# US 2020/0004813 A1 · Social autonomous agent implementation using lattice queries and relevancy detection

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0004813 A1
- Document ID: US20200004813A1
- Status: Published application
- Kind: A1
- Application number: 16/460,503
- Filed: 2019-07-02
- Published: 2020-01-02
- Earliest claimed: 2018-07-02
- Assignees: Oracle International Corporation
- Applicants: Oracle International Corporation
- Inventors: Boris Galitsky
- Web page: https://patentgrep.com/patent/US20200004813A1

## CPC

- G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
- G06F 16/3329: Natural language query formulation
- G06F 16/9027: {Trees}
- G06F 16/9538: Presentation of query results
- G06F 40/279: Recognition of textual entities
- G06F 40/30: Semantic analysis
- G06F 40/35: Discourse or dialogue representation
- G06F 40/51: Translation evaluation
- H04L 51/02: using automatic reactions or user delegation, e.g. automatic replies or chatbot-generated messages
- H04L 51/216: Handling conversation history, e.g. grouping of messages in sessions or threads
- H04L 51/52: for supporting social networking services
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Techniques for computer-generated conversation are disclosed. In an example, a method identifies text postings from a conversation. The method creates, for each text fragment of each text posting, a syntactic tree and a discourse tree. The method creates parse thickets, each parse thicket including the syntactic tree and discourse tree of a unique pair of text postings. The method extracts, from each parse thicket, a common text segment and obtains a set of candidate search results by providing the common text segments to a search engine. The candidate search results can be further refined for relevancy and mental state and posted as a response to a conversation.

## Claims

### Claim 1

A method of performing a computer-generated conversation, the method comprising:
identifying, from a conversation, text posts, wherein each text post comprises text fragments;
creating, for each text fragment, a syntactic tree;
creating, for each text fragment, a discourse tree that comprises terminal nodes and nonterminal nodes, wherein each terminal node is associated with a fragment of the respective text and each nonterminal node represents a rhetorical relationship between two of the fragments;
creating parse thickets, each parse thicket comprising a first discourse tree and first syntactic tree from a first text post and a second discourse tree and second syntactic tree from a second text post, wherein the a first text post and a second text post are a unique pair and wherein the parse thicket comprises one or more relationships between the first text post and the second text post;
extracting, from each parse thicket, a common text segment comprising text that (i) corresponds to a communicative action that is common to the respective first text post and the respective second text post, (ii) corresponds to a rhetorical relation that is common to the respective first text post and the respective second text post, or (iii) has a common verb signature between the respective first text post and a verb in the respective second text post; and
obtaining candidate search results by providing the common text segments to a search engine.

### Claim 2 (depends on 1)

The method of claim 1, wherein the relationships comprise one or more of: (i) two nouns represented by fragments in a syntactic tree of the first text post and the syntactic tree of the second text post, wherein the nouns represent a common entity or (ii) a rhetorical relation between fragments in a first discourse tree of the first text post and a second discourse tree of the second text post.

### Claim 3 (depends on 1)

The method of claim 1, further comprising:
determining, for each combination of candidate search result and text post, a candidate parse thicket;
calculating, from the candidate parse thickets and for each candidate search result, an aggregate similarity score representing a rhetorical similarity of the respective candidate search result to the text posts; and
selecting, as an output, the candidate search result that comprises a highest score.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
determining, from the text posts, a last text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the first discourse tree comprises one or more rhetorical relations of type elaboration, selecting, a candidate search result that corresponds to a discourse tree that comprises a rhetorical relation of (i) attribution, (ii) contrast, (iii) explanation, or (iv) cause.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
determining, from the text posts, a text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the one or more nodes of the first discourse tree comprise communicative actions, selecting a candidate search result that corresponds to a discourse tree that comprises a rhetorical relation of (i) restatement, (ii) confirmation, (iii) attribution, (iv) summary, or (v) evidence.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
determining, from the text posts, a text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the first discourse tree comprises one or more rhetorical relations of type elaboration, joint, or sequence, selecting a candidate search result that comprises a node of a discourse tree, the node labeled with communicative actions and does not comprise nodes that are labeled with mental actions.

### Claim 7 (depends on 1)

The method of claim 1, wherein a first text post is an original post on a social media page and a second text post is derivative of the first text post.

### Claim 8 (depends on 1)

The method of claim 1, further comprising providing one of the candidate search results to a user interface or transmitting one of the candidate search results to a website.

### Claim 9

A system comprising:
a computer-readable medium storing non-transitory computer-executable program instructions for applying an image effect within an image processing application; and
a processing device communicatively coupled to the computer-readable medium for executing the non-transitory computer-executable program instructions, wherein executing the non-transitory computer-executable program instructions configures the processing device to perform operations comprising:
identifying, from a conversation, text posts, wherein each text post comprises text fragments;
creating, for each text fragment, a syntactic tree;
creating, for each text fragment, a discourse tree that comprises terminal nodes and nonterminal nodes, wherein each terminal node is associated with a fragment of the respective text and each nonterminal node represents a rhetorical relationship between two of the fragments;
creating parse thickets, each parse thicket comprising a first discourse tree and first syntactic tree from a first text post and a second discourse tree and second syntactic tree from a second text post, wherein the a first text post and a second text post are a unique pair and wherein the parse thicket comprises one or more relationships between the first text post and the second text post;
extracting, from each parse thicket, a common text segment comprising text that (i) corresponds to a communicative action that is common to the respective first text post and the respective second text post, (ii) corresponds to a rhetorical relation that is common to the respective first text post and the respective second text post, or (iii) has a common verb signature between the respective first text post and a verb in the respective second text post; and
obtaining candidate search results by providing the common text segments to a search engine.

### Claim 10 (depends on 9)

The system of claim 9, wherein the relationships comprise one or more of: (i) two nouns represented by fragments in a syntactic tree of the first text post and the syntactic tree of the second text post, wherein the nouns represent a common entity or (ii) a rhetorical relation between fragments in a first discourse tree of the first text post and a second discourse tree of the second text post.

### Claim 11 (depends on 9)

The system of claim 9, wherein the operations further comprise:
determining, for each combination of candidate search result and text post, a candidate parse thicket;
calculating, from the candidate parse thickets and for each candidate search result, an aggregate similarity score representing a rhetorical similarity of the respective candidate search result to the text posts; and
selecting, as an output, the candidate search result that comprises a highest score.

### Claim 12 (depends on 9)

The system of claim 9, wherein the operations further comprise:
determining, from the text posts, a last text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the first discourse tree comprises one or more rhetorical relations of type elaboration, selecting, a candidate search result that corresponds to a discourse tree that comprises a rhetorical relation of (i) attribution, (ii) contrast, (iii) explanation, or (iv) cause.

### Claim 13 (depends on 9)

The system of claim 9, wherein the operations further comprise:
determining, from the text posts, a text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the one or more nodes of the first discourse tree comprise communicative actions, selecting, a candidate search result that corresponds to a discourse tree that comprises a rhetorical relation of (i) restatement, (ii) confirmation, (iii) attribution, (iv) summary, or (v) evidence.

### Claim 14 (depends on 9)

The system of claim 9, wherein the operations further comprise:
determining, from the text posts, a text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the first discourse tree comprises one or more rhetorical relations of type elaboration, joint, or sequence, selecting, a candidate search result that comprises a node of a discourse tree, the node labeled with communicative actions and does not comprise nodes that are labeled with mental actions.

### Claim 15 (depends on 9)

The system of claim 9, wherein a first text post is an original post on a social media page and a second text post is derivative of the first text post.

### Claim 16 (depends on 9)

The system of claim 9, wherein the operations further comprise providing one of the candidate search results to a user interface or transmitting one of the candidate search results to a website.

### Claim 17

A non-transitory computer-readable storage medium storing computer-executable program instructions, wherein when executed by a processing device, the program instructions cause the processing device to perform operations comprising:
identifying, from a conversation, text posts, wherein each text post comprises text fragments;
creating, for each text fragment, a syntactic tree;
creating, for each text fragment, a discourse tree that comprises terminal nodes and nonterminal nodes, wherein each terminal node is associated with a fragment of the respective text and each nonterminal node represents a rhetorical relationship between two of the fragments;
creating parse thickets, each parse thicket comprising a first discourse tree and first syntactic tree from a first text post and a second discourse tree and second syntactic tree from a second text post, wherein the a first text post and a second text post are a unique pair and wherein the parse thicket comprises one or more relationships between the first text post and the second text post;
extracting, from each parse thicket, a common text segment comprising text that (i) corresponds to a communicative action that is common to the respective first text post and the respective second text post, (ii) corresponds to a rhetorical relation that is common to the respective first text post and the respective second text post, or (iii) has a common verb signature between the respective first text post and a verb in the respective second text post; and
obtaining candidate search results by providing the common text segments to a search engine.

### Claim 18 (depends on 17)

The non-transitory computer-readable storage medium of claim 17, wherein the operations further comprise:
determining, for each combination of candidate search result and text post, a candidate parse thicket;
calculating, from the candidate parse thickets and for each candidate search result, an aggregate similarity score representing a rhetorical similarity of the respective candidate search result to the text posts; and
selecting, as an output, the candidate search result that comprises a highest score.

### Claim 19 (depends on 17)

The non-transitory computer-readable storage medium of claim 17, wherein the operations further comprise:
determining, from the text posts, a last text post that occurred after the other text posts;
determining, a first discourse tree for the last text post;
determining a discourse tree for each candidate search result; and
responsive to determining that the first discourse tree comprises one or more rhetorical relations of type elaboration, selecting, a candidate search result that corresponds to a discourse tree that comprises a rhetorical relation of (i) attribution, (ii) contrast, (iii) explanation, or (iv) cause.

### Claim 20 (depends on 17)

The non-transitory computer-readable storage medium of claim 17, wherein a first text post is an original post on a social media page and a second text post is derivative of the first text post.
