Intelligent assistant system and method thereof

Abstract

An intelligent assistant system includes a transceiver interface and a processing unit. The processing unit is electrically connected to the transceiver interface and is used to execute a knowledge model. The knowledge model includes a context-aware model and an event-driven module. The context-aware model is trained based on multiple workplace information items and multiple prompt words to find the corresponding workplace information items based on the prompt words. The knowledge model includes the following operations: receiving input information through the transceiver interface; determining whether the input information triggers at least one of the prompt words through the context-aware model; when the input information triggers at least one of the prompt words, finding the workplace information corresponding to at least one of the prompt words through the context-aware model; generating output information based on the workplace information through the event-driven module; and sending the output information through the transceiver interface.

Description

[Technical Field]

[0001] This disclosure relates to an intelligent assistant system and method thereof, and more particularly to an intelligent assistant system and method thereof for proactively resolving employee problems. [Previous Technology]

[0002] Currently, AI chatbots used by company employees are a technology specifically designed to support and optimize internal employee communication and operations. These AI chatbots utilize natural language processing (NLP) and artificial intelligence technologies to assist employees with technical issues, such as onboarding processes, leave applications, internal mobility, and training. However, the aforementioned AI chatbot applications are passive and lack initiative.

[0003] Therefore, how to provide a proactive assistant system is a goal that the industry urgently needs to strive for. [Summary of the Invention]

[0004] One object of this disclosure is to provide an intelligent assistant system. The intelligent assistant system includes a transceiver interface and a processing unit. The processing unit is electrically connected to the transceiver interface and is used to execute a knowledge model. The knowledge model includes a context-aware model and an event-driven module. The context-aware model is trained based on multiple workplace information items and multiple prompt words to find corresponding workplace information items based on these prompt words. These workplace information items include multiple case handling information items, multiple human resource information items, multiple enterprise management information items, multiple customer relationship information items, and multiple schedule information items. The knowledge model includes the following operations: receiving input information through the transceiver interface; determining whether the input information triggers at least one of the prompt words through the context-aware model; when the input information triggers at least one of the prompt words, finding workplace information corresponding to at least one of the prompt words through the context-aware model; generating output information based on the workplace information through the event-driven module; and sending the output information through the transceiver interface.

[0005] Another object of this disclosure is to provide a method. This method is used in an intelligent assistant system, which includes a transceiver interface and a processing unit electrically connected to the transceiver interface and used to execute a knowledge model. The knowledge model includes a context-aware model and an event-driven module. The context-aware model is trained based on multiple workplace information items and multiple prompt words to find corresponding workplace information items based on these prompt words. These workplace information items include multiple case handling information items, multiple human resource information items, multiple enterprise management information items, multiple customer relationship information items, and multiple schedule information items. The method includes: receiving input information; determining, through the context-aware model, whether the input information triggers at least one of the prompt words; when the input information triggers at least one of the prompt words, finding workplace information corresponding to at least one of the prompt words through the context-aware model; generating output information based on the workplace information; and sending the output information.

[0006] This disclosure provides an intelligent assistant system and method, whose context-aware model is trained based on multiple workplace information and multiple prompt words, enabling the context-aware model to proactively solve employee problems when it perceives various employee problems (i.e., when it perceives various situations). Therefore, the intelligent assistant system and method of this disclosure are proactive.

[0007] The following detailed technical and implementation methods of this disclosure are illustrated in conjunction with the figures, so that those skilled in the art to which this disclosure pertains can understand the technical features of the claimed invention.

Implementation Method

[0009] The following disclosure provides numerous different embodiments or examples for implementing various features of this disclosure. Elements and configurations in the specific examples are used in the following discussion to simplify this disclosure. Any examples discussed are for illustrative purposes only and do not in any way limit the scope or meaning of this disclosure or its examples. Where appropriate, the same reference numerals are used between figures and in corresponding text descriptions to represent the same or similar elements.

[0010] Please refer to Figure 1, which illustrates a block diagram of an intelligent assistant system 100 according to an embodiment of this disclosure. As shown in Figure 1, the intelligent assistant system 100 includes a transceiver interface 110, a processing unit 120, a case management unit 130, and a smart service desk dashboard 140. In some embodiments, the intelligent assistant system 100 may be a computer, a smartphone, a tablet computer, a data processing server, or any equivalent processing device.

[0011] Currently, AI chatbots used by companies in the workplace can assist employees with onboarding processes, leave applications, internal mobility, and training, among other workplace issues. However, these AI chatbot applications are passive and lack initiative. Therefore, the intelligent assistant system 100 of this disclosure is trained based on a plurality of workplace information W1 and a plurality of prompt words I1, enabling the intelligent assistant system 100 to proactively resolve employee problems when it perceives various employee issues (i.e., when it perceives various situations). Therefore, the intelligent assistant system 100 of this disclosure is proactive.

[0012] The specific contents and implementation details of the intelligent assistant system 100's send / receive interface 110, processing unit 120, case management unit 130, and smart service desk dashboard 140 will be described in detail later.

[0013] As shown in Figure 1, the processing unit 120 is electrically connected to the transceiver interface 110 and is used to execute the knowledge model 122. In some embodiments, the transceiver interface 110 may be a text interface or a video interface. The text interface may include a communication circuit and corresponding network communication software (e.g., Line) to exchange or send and receive signals with the server of the communication software such as Line; the video interface may include a communication circuit and corresponding network video software (e.g., Zoom, Microsoft Teams, Google Meet, Skype, or Webex) to conduct video conversations through the image of a virtual assistant; the video interface may also include a communication circuit and corresponding network voice software (e.g., ASR or TTS model) to conduct voice conversations through the ASR or TTS model. In some embodiments, the processing unit 120 may be a central processing unit (CPU), a tensor processing unit (TPU), a microcontroller unit (MCU), an application-specific integrated circuit (ASIC), or any equivalent processing circuit. In some embodiments, the knowledge model 122 may be an artificial intelligence model (AI model).

[0014] As shown in Figure 1, the knowledge model 122 includes a context-aware model 122A and an event-driven module 122B. The event-driven module 122B is coupled to the context-aware model 122A. The context-aware model 122A is trained based on a plurality of workplace information W1 and a plurality of cue words I1 to find (i.e. distinguish) the corresponding workplace information W1 based on the cue words I1. In one embodiment, the workplace information W1 may include a plurality of case handling information W11 (e.g., software case handling information, hardware case handling information, system case handling information, company policy or company procedure case handling information, and employee personal issue case handling information), a plurality of human resources information W12 (e.g., employee profiles, job titles and responsibilities provided by a human resources system (HRMS), a plurality of enterprise management information W13 (e.g., internal resource management information on enterprise finance, logistics and production data provided by an enterprise resource planning system (ERP), a plurality of customer relationship information W14 (e.g., customer information and historical information on customer interactions provided by a customer relationship management system (CRM), and a plurality of schedule information W15 (e.g., employee schedules and meeting plans provided by a self-calendar system). In one embodiment, the prompt word I1 may include software case prompt word I11, hardware case prompt word I12, system case prompt word I13, company policy or company procedure case prompt word I14, employee personal issue case prompt word I15, human resources information prompt word I16, enterprise management information prompt word I17, customer relationship information prompt word I18, and schedule information prompt word I19. For example, if relevant prompt words and case handling information for "Word, PowerPoint, and Outlook" are input into the context-aware model 122A for training, the trained context-aware model 122A can identify and find the corresponding Word, PowerPoint, and Outlook software case handling information (i.e., case handling information W11 of workplace information W1) for the prompt word "Word, PowerPoint, and Outlook" (i.e., software case prompt word I11).

[0015] In one embodiment, as shown in Figure 1, the context-aware model 122A is further trained based on a plurality of action codes AC corresponding to input information IM, which are mapped to a plurality of management information systems (MIS) 150A-150N. In one embodiment, the input information IM may include software problem IM1, hardware problem IM2, system problem IM3, company policy problem IM4, company procedure problem IM5, and employee personal information problem IM6. In one embodiment, the action code AC may be a binary number or a decimal number. In one embodiment, the management information systems 150A-150N may be the company's software maintenance center, hardware maintenance center, and system maintenance center, or the company's employee customer service system. For details regarding the specific functional operation and implementation of the input information IM, action codes AC, and management information systems 150A-150N, please refer to the detailed description of method 200 in Figure 2 below.

[0016] In one embodiment, as shown in Figure 1, the case management unit 130 is electrically connected to the processing unit 120 and is used to store the case processing information W11 from the context-aware model 122A. The smart service desk dashboard 140 is electrically connected to the case management unit 130 and is used to generate a plurality of case analysis information S1 based on the case processing information W11. For example, if the case management unit 130 receives and stores historical information about cases handled by company staff where employees cannot open Outlook (i.e., case processing information W11), the smart service desk dashboard 140 analyzes and displays the current processing status information (i.e., case analysis information S1) of the cases handled by the staff based on the historical case information.

[0017] In one embodiment, as shown in Figure 1, the smart service desk dashboard 140 also transmits the case analysis information S1 to the context-aware model 122A, so that the context-aware model 122A can be trained based on the case analysis information S1. For example, if the smart service desk dashboard 140 analyzes and displays the current processing status information (i.e., case analysis information S1) of an employee being unable to open Outlook, the smart service desk dashboard 140 sends the current processing status information back to the context-aware model 122A for training.

[0018] Please refer to Figures 1-2. Figure 2 shows a flowchart of the method 200 for operating the intelligent assistant system 100 in Figure 1.

[0019] First, as shown in Figures 1-2, step S210 is executed, and the knowledge model 122 receives input information IM through the transceiver interface 110. For example, if a user (e.g., an employee) has a technical problem with Outlook operation (i.e., software problem IM1), they can input the technical problem to the knowledge model 122 through the transceiver interface 110, so that the knowledge model 122 receives the software problem IM1 through the transceiver interface 110.

[0020] Next, as shown in Figures 1-2, step S220 is executed. The knowledge model 122 determines, through the context-aware model 122A, whether the input information IM triggers at least one of the prompt words I1. If yes, proceed to step S230; if no, proceed to step S221. For example, if the knowledge model 122 receives a technical problem (i.e., hardware problem IM2) related to an employee's computer malfunction, then through the context-aware model 122A, it determines the software case prompt word I11 that triggers the prompt word I1 for the aforementioned technical problem.

[0021] When the input information IM triggers at least one of the prompt words I1, as shown in Figures 1-2, step S230 is executed. The knowledge model 122 uses the context-aware model 122A to find the workplace information W1 corresponding to at least one of the prompt words I1. For example, when the knowledge model 122 receives a technical problem (i.e., system problem IM3) where an employee forgets their password and cannot log in to the company system, and this technical problem triggers the system case prompt word I13, the context-aware model 122A can find the password information for logging into the company system (i.e., find the case processing information W11 corresponding to the system case prompt word I13).

[0022] Next, as shown in Figures 1-2, step S240 is executed, whereby the knowledge model 122 generates output information OM based on workplace information W1 through the event-driven module 122B. Specifically, in step S240, the knowledge model 122 analyzes workplace information W1 to generate proactive reminder information (e.g., reminder messages based on employee meeting schedules), personalized information (e.g., customized suggestions or decision-making suggestions based on employee job level and responsibilities), contextualized notification information (e.g., notification messages issued based on the employee's actual personal situation), or analytical solution suggestion information (e.g., feasible solution information analyzed based on the employee's actual personal situation), and generates output information OM based on proactive reminder information, personalized information, contextualized notification information, or analytical solution suggestion information. For example, if the event-driven module 122B receives the schedule information of an employee meeting (i.e., the schedule information W15 of workplace information W1), the event-driven module 122B will proactively remind the employee of the upcoming meeting based on the aforementioned employee meeting schedule information (i.e., proactive reminder information), and will provide the proactive reminder information and the relevant materials required for the meeting (i.e., output information OM) to the employee.

[0023] When the input information IM does not trigger at least one of the prompt words I1, as shown in Figures 1-2, step S221 is executed. The knowledge model 122, through the context-aware model 122A, calls the management information systems 150A-150N mapped by the action code AC corresponding to the input information IM. For example, if the action code AC corresponding to system problem IM3 is 10, and the management information system 150A mapped by the action code AC is the company's system maintenance center, then when the knowledge model 122 receives a technical problem (i.e., system problem IM3) where an employee cannot log in to the company system and this technical problem does not trigger the prompt word I1, the context-aware model 122A can call the company's system maintenance center according to the action code 10 corresponding to the above-mentioned technical problem.

[0024] Next, as shown in Figures 1-2, step S222 is executed. The knowledge model 122 transmits the input information IM to the management information systems 150A-150N through the context-aware model 122A, so that the management information systems 150A-150N generate output information OM based on the input information IM. For example, when the knowledge model 122 receives a technical problem (i.e., system problem IM3) that an employee cannot log in to the company system, and the context-aware model 122A calls the management information system 150A (e.g., the company's system maintenance center) mapped to the action code 10 according to the action code (e.g., 10) corresponding to the above technical problem, the context-aware model 122A transmits the above technical problem to the system maintenance center, so that the system maintenance center replies with the solution information for the inability to log in to the company system (i.e., output information OM) based on the above technical problem.

[0025] Finally, as shown in Figures 1-2, step S250 is executed, and the knowledge model 122 sends the output information OM through the transceiver interface 110. For example, if the management information system 150A (e.g., the company's system maintenance center) replies with a solution message (i.e., output information OM) indicating that the user cannot log in to the company system, the knowledge model 122 sends this solution message to the employee through a communication software such as Line (i.e., the transceiver interface 110).

[0026] It is worth noting that, in one embodiment of step S250, the knowledge model 122 may also transmit the output information OM to the context-aware model 122A, so that the context-aware model 122A can be trained based on the output information OM.

[0027] Please refer to Figures 3A-3B. Figure 3A shows the actual operation flowchart of the intelligent assistant system 100 when the input information IM is a software problem, and Figure 3B shows the actual operation flowchart of the intelligent assistant system 100 when the input information IM is a hardware problem.

[0028] As shown in Figure 3A, the knowledge model 122 of the intelligent assistant system 100 receives a software problem from the user 160 (e.g., a company employee) through the send / receive interface 110: "My Outlook cannot send or receive emails. What should I do?". When the above software problem triggers the software case prompt word I11 (e.g., Outlook), the context-aware model 122A finds the Outlook software case handling information (i.e., case handling information W11) corresponding to the software case prompt word "outlook", and the event-driven module 122B replies (i.e. generates) the solution information "Because your Outlook settings are incorrect, you cannot send or receive emails correctly. Therefore, it is recommended that you set up Outlook in the following way" (i.e., output information OM), so as to send the above solution information to the user 160 through the send / receive interface 110.

[0029] As shown in Figure 3B, the knowledge model 122 of the intelligent assistant system 100 receives the hardware problem "My laptop is malfunctioning and cannot be turned on" from the user 160 through the transceiver interface 110. When the hardware problem does not trigger the hardware case prompt word I12 of the prompt word I1, the context-aware model 122A calls the hardware maintenance center (e.g., management information system 150B) mapped by the action code AC (e.g., 11) corresponding to the hardware problem, and transmits the hardware problem to the hardware maintenance center. The hardware maintenance center then replies (i.e. generates) the solution information "Someone has been dispatched for repair" (i.e., output information OM) to the knowledge model 122, and the knowledge model 122 sends the solution information to the user 160 through the transceiver interface 110.

[0030] It is worth noting that, in the embodiment shown in Figure 3B, the smart service desk dashboard 140 also displays case processing status information CS for hardware or system problems handled by the management information system 150B, and generates case alert information CW based on the case processing status information CS. Finally, the smart service desk dashboard 140 transmits the case processing status information CS to the context-aware model 122A, enabling the context-aware model 122A to be trained based on the case processing status information CS. Therefore, the user 160 (e.g., a company staff member on duty) can learn about the current case status of hardware or system problems handled by the management information system 150B through the case alert information CW.

[0031] Please refer to Figure 4, which illustrates the actual operation flowchart of the intelligent assistant system 100 when the input information IM is a company policy issue or a company procedure issue.

[0032] As shown in Figure 4, the knowledge model 122 of the intelligent assistant system 100 receives the company procedure question "How do I reimburse business travel expenses?" from the user 160 through the send / receive interface 110. When the above company procedure question triggers the company procedure case prompt I14 (e.g., business travel expense reimbursement), the context-aware model 122A finds the business travel expense reimbursement processing information (i.e., case processing information W11) corresponding to the company procedure case prompt "business travel expense reimbursement", and the event-driven module 122B replies (i.e. generates) the resolution information "According to Chapter 3.1 of the company procedure specification, you need to first click on the website http: / / twm / portal-web to download the business travel expense reimbursement form, fill in the business travel expense reimbursement form to apply for reimbursement" and the business travel expense reimbursement form in PDF format (i.e., output information OM), and sends the above resolution information and business travel expense reimbursement form to the user 160 through the send / receive interface 110.

[0033] Please refer to Figure 5, which illustrates the actual operation flowchart of the intelligent assistant system 100 when the input information IM is an employee's personal information question.

[0034] As shown in Figure 5, the knowledge model 122 of the intelligent assistant system 100 receives the employee personal information question "How many days of special leave do I have left this year? If there are any remaining special leave days, please help me apply for leave on Friday" from the user 160 through the send / receive interface 110. When the above employee personal information question does not trigger the employee personal information question prompt word I15, the context-aware model 122A calls the employee customer service system (e.g., management information system 150N) mapped by the action code AC (e.g., 13) corresponding to the above employee personal information question, and transmits the above employee personal information question to the employee customer service system. The employee customer service system then replies (i.e. generates) the solution information "User 160 has 7 days of special leave left this year and has already applied for leave on Friday" (i.e., output information OM) to the knowledge model 122, so that the knowledge model 122 sends the above solution information to the user 160 through the send / receive interface 110.

[0035] Therefore, through the above embodiments, the context awareness model 122A of the intelligent assistant system 100 of this disclosure is trained based on a plurality of workplace information W1 and a plurality of prompt words I1, so that the context awareness model 122A has the ability to proactively solve employee problems when it perceives various employee problems (i.e., when it perceives various situations). Therefore, the intelligent assistant system 100 and its method 200 of this disclosure are proactive.

[0036] Although this disclosure has been described above with reference to embodiments, it is not intended to limit this disclosure. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of this disclosure. Therefore, the scope of protection of this disclosure shall be determined by the appended claims. [Simplified Explanation of the Diagram]

[0008] To make the above and other objects, features and embodiments of this disclosure more apparent and understandable, the accompanying drawings are described as follows: Figure 1 is a block diagram of an intelligent assistant system according to an embodiment of this disclosure; Figure 2 is a flowchart of the method of operating the intelligent assistant system of Figure 1; Figure 3A is a flowchart of the actual operation of the intelligent assistant system when the input information is a software problem; Figure 3B is a flowchart of the actual operation of the intelligent assistant system when the input information is a hardware problem; Figure 4 is a flowchart of the actual operation of the intelligent assistant system when the input information is a company policy problem or a company procedure problem; Figure 5 is a flowchart of the actual operation of the intelligent assistant system when the input information is an employee personal information problem. [Biomaterial Storage]

[0038] Domestic storage information (please note in order of storage institution, date, and number): None. International storage information (please note in order of storage country, institution, date, and number): None.

Claims

1. An intelligent assistant system, comprising: a transceiver interface; and a processing unit electrically connected to the transceiver interface and configured to execute a knowledge model, the knowledge model including a context-aware model and an event-driven module, the event-driven module being coupled to the context-aware model, the context-aware model being trained based on a plurality of workplace information, a plurality of prompt words, and a plurality of action codes to find corresponding workplace information based on the prompt words, the workplace information including a plurality of case handling information, a plurality of human resource information, a plurality of enterprise management information, a plurality of customer relationship information, and a plurality of schedule information, and the action codes being mapped to a plurality of management information systems, wherein the knowledge model includes the following operations: receiving input information through the transceiver interface; determining, through the context-aware model, whether the input information triggers at least one of the prompt words; when the input information triggers the at least one of the prompt words, the knowledge model includes the following operations: The context-aware model identifies workplace information corresponding to at least one of the prompt words; and the event-driven module generates output information based on the workplace information. When the input information does not trigger at least one of the prompt words, the knowledge model includes the following operations: The context-aware model calls the management information system mapped to the action code corresponding to the input information; and the context-aware model transmits the input information to the management information system, causing the management information system to generate the output information based on the input information; and the output information is sent through the transceiver interface. A case management unit is electrically connected to the processing unit and is used to store case processing information from the context-aware model; and a smart service desk dashboard is electrically connected to the case management unit and includes the following operations: Displaying case processing status information of a management information system handling a software problem, a hardware problem, or a system problem; and transmitting the case processing status information to the context-aware model, causing the context-aware model to be trained based on the case processing status information.

2. The intelligent assistant system as described in claim 1, wherein generating the output information based on the workplace information through the event-driven module includes the following operations: analyzing the workplace information to generate a proactive reminder, a personalized message, a contextual notification, or an analysis solution suggestion; and generating the output information based on the proactive reminder, the personalized message, the contextual notification, or the analysis solution suggestion.

3. The intelligent assistant system as described in claim 1, the intelligent service desk dashboard further includes the following operation: generating multiple case analysis information based on the case processing information.

4. The intelligent assistant system as described in claim 3, wherein the smart service desk dashboard also transmits the case analysis information to the context-aware model, so that the context-aware model can be trained based on the case analysis information.

5. The intelligent assistant system as described in claim 1, wherein when the input information is a software problem, a hardware problem, or a system problem, the knowledge model includes the following operations: When a software problem, hardware problem, or system problem triggers a software case prompt, hardware case prompt, or system case prompt, the context-aware model finds the corresponding case handling information for that prompt and generates output information based on the case handling information through the event-driven module. Conversely, when a software problem, hardware problem, or system problem does not trigger a software case prompt, hardware case prompt, or system case prompt, the context-aware model calls the management information system mapped to the action code corresponding to the software problem, hardware problem, or system problem, and transmits the software problem, hardware problem, or system problem to the management information system, causing the management information system to generate output information for the knowledge model based on the software problem, hardware problem, or system problem.

6. The intelligent assistant system as described in claim 5, wherein the intelligent service desk dashboard further includes the following operation: generating a case alert based on the case processing status information.

7. The intelligent assistant system as described in claim 1, wherein when the input information is a company policy issue or a company procedure issue, the knowledge model includes the following operations: when the company policy issue or company procedure issue triggers a company policy or company procedure case prompt word, the context-aware model finds the case handling information corresponding to the company policy or company procedure case prompt word, and the event-driven module generates the output information based on the case handling information.

8. The intelligent assistant system as described in claim 1, wherein when the input information is an employee personal information question, the knowledge model further includes the following operation: when the employee personal information question does not trigger the prompt words of an employee personal question case prompt word, the context-aware model calls the management information system mapped by the action code corresponding to the employee personal information question, and transmits the employee personal information question to the management information system, so that the management information system generates the output information to the knowledge model based on the employee personal information question.

9. A method for an intelligent assistant system, the intelligent assistant system comprising a transceiver interface and a processing unit electrically connected to the transceiver interface and a case management unit, and a smart service dashboard electrically connected to the case management unit, the processing unit being configured to execute a knowledge model comprising a context-aware model and an event-driven module coupled to the context-aware model, the context-aware model being trained based on a plurality of workplace information, a plurality of prompt words, and a plurality of action codes to find corresponding workplace information based on the prompt words, the workplace information including a plurality of case handling information, a plurality of human resources information, a plurality of enterprise management information, a plurality of customer relationship information, and a plurality of schedule information, and the action codes being mapped to a plurality of management information systems, the method comprising: receiving input information; determining, through the context-aware model, whether the input information triggers at least one of the prompt words; when the input information triggers the at least one of the prompt words, the method comprising: The method involves: using the context-aware model to find workplace information corresponding to at least one of the prompt words; and generating output information based on the workplace information; when the input information does not trigger at least one of the prompt words, the method includes: using the context-aware model to call the management information system mapped to the action code corresponding to the input information; and using the context-aware model to transmit the input information to the management information system, causing the management information system to generate the output information based on the input information; sending the output information; storing case processing information from the context-aware model in the case management unit; displaying case processing status information of a management information system handling a software problem, a hardware problem, or a system problem through the smart service desk dashboard; and transmitting the case processing status information to the context-aware model through the smart service desk dashboard, causing the context-aware model to be trained based on the case processing status information.

Images