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 .
Status of Claims
Claims 1-20 are pending and examined herein.
Claims 1-20 are objected to.
Claims 1-14 are rejected under 35 U.S.C. 112(b).
Claims 1-20 are rejected under 35 U.S.C. 101.
Claims 1-20 are rejected under 35 U.S.C. 103.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-14 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 1 and 8 recite the limitation “reason text” in the second to last paragraph of each claim. It is unclear whether this limitation refers to the “reason text” in the preamble of the respective claim or if a second reason text is provided. Thus, the claims are rendered indefinite.
Dependent claims 2-7 and 9-14 fail to resolve the issue and are rejected with the same rationale.
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 15-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter.
The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because they claim a product without any structural recitations; the claims are directed to “software per se”. The claimed “computing device” and “computer-readable storage device” are not limited to computer hardware, but rather encompass computer software as well. Paragraph [0059] of the specification also states that the apparatus “can be implemented in a computer program product”, which suggests that the system/apparatus can be implemented as software per se. Examiner recommends that the “computing device” be amended to recite a computer processor.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
MPEP § 2109(III) sets out steps for evaluating whether a claim is drawn to patent-eligible subject
matter. The analysis of claims 1-20, in accordance with these steps, follows.
Step 1 Analysis:
Step 1 is to determine whether the claim is directed to a statutory category (process, machine,
manufacture, or composition of matter. Claims 1-7 are directed to a process, claims 8-14 are directed to an article of manufacture, and claims 15-20 would be directed to a machine if amended. If amended, all claims would be directed to statutory categories and analysis proceeds.
Step 2A Prong One, Step 2A Prong Two, and Step 2B Analysis:
Step 2A Prong One asks if the claim recites a judicial exception (abstract idea, law of nature, or natural phenomenon). If the claim recites a judicial exception, analysis proceeds to Step 2A Prong Two, which asks if the claim recites additional elements that integrate the abstract idea into a practical application. If the claim does not integrate the judicial exception, analysis proceeds to Step 2B, which asks if the claim amounts to significantly more than the judicial exception. If the claim does not amount to significantly more than the judicial exception, the claim is not eligible subject matter under 35 U.S.C. 101.
None of the claims represent an improvement to technology.
Regarding claim 1, the following are abstract ideas:
generating user-specific recommendations with reason text (Generating user-specific recommendations with reason text can be practically performed in the human mind. This is a mental process.)
aggregating user data and event data, the user data representative of a user that as an addressee of the communication, the event data representative of an event described in the communication; (Aggregating data can be practically performed in the human mind. This is a mental process.)
determining, from a graph neural network (GNN), a sub-GNN that is specific to the user, the GNN comprising a data structure that represents users of an enterprise and relationships between users, the sub-GNN representing a portion of the GNN; (Determining a sub-GNN, as supported by [0034] of the specification, is determining a sub-graph of the graph comprised in the GNN. Therefore, determining a sub-GNN (sub-graph) can be practically performed in the human mind. This is a mental process.)
generating a recommendation regarding the event using the updated sub-GNN, the recommendation being specific to the user; (Generating a recommendation using the graph of the sub-GNN can be practically performed in the human mind. This is a mental process.)
The following claim elements are additional elements which, taken alone or in combination with the other additional elements, do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception:
A computer-implemented method for …, the method being executed by one or more processors and comprising: (This simply recites that the abstract idea of generating user-specific recommendations is performed on a computer, which amounts to mere instructions to apply an exception.)
receiving a communication; (Receiving data is a known process in computers. This amounts to mere instructions to apply an exception.)
providing an updated sub-GNN based on the event data; (Transmitting data is a known process in computers. This amounts to mere instructions to apply an exception.)
providing reason text from a large language model (LLM) responsive to a prompt; and (Transmitting data is a known process in computers. This also recites the use of a machine learning component (LLM) at a high-level of generality without detail. This amounts to mere instructions to apply an exception.)
transmitting a notification to the user, the notification comprising the recommendation and the reason text. (Transmitting data is a known process in computers. This amounts to mere instructions to apply an exception.)
Regarding claim 2, the rejection of claim 1 is incorporated herein. The following is an abstract idea:
wherein providing an updated sub-GNN based on the event data comprises updating weight matrices of edges of the sub-GNN based on the event data. (Updating weight matrices of edges of the graph can be practically performed in the human mind, for example, assigning a weight to each edge based on perceived importance of the edge. This is a mental process.)
Regarding claim 3, the rejection of claim 2 is incorporated herein. The following is an abstract idea:
wherein updating comprises each node of the sub-GNN executing computation and aggregation to update a respective weight matrix. (Executing computation and aggregation can be practically performed in the human mind, i.e. for each node, adding values to a weight matrix. This is a mental process.)
Regarding claim 4, the rejection of claim 1 is incorporated herein. The following is an abstract idea:
wherein the prompt is generated using a prompt template. (Generating a prompt using a prompt template can be practically performed in the human mind. This is a mental process.)
Regarding claim 5, the rejection of claim 1 is incorporated herein. The following claim elements are additional elements which, taken alone or in combination with the other additional elements, do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception:
receiving feedback from the user regarding the event, at least a portion of the feedback being stored as an attribute in a node of the GNN, the node representing the user. (Receiving and storing data are known processes in computing. This amounts to mere instructions to apply an exception.)
Regarding claim 6, the rejection of claim 1 is incorporated herein. The following is an abstract idea:
wherein the sub-GNN is a k-hop representation from a node of the user within the GNN. (Following claim 1, the generation of the sub-GNN is a mental process. Generating a sub-GNN wherein the sub-GNN is a k-hop representation of a node of the user within the GNN can still be practically performed in the human mind. For example, one could determine the nodes and edges within a k-hop distance of the user node. This is a continuation of the mental process of the step of "determining, from a graph neural network (GNN), a sub-GNN that is specific to the user, the GNN comprising a data structure that represents users of an enterprise and relationships between users, the sub-GNN representing a portion of the GNN" in claim 1. )
Regarding claim 7, the rejection of claim 8 is incorporated herein. The following is an abstract idea:
wherein aggregating user data and event data is executed in response to determining that the communication describes the event. (Aggregating data in response to a determination can be practically performed in the human mind. This is a mental process.)
Regarding claim 8, the following is an abstract idea.
generating user-specific recommendations with reason text (Generating user-specific recommendations with reason text can be practically performed in the human mind. This is a mental process.)
The following claim elements are additional elements which, taken alone or in combination with the other additional elements, do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception:
A non-transitory computer-readable storage medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for … the operations comprising: (This recites generic computer processes (data storage, instruction execution) and generic computer components (non-transitory computer-readable storage medium, processors, computer operations/instructions), which implement the abstract idea of generating user-specific recommendations. This amounts to mere instructions to apply an exception.)
The remainder of claim 8 recites substantially similar subject matter to claim 1 and is rejected with the same rationale, mutatis mutandis.
Claims 9-14 recite substantially similar subject matter to claims 2-7 respectively and are rejected with the same rationale, mutatis mutandis.
Regarding claim 15, the following claim elements are additional elements which, taken alone or in combination with the other additional elements, do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception:
A system, comprising: (This recites a generic system on which to apply the judicial exceptions. This amounts to mere instructions to apply an exception.)
a computing device; and (This recites a generic computing device on which to apply the judicial exceptions. This amounts to mere instructions to apply an exception.)
a computer-readable storage device coupled to the computing device and having instructions stored thereon which, when executed by the computing device, cause the computing device to: (This recites generic computer components (computer-readable storage and program instructions) on which to apply the judicial exceptions. This amounts to mere instructions to apply an exception.)
The remainder of claim 15 recites substantially similar subject matter to claim 1 and is rejected with the same rationale, mutatis mutandis.
Claims 16-20 recite substantially similar subject matter to claims 2-6 respectively and are rejected with the same rationale, mutatis mutandis.
Claim Rejections - 35 USC § 103
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 (i.e., changing from AIA to pre-AIA ) 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.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Gao (“Chat-Rec: Towards Interactive and Explainable LLMs-Augmented Recommender System”, April 2023) and Wang (“Neural Graph Collaborative Filtering”, 2020).
Regarding claim 1, Gao teaches
A computer-implemented method for generating user-specific recommendations with reason text, the method being executed by one or more processors and comprising: (Page 3 states "We introduce a novel and effective paradigm called Chat-Rec, which combines traditional recommender systems with LLMs through prompts, leveraging LLMs' ability to learn from context. Chat-Rec employs LLMs as a recommender system interface, enabling multi-round recommendations, enhancing interactivity and explainability." The paradigm is interpreted as the method, which one of ordinary skill in the art would recognize as performed on a computer. Page 5, Fig. 1 shows the dialogue, wherein the second answer comprises reason text, i.e. an explanation of the recommendation. Page 5 states "User-item history interactions, which refers to the user's past interactions with items, such as items they have clicked, purchased, or rated. This information is used to understand the user's preferences and to personalize the recommendation." Therefore, the recommendations are user-specific. )
receiving a communication; (The middle of Fig. 1 on page 5 shows that the user query, interpreted as the communication, is received.)
aggregating user data and event data, the user data representative of a user that as an addressee of the communication, the event data representative of an event described in the communication; (Page 5 states "The prompt constructor module in the enhanced recommender system takes multiple inputs to generate a natural language paragraph that captures the user's query and recommendation information. The inputs are as follows: User-item history interactions, which refers to the user's past interactions with items, such as items they have clicked, purchased, or rated. This information is used to understand the user's preferences and to personalize the recommendation. User profile, which contains demographic and preference information about the user. This may include age, gender, location, and interests. The user profile helps the system understand the user's characteristics and preferences. User query Qi, which is the user's specific request for information or recommendation. This may include a specific item or genre they are interested in, or a more general request for recommendations in a particular category." User-item history interactions are interpreted as event data, and the user profile is interpreted as the user data. Pages 5-6 state "Formally, based on the aforementioned inputs, the prompt constructor module generates a natural language paragraph that summarizes the user's query and recommendation information, and provides a more personalized and relevant response to the user's request." Therefore, when a recommendation is requested, the event and user data are aggregated to generate a response. "As the user-item interactions are used to determine the recommendation for the event (specific request for recommendation) in the user query (which is in the communication), the event data are representative of event described in the communication.)
providing reason text from a large language model (LLM) responsive to a prompt; and (Page 7 states "The dialogue on the left shows that when a user asks why the movie was recommended, LLM can give an explanation based on the user's preferences and specific information about the recommended movie." The explanation is interpreted as the reason text. Page 16, Fig. 6 and Fig. 7 show the prompts for recommendation. As the reason text is given after the prompt for recommendation, the reason text is responsive to the prompt.)
transmitting a notification to the user, the notification comprising the recommendation and the reason text. (Fig. 2 shows that the recommendation and reason text are transmitted to the user. The answers that include the recommendation and reason text are interpreted as the notification.)
Gao does not appear to explicitly teach
determining, from a graph neural network (GNN), a sub-GNN that is specific to the user, the GNN comprising a data structure that represents users of an enterprise and relationships between users, the sub-GNN representing a portion of the GNN;
providing an updated sub-GNN based on the event data;
generating a recommendation regarding the event using the updated sub-GNN, the recommendation being specific to the user;
However, Wang—directed to analogous art—teaches
determining, from a graph neural network (GNN), a sub-GNN that is specific to the user, the GNN comprising a data structure that represents users of an enterprise and relationships between users, the sub-GNN representing a portion of the GNN; (Page 2 states "Figure 1 illustrates the concept of high-order connectivity. The user of interest for recommendation is u1, labeled with the double circle in the left subfigure of user-item interaction graph. The right subfigure shows the tree structure that is expanded from u1. The high-order connectivity denotes the path that reaches u1 from any node with the path length l larger than 1. Such high-order connectivity contains rich semantics that carry collaborative signal. For example, the path u1 ←i2 ←u2 indicates the behavior similarity between u1 and u2, as both users have interacted with i2; the longer path u1 ←i2 ←u2 ←i4 suggests that u1 is likely to adopt i4, since her similar useru2 has consumed i4 before. Moreover, from the holistic view of l = 3, item i4 is more likely to be of interest to u1 than item i5, since there are two paths connecting <i4,u1>, while only one path connects <i5,u1>." Figure 1 shows that the graph represents users and their relationships (behavior similarities). Page 3 states "With the representations augmented by first-order connectivity modeling, we can stack more embedding propagation layers to explore the high-order connectivity information. Such high-order connectivities are crucial to encode the collaborative signal to estimate the relevance score between a user and item. By stacking l embedding propagation layers, a user (and an item) is capable of receiving the messages propagated from its l-hop neighbors." Therefore, the sub-graph representation formed by the user and its l-hop neighbors is interpreted as the sub-GNN. Page 6 states "To evaluate the effectiveness of NGCF, we conduct experiments on three benchmark datasets: Gowalla, Yelp2018∗2, and Amazon-book, which are publicly accessible and vary in terms of domain, size, and sparsity. We summarize the statistics of three datasets in Table 1. Gowalla: This is the check-in dataset [21] obtained from Gowalla, where users share their locations by checking-in. To ensure the quality of the dataset, we use the 10-core setting [10], i.e., retaining users and items with at least ten interactions. Yelp2018∗: This dataset is adopted from the 2018 edition of the Yelp challenge. Wherein, the local businesses like restaurants and bars are viewed as the items. We use the same 10-core setting in order to ensure data quality. Amazon-book: Amazon-review is a widely used dataset for product recommendation [9]. We select Amazon-book from the collection. Similarly, we use the 10-core setting to ensure that each user and item have at least ten interactions." Each dataset includes users of an enterprise, Gowalla, Yelp, and Amazon respectively. Therefore, the user-interaction graph created using these datasets would represent the users of an enterprise.)
providing an updated sub-GNN based on the event data; (Pages 3-4 state "By stacking l embedding propagation layers, a user (and an item) l is capable of receiving the messages propagated from its l-hop neighbors. As Figure 2 displays, in the l-th step, the representation of user u is recursively formulated as: [Eq. (5)] wherein the messages being propagated are defined as follows, [Eq. 6], where
W
1
l
,
W
2
l
,
∈
R
d
l
×
d
l
-
1
are the trainable transformation matrices, and
d
l
is the transformation size;
e
i
l
-
1
is the item representation generated from the previous message-passing steps, memorizing the representation of user
u
at layer
l
." Therefore, when the weight matrices are trained, the sub-graph representation of user
u
is updated. As the training data is the user-item interactions, see page 6, section 4.1, the sub-GNN is updated based on the event data.)
generating a recommendation regarding the event using the updated sub-GNN, the recommendation being specific to the user; (Page 4 states "Finally, we conduct the inner product to estimate the user’s preference towards the target item: [Eq. 10]." Estimating the user’s preference is interpreted as generating a recommendation. As the user-item interactions are involved in the GNN, the recommendation is regarding the event. As Eq. 10 takes a specific user as input, the recommendation is specific to the user.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Gao and Wang because, as Wang states in the abstract, "In this work, we propose to integrate the user-item interactions — more specifically the bipartite graph structure—into the embedding process. We develop a new recommendation framework Neural Graph Collaborative Filtering (NGCF), which exploits the user-item graph structure by propagating embeddings on it. This leads to the expressive modeling of high-order connectivity in user-item graph, effectively injecting the collaborative signal into the embedding process in an explicit manner. We conduct extensive experiments on three public benchmarks, demonstrating significant improvements over several state-of-the-art models like HOPRec [40] and Collaborative Memory Network [5]. Further analysis verifies the importance of embedding propagation for learning better user and item representations, justifying the rationality and effectiveness of NGCF."
Regarding claim 2, the rejection of claim 1 is incorporated herein. Gao does not appear to explicitly teach
wherein providing an updated sub-GNN based on the event data comprises updating weight matrices of edges of the sub-GNN based on the event data
However, Wang—directed to analogous art—teaches
wherein providing an updated sub-GNN based on the event data comprises updating weight matrices of edges of the sub-GNN based on the event data. (Page 3 states "For a connected user-item pair
(
u
,
i
)
, we define the message from
i
to
u
as:
m
u
←
i
=
f
e
i
,
e
u
,
p
u
i
,
(
2
)
where
m
u
←
i
is the message embedding (i.e., the information to be propagated)" As the nodes of the graph are the users and items, the messages passed between the nodes are interpreted as the edges. Pages 3-4 state "By stacking l embedding propagation layers, a user (and an item) l is capable of receiving the messages propagated from its l-hop neighbors. As Figure 2 displays, in the l-th step, the representation of user u is recursively formulated as: [Eq. (5)] wherein the messages being propagated are defined as follows, [Eq. 6], where
W
1
l
,
W
2
l
,
∈
R
d
l
×
d
l
-
1
are the trainable transformation matrices, and
d
l
is the transformation size;
e
i
l
-
1
is the item representation generated from the previous message-passing steps, memorizing the representation of user
u
at layer
l
." Eq. 6 shows equations for
m
u
←
i
l
and
m
u
←
u
l
(messages passed between nodes, interpreted as edges), which include the trainable transformation matrices, interpreted as the weights. Therefore, when the weight matrices are trained, the sub-graph representation of user
u
is updated. As the training data is the user-item interactions, see page 6, section 4.1, the sub-GNN is updated based on the event data.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Gao and Wang for the reasons given above in regards to claim 1.
Regarding claim 3, the rejection of claim 2 is incorporated herein. Gao does not appear to explicitly teach
wherein updating comprises each node of the sub-GNN executing computation and aggregation to update a respective weight matrix.
However, Wang—directed to analogous art—teaches
wherein updating comprises each node of the sub-GNN executing computation and aggregation to update a respective weight matrix. (Page 3 states "Intuitively, the interacted items provide direct evidence on a user’s preference [16, 39]; analogously, the users that consume an item can be treated as the item’s features and used to measure the collaborative similarity of two items. We build upon this basis to perform embedding propagation between the connected users and items, formulating the process with two major operations: message construction and message aggregation." As the nodes of the sub-GNN are the users and items, the tasks are executed by each node of the sub-GNN. The message construction is interpreted as the computation and the message aggregation is interpreted as the computation. As propagation is required for training, the propagation updates the respective weight matrixes shown in Eq. 3 for computation and Eq. 4 for aggregation.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Gao and Wang for the reasons given above in regards to claim 1.
Regarding claim 4, the rejection of claim 1 is incorporated herein. Gao teaches
wherein the prompt is generated using a prompt template. Page 16, Fig. 6 and Fig. 7 show the templates for the prompts for recommendation.)
Regarding claim 5, the rejection of claim 1 is incorporated herein. Gao teaches
receiving feedback from the user regarding the event, (Page 5 states ""The prompt constructor module in the enhanced recommender system takes multiple inputs to generate a natural language paragraph that captures the user's query and recommendation information. The inputs are as follows: User-item history interactions, which refers to the user's past interactions with items, such as items they have clicked, purchased, or rated. This information is used to understand the user's preferences and to personalize the recommendation. User profile, which contains demographic and preference information about the user. This may include age, gender, location, and interests. The user profile helps the system understand the user's characteristics and preferences. User query Qi, which is the user's specific request for information or recommendation. This may include a specific item or genre they are interested in, or a more general request for recommendations in a particular category." User-item history interactions are interpreted as feedback. As the user-item interactions are used to determine the recommendation for the event, the event data regards the event.)
Gao does not appear to explicitly teach
at least a portion of the feedback being stored as an attribute in a node of the GNN, the node representing the user.
However, Wang—directed to analogous art—teaches
at least a portion of the feedback being stored as an attribute in a node of the GNN, the node representing the user. (Pages 3-4 state "By stacking l embedding propagation layers, a user (and an item) l is capable of receiving the messages propagated from its l-hop neighbors. As Figure 2 displays, in the l-th step, the representation of user u is recursively formulated as: [Eq. (5)] wherein the messages being propagated are defined as follows, [Eq. 6], where
W
1
l
,
W
2
l
,
∈
R
d
l
×
d
l
-
1
are the trainable transformation matrices, and
d
l
is the transformation size;
e
i
l
-
1
is the item representation generated from the previous message-passing steps, memorizing the representation of user
u
at layer
l
." As the nodes of the graphs are the users and items, the representation of the user is an attribute of the graph. As the GNN models the user-item interactions and creates the representation by message passing, the user-item interactions, interpreted as the feedback, are stored in the representation, which is stored as an attribute in the user and item nodes.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Gao and Wang for the reasons given above in regards to claim 1.
Regarding claim 6, the rejection of claim 1 is incorporated herein. Gao does not appear to explicitly teach
wherein the sub-GNN is a k-hop representation from a node of the user within the GNN.
However, Wang—directed to analogous art—teaches
wherein the sub-GNN is a k-hop representation from a node of the user within the GNN. ((Page 2 states "Figure 1 illustrates the concept of high-order connectivity. The user of interest for recommendation is u1, labeled with the double circle in the left subfigure of user-item interaction graph. The right subfigure shows the tree structure that is expanded from u1. The high-order connectivity denotes the path that reaches u1 from any node with the path length l larger than 1. Such high-order connectivity contains rich semantics that carry collaborative signal. For example, the path u1 ←i2 ←u2 indicates the behavior similarity between u1 and u2, as both users have interacted with i2; the longer path u1 ←i2 ←u2 ←i4 suggests that u1 is likely to adopt i4, since her similar useru2 has consumed i4 before. Moreover, from the holistic view of l = 3, item i4 is more likely to be of interest to u1 than item i5, since there are two paths connecting <i4,u1>, while only one path connects <i5,u1>." Figure 1 shows that the graph represents users and their relationships (behavior similarities). Page 3 states "With the representations augmented by first-order connectivity modeling, we can stack more embedding propagation layers to explore the high-order connectivity information. Such high-order connectivities are crucial to encode the collaborative signal to estimate the relevance score between a user and item. By stacking l embedding propagation layers, a user (and an item) is capable of receiving the messages propagated from its l-hop neighbors." Therefore, the sub-graph representation formed by the user and its l-hop neighbors is interpreted as the sub-GNN.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Gao and Wang for the reasons given above in regards to claim 1.
Regarding claim 7, the rejection of claim 8 is incorporated herein. Gao teaches
wherein aggregating user data and event data is executed in response to determining that the communication describes the event. (Page 5 states "The prompt constructor module in the enhanced recommender system takes multiple inputs to generate a natural language paragraph that captures the user's query and recommendation information. The inputs are as follows: User-item history interactions, which refers to the user's past interactions with items, such as items they have clicked, purchased, or rated. This information is used to understand the user's preferences and to personalize the recommendation. User profile, which contains demographic and preference information about the user. This may include age, gender, location, and interests. The user profile helps the system understand the user's characteristics and preferences. User query Qi, which is the user's specific request for information or recommendation. This may include a specific item or genre they are interested in, or a more general request for recommendations in a particular category." User-item history interactions are interpreted as event data, and the user profile is interpreted as the user data. Pages 5-6 state "Formally, based on the aforementioned inputs, the prompt constructor module generates a natural language paragraph that summarizes the user's query and recommendation information, and provides a more personalized and relevant response to the user's request." Page 5 states "If the task is determined to be a recommendation task, the module uses R to generate a candidate set of items. Otherwise, it directly outputs a response to the user, such as an explanation of a generation task or a request for item details." Therefore, when a recommendation task is determined, which would include an event (type of recommendation) in the communication, the event and user data are aggregated to generate a response.)
Regarding claim 8, Gao teaches
A non-transitory computer-readable storage medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for generating user-specific recommendations with reason text the operations comprising: (Page 3 states "We introduce a novel and effective paradigm called Chat-Rec, which combines traditional recommender systems with LLMs through prompts, leveraging LLMs' ability to learn from context. Chat-Rec employs LLMs as a recommender system interface, enabling multi-round recommendations, enhancing interactivity and explainability." The paradigm is interpreted as the method, which one of ordinary skill in the art would recognize as performed on a computer, which would necessarily include a non-transitory computer-readable storage medium coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations for performing the method, the operations comprising the method. Page 5, Fig. 1 shows the dialogue, wherein the second answer comprises reason text, i.e. an explanation of the recommendation. Page 5 states "User-item history interactions, which refers to the user's past interactions with items, such as items they have clicked, purchased, or rated. This information is used to understand the user's preferences and to personalize the recommendation." Therefore, the recommendations are user-specific. )
The remainder of claim 8 recites substantially similar subject matter to claim 1 and is rejected with the same rationale, mutatis mutandis.
Claims 9-14 recite substantially similar subject matter to claims 2-7 respectively and are rejected with the same rationale, mutatis mutandis.
Regarding claim 15, Gao teaches
A system, comprising: a computing device; and a computer-readable storage device coupled to the computing device and having instructions stored thereon which, when executed by the computing device, cause the computing device to: (Page 3 states "We introduce a novel and effective paradigm called Chat-Rec, which combines traditional recommender systems with LLMs through prompts, leveraging LLMs' ability to learn from context. Chat-Rec employs LLMs as a recommender system interface, enabling multi-round recommendations, enhancing interactivity and explainability." The paradigm is interpreted as the method, which one of ordinary skill in the art would recognize as performed on a computer system, which would necessarily include a computing device; and a computer-readable storage device coupled to the computing device and having instructions stored thereon which, when executed by the computing device, cause the computing device to perform the method.)
The remainder of claim 15 recites substantially similar subject matter to claim 1 and is rejected with the same rationale, mutatis mutandis.
Claims 16-20 recite substantially similar subject matter to claims 2-6 respectively and are rejected with the same rationale, mutatis mutandis.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JESSICA THUY PHAM whose telephone number is (571)272-2605. The examiner can normally be reached Monday - Friday, 9 A.M. - 5:00 P.M..
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, Li Zhen can be reached at (571) 272-3768. 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.
/J.T.P./Examiner, Art Unit 2121
/Li B. Zhen/Supervisory Patent Examiner, Art Unit 2121
Prosecution Insights