoder to detecting the link using knowledge graph embeddings that includes multiple entity embeddings and the graph relationship and the relation type embeddings including the link), ([0006], [0028], [0041-0063] wherein Jiao discloses operations for detecting the link prediction by ranking and comparing the link predictions throughout the instances).
Regarding claim 8, Whang as modified by Jiao teach wherein the predicting of the mask element comprises inputting, to the predictor, element embeddings of a knowledge instance comprising the mask element (Abstract, wherein Jiao discloses a hierarchical Transformer model learns entity embeddings in knowledge graphs. The model consists of two different Transformer blocks where the bottom block generates relation-dependent embeddings for the source entity and its neighbors, and the top block aggregates the outputs from the bottom block to produce the target entity embedding. To balance the information from contextual entities and the source entity itself, a masked entity model (MEM) task is combined with a link prediction task in model training).
Regarding claim 9, Whang as modified by Jiao teach reflecting, in the element embeddings, information for distinguishing types of elements from each other, and inputting, to the predictor, the element embeddings in which the information is reflected (FIG. 3, [0048] wherein Jiao aggregates contextual information together with the information from the source entity embedding and the predicate embedding by using structural features extracted from the output vector representations/intermediate embeddings of the Transformer block. The intermediate embeddings are input into Transformer in the block to create Target embedding which can be used for link prediction., any number of Transformer encoders may be used as shown by N.sub.E 438 and N.sub.C 444. in FIG. 3, it is preferable to use a multiple (e.g., multilayer) Transformer encoders to receive more accurate link prediction. Multiple encoders provide more interaction information to improve accuracy of the link prediction).
Regarding claim 10, Whang as modified by Jiao teach wherein type information of the knowledge instance corresponding to the instance embedding is reflected in the instance embedding associated with the mask element, and wherein the instance embedding in which the type information is reflected is input to the predictor ([0059] wherein Jiao performs positional encoding type encoding, the type encoding encodes each embedding as a source node or as a predicate relationship).
Regarding claim 11, Whang as modified by Jiao teach wherein the predictor comprises a context reflector and a prediction layer, and wherein the predicting of the mask element comprises: generating a plurality of embeddings by reflecting, through the context reflector, the instance embedding associated with the mask element in the element embeddings; and predicting the mask element by inputting, to the prediction layer, an embedding corresponding to the mask element among the plurality of embeddings (FIG. 3, [0048] wherein Jiao aggregates contextual information together with the information from the source entity embedding and the predicate embedding by using structural features extracted from the output vector representations/intermediate embeddings of the Transformer block. The intermediate embeddings are input into Transformer in the block to create Target embedding which can be used for link prediction., any number of Transformer encoders may be used as shown by N.sub.E 438 and N.sub.C 444. in FIG. 3, it is preferable to use a multiple (e.g., multilayer) Transformer encoders to receive more accurate link prediction. Multiple encoders provide more interaction information to improve accuracy of the link prediction)
Regarding claim 12, Whang as modified by Jiao teach wherein the context reflector is implemented based on a self-attention module (Abstract, [0025], [0029-0032] wherein Jiao it applies a self-attention mechanism which directly models relationships).
Regarding claim 13, Whang as modified by Jiao teach wherein the mask element is a discrete entity, and wherein the predictor comprises a classification layer configured to output a predicted probability distribution for predefined entities ([0051], [0056] wherein Jiao discloses a masking strategy that is applied to the source entity embedding of each training example as follows. During training, a proportion of training examples are randomly selected in a batch. With certain probabilities, the input source entity is replaced with a special mask token, a random chosen entity, or left unchanged. The purpose of these perturbations is to introduce extra noise to the information from the source entity embedding, thus requiring the model to learn contextual representations. The probability of each category (masked, random, and unchanged) is a dataset-specific hyper-parameter: for example, the source entity embedding can be masked out more frequently if its graph neighborhood is denser (in which case, the source entity embedding can be easily replaced by the additional contextual information).
Regarding claim 14, Whang as modified by Jiao teach wherein the mask element is a relation, and wherein the predictor comprises a classification layer configured to output a predicted probability distribution for predefined relations ([0051], [0056] wherein Jiao discloses a masking strategy that is applied to the source entity embedding of each training example as follows. During training, a proportion of training examples are randomly selected in a batch. With certain probabilities, the input source entity is replaced with a special mask token, a random chosen entity, or left unchanged. The purpose of these perturbations is to introduce extra noise to the information from the source entity embedding, thus requiring the model to learn contextual representations. The probability of each category (masked, random, and unchanged) is a dataset-specific hyper-parameter: for example, the source entity embedding can be masked out more frequently if its graph neighborhood is denser (in which case, the source entity embedding can be easily replaced by the additional contextual information).
Regarding claim 15, Whang as modified by Jiao teach wherein the mask element is a numeric entity, and wherein the predictor comprises a regression layer configured to output a predicted numeric value (page. 6, ¶ 3 wherein Whang calculates calculate a loss function for higher level knowledge. At this time, wherein the processor expresses each head entity, relation, and tail entity of the triplets of the base knowledge as expression vectors, and then combines the expression vectors using an appropriate aggregator to express the triplets. Each can be expressed as one expression vector. For example, the aggregator can apply linear transformation after combining vectors. Accordingly, the processor can calculate the loss function by expressing the head triplet, upper relation, and tail triplet of each upper triplet as expression vectors, and then calculating the score of the upper triplet using a scoring function. For example, the loss function for upper-level knowledge can be defined as an equation. That is, the loss function for top knowledge is based on the score of the top triplet and the scores of contaminated top triplets generated from the top triplet of top knowledge, and the contaminated top triplets are the head triplet from each of the top triplets. Alternatively, it can be generated by replacing the tail triplet with a random triplet from the base knowledge. Accordingly, the processor 140 can learn the knowledge graph embedding model in a way that increases the score of the top triplet and simultaneously lowers the score of the contaminated top triplet), ([0038] wherein Jiao incorporates multi-attention head layers in both the encoder and decoder, there are pointwise feed-forward layers , which may have identical parameters for each position, and which can be described as a separate, identical linear transformation of each element from the given sequence).
Regarding claim 16, Whang as modified by Jiao teach wherein the predictor comprises a first predictor associated with a first prediction task and a second predictor associated with a second prediction task, and wherein, based on at least one of a type of the mask element, a format of a value of the mask element, and a type of a knowledge instance comprising the mask element, the first prediction task and the second prediction task are distinguished from each other ([0050], [0077], [0085] wherein Jiao supplies contextual information to the model during training might cause problems. On the one hand, since a source entity often contains particular information for link prediction, the model may learn to ignore the additional contextual information, which could also be noisy. On the other hand, the introduction of rich contextual information could in turn downgrade information from the source entity and cause potential over-fitting problems. To solve these problems, a Masked Entity Prediction (MEP) task is used to balance the process of contextualization during training)
Regarding claim 17, Whang as modified by Jiao teach wherein the first predictor and the second predictor are configured to share at least one weight parameter with each other ([0035], [0052], [0057] wherein Jiao describes sharing the weight matrix).
Regarding claim 18, Whang as modified by Jiao teach generating a complete knowledge instance based on a result of the prediction in a knowledge instance comprising the mask element; and inserting the generated complete knowledge instance into the knowledge graph (Abstract, wherein Whang proposes a method capable of augmenting knowledge not included in the existing knowledge graph to induce more effective learning of the knowledge graph embedding models. A method for embedding a multi-layer knowledge graph performed in a computer system comprises the steps of: configuring a multi-layer knowledge graph including base knowledge represented by a plurality of triples and higher-level knowledge in which a relationship between the plurality of triples is represented by at least one higher triple; and learning a knowledge graph embedding model by using the base knowledge and the higher-level knowledge).
Regarding claim 19, the claim is similar in scope to claim 1 therefore the claims are rejected under similar rationale.
Regarding claim 20, the claim is similar in scope to claim 1 therefore the claims are rejected under similar rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HASSAN MRABI whose telephone number is (571)272-8875. The examiner can normally be reached on Monday-Friday, 7:30am-5pm. Alt, Friday, EST.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached on 571-270-5871. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/HASSAN MRABI/Examiner, Art Unit 2144
Prosecution Insights