US-12626696-B2 - System and method for generating a closed domain conversation

US12626696B2US 12626696 B2US12626696 B2US 12626696B2US-12626696-B2

Abstract

An embodiment herein provides a system and a method for generating a closed domain conversation with a user ( 102 ) in real-time based on user's interest. The system includes a robot ( 104 ) including a memory ( 110 ) that includes one or more instructions and a processor ( 112 ) that executes the one or more instructions. The processor ( 112 ) is configured to initiate a conversation with the user ( 102 ) which is a machine-initiated conversation, determine a flow of the conversation with the user ( 102 ) by analyzing replies of the user ( 102 ), generate the closed domain conversation by providing one or more contents related to at least one topic or category which is personalized based on a combination of one or more properties of the user's interest, and enable the user ( 102 ) to interact with the robot ( 104 ) using the closed domain conversation in real-time based on the user's interest.

Inventors

PRASHANT IYENGAR
Hardik Godara

Assignees

RN Chidakashi Technologies Private Limited

Dates

Publication Date: 20260512
Application Date: 20220323
Priority Date: 20210324

Claims (8)

1 . A computer-implemented conversational robot for generating a closed domain conversation with a user in real-time based on user's interest, the computer-implemented conversational robot comprising: a memory comprising one or more instructions and storing micro-conversation data structures, user-interest parameters, and instructions for generating a closed-domain conversation; and a processor that executes the one or more instructions, wherein the processor is configured to: initiate a conversation with the user, wherein the conversation comprises a machine-initiated conversation; generate an input conversation from the user in real time by capturing (i) an audio input of the user using an audio input unit of the robot and (ii) a video input of the user using a video input unit of the robot; convert the input conversation into text; analyze the converted text to detect one or more user-specific properties comprising at least one of: age, location, gender, likes or dislikes, sentiments derived from the input conversation, or environmental or contextual parameters; select, from the memory or from a server communicatively connected to the robot over a wired or wireless network, a micro-conversation from a plurality of micro-conversations based on the one or more user-specific properties; determine a flow of the conversation with the user by transitioning between the plurality of micro-conversations that evaluates the input conversation, the sentiments, or the environmental parameters; generate, using the selected micro-conversation and the determined flow, the closed domain conversation by selecting one or more contents related to at least one topic or category which is personalized based on the user's interest; output, through an audio output unit or an expression output unit of the robot, a conversation response comprising at least one of an audio output or a robotic expression; and in response to detecting that the conversation is not relevant to the user's interest, automatically clear stored conversation context and initiate a new conversation with updated micro-conversation selections, thereby enabling the user to interact with the robot using the closed domain conversation in real-time based on the user's interest.
2 . The computer-implemented conversational robot as claimed in claim 1 , wherein the processor is configured to: generate one or more opinion sets by analyzing the conversation response of the user during the conversation, wherein the processor determines the user's interest based on at least one of user preferences, likes and dislikes of the user, wherein the one or more opinion sets enables the robot to update the user's interest and the user preferences.
3 . The computer-implemented conversational robot as claimed in claim 1 , wherein the processor is configured to: determine and redirect the flow of conversation with the sentiments by analyzing the conversation response of the user; and generate the closed domain conversation by filtering the input conversation with the one or more contents.
4 . The computer-implemented conversational robot as claimed in claim 1 , wherein the robot comprises a robot control system that controls a movement of the robot by analyzing the audio output and the robotic expression.
5 . The computer-implemented conversational robot as claimed in claim 1 , wherein the processor is configured to: receive the input conversation from the user, wherein the input conversation is a user-initiated conversation; recognize a voice of the user to identify the user; and automatically initiate the conversation with the user by analyzing the input conversation.
6 . The computer-implemented conversational robot as claimed in claim 1 , wherein the processor is configured to: control the flow of the conversation in at least one of sequential order or random order, wherein the sequential order comprises a greeting message, a main conversation body, an end message, or a goodbye message which are connected in the sequential order, wherein the random order comprises the greeting message, the main conversation body, the end message, or the goodbye message which are connected in the random order that selects the flow of the conversation randomly to generate a different flow of conversation.
7 . A method for generating a closed domain conversation with a user in real-time based on user's interest, wherein the method comprises, initiating a conversation with the user, wherein the conversation comprises a machine-initiated conversation; generating an input conversation from the user in real-time by capturing (i) an audio input of the user using an audio input unit of the robot and (ii) a video input of the user using a video input unit of the robot; converting the input conversation into text; analyzing the converted text to detect one or more user-specific properties comprising at least one of: age, location, gender, likes or dislikes, sentiments derived from the input conversation, or environmental or contextual parameters; selecting, from a memory or from a server communicatively connected to the robot over a wired or wireless network, a micro-conversation from a plurality of micro-conversations based on the one or more user-specific properties determining a flow of the conversation with the user by transitioning between the plurality of micro-conversations that evaluates the input conversation, the sentiments, or the environmental parameters; generating, using the selected micro-conversation and the determined flow, the closed domain conversation by selecting one or more contents related to at least one topic or category which is personalized based on the user's interest; outputting, through an audio output unit or an expression output unit of the robot, a conversation response comprising at least one of an audio output or a robotic expression; and in response to detecting that the conversation is not relevant to the user's interest, automatically clear stored conversation context and initiate a new conversation with updated micro-conversation selections, thereby enabling the user to interact with the robot using the closed domain conversation in real-time based on the user's interest.
8 . The method as claimed in claim 7 , wherein the method comprises, receiving the input conversation from the user, wherein the input conversation is a user-initiated conversation; recognizing a voice of the user to identify the user; and automatically initiating the conversation with the user based on the input conversation.

Description

BACKGROUND Technical Field Embodiments of this disclosure generally relate to a closed domain conversation, and more particularly, to a system and a method for generating the closed domain conversation with a user in real-time based on user's interest. Description of the Related Art Closed domain conversation is a type of conversation that allows a user to interact with a plurality of applications, a plurality of websites, and a plurality of devices, etc. via text, voice, audio, video, etc. The closed domain conversation uses a plurality of technologies to process and contextualize user input to respond to the user. Nowadays, the closed domain conversation has been used for businesses in order to create personalized customer experiences. Companies continue to develop a variety of closed domain conversations to interact with customers. Though a variety of closed domain conversation emerges day by day, more research are still going on to develop the closed domain conversation that enables fastest user interaction which in turn improves the user's conversational experience. An architecture for the closed domain conversation includes a plurality of nodes that illustrates execution steps of operations and edges that illustrates corresponding interconnections between the nodes. Conventional architecture for the closed domain conversation consumes more time in terms of computer resource processing and the response time for the user input is high. Thus, the conventional architecture for closed domain conversation may impede the functioning of a computing device that is used for a conversational purpose. Further, the existing architecture for closed domain conversation is most complex in terms of design and does not support to continue any form of user-initiated conversation or machine-initiated conversation. Further, the existing system does not continue the conversation with a new topic which are liked by the user and the existing system does not continue the conversation based on interesting facts. Further, the existing system does not move to the next conversation if the user won't like the topic and does not continue the chat with the user interestingly. Further, the existing system does not continue the conversation based on the user interest or does not change the topic based on the user interest. Existing closed domain conversation requires different conversational architecture for supporting different input modalities which in turn takes up more computational resources and increases the complexity of implementation. Hence, the existing closed domain conversation system supports only limited input topics. Still, research is going on for the closed domain conversation that is capable of supporting any topics initiated by the user or the machine. Accordingly, there remains a need for a system for generating the closed domain conversation that is capable of continuing the conversation with the user in real-time based on the user's interest and supporting any form of user-initiated conversation. SUMMARY An embodiment herein provides a system for generating a closed domain conversation with a user in real-time based on user's interest. The system includes a robot including a memory and a processor. The memory includes one or more instructions and the processor executes the one or more instructions. The processor is configured to initiate a conversation with the user. The conversation includes a machine-initiated conversation. The processor is configured to determine a flow of the conversation with the user by analyzing replies of the user. The processor determines and redirects the flow of the conversation with any of an input query or one or more parameters extracted from the input query, entity or sentiments, a location of the user, or environmental factors from the replies of the user. The processor is configured to generate the closed domain conversation by providing one or more contents related to at least one topic or category which is personalized based on a combination of one or more properties of the user's interest. The processor is configured to enable the user to interact with the robot using the closed domain conversation in real-time based on the user's interest. In some embodiments, the processor is configured to generate one or more opinion sets by analyzing the replies of the user during the conversation. The processor determines the user's interest based on at least one of the user preferences, likes, and dislikes of the user, wherein the one or more opinion sets enables the robot to update the user's interest and the user preferences. In some embodiments, the processor is configured to determine and redirect the flow of conversation with user sentiments by analyzing the replies of the user, and generate the closed domain conversation by filtering the conversation with the one or more contents. In some embodiments, the conversation includes one or more micro conversations pre-determined in the m