Human-computer interaction processing method, device, storage medium and electronic device

Abstract

The present invention discloses a human-computer interaction processing method, device, storage medium and electronic device. The method comprises: when an interaction request of a target object is detected, obtaining the initial interaction information of the target object, wherein the initial interaction information is obtained based on the summary of historical industry issues; if there is an interaction application scenario associated with the initial interaction information, executing the human-computer interaction rules corresponding to the interaction application scenario; if there is no interaction application scenario associated with the initial interaction information, executing the scene selection rules; if the scene selection rules locate the interaction application scenario, executing the human-computer interaction rules corresponding to the interaction application scenario. The present invention solves the technical problems of inaccurate information push and poor user satisfaction caused by the inability of the human-computer interaction processing method in the prior art to accurately infer customer intentions.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, in particular to the field of human-computer question answering and human-computer interaction technology, and specifically to a human-computer interaction processing method, device, storage medium and electronic equipment. Background Art

[0002] With the continuous development of artificial intelligence technology, human-computer question-answering and human-computer interaction technologies have been widely used in the field of customer service. The human-computer interaction process in the existing technology is mainly to determine the customer's initial intention by understanding the customer's language, and select appropriate scene execution rules according to the customer's initial intention, and interact with the customer under this scene execution rule. However, the above method does not perform the process of customer intention reasoning, and the accuracy of customer intention guessing is low, which is easy to cause misjudgment of customer intention, and then cause inaccurate information push, affecting customer satisfaction.

[0003] In addition, the configuration methods of scenario execution rules in the prior art, such as expert system method, triple definition method, IF-THEN definition method, etc., have a relatively complex configuration process and are difficult to configure for business personnel with weak professional knowledge.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] The embodiments of the present invention provide a human-computer interaction processing method, device, storage medium and electronic device to at least solve the technical problems of inaccurate information push and poor user satisfaction caused by the inability of the human-computer interaction processing method in the prior art to accurately infer customer intentions.

[0006] According to one aspect of an embodiment of the present invention, a human-computer interaction processing method is provided, comprising: when an interaction request of a target object is detected, obtaining initial interaction information of the target object, wherein the initial interaction information is obtained based on a summary of historical industry issues; if there is an interaction application scenario associated with the initial interaction information, executing a human-computer interaction rule corresponding to the interaction application scenario; if there is no interaction application scenario associated with the initial interaction information, executing a scenario selection rule, wherein the scenario selection rule is used to locate the interaction application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; if the scenario selection rule locates the interaction application scenario, executing the human-computer interaction rule corresponding to the interaction application scenario.

[0007] Optionally, the method also includes: if the scene selection rule cannot locate the interactive application scenario, determining whether the initial interaction information is mapped to the historical interaction scenario; if the initial interaction information has been mapped to the historical interaction scenario, establishing an association relationship between the initial interaction information and the historical interaction scenario, and determining the corresponding human-computer interaction rule based on the association relationship; if the initial interaction information is not mapped to the historical interaction scenario, instructing the staff to use the human-computer interaction system to configure new human-computer interaction rules for the initial interaction information.

[0008] Optionally, the above-mentioned execution of the human-computer interaction rules corresponding to the above-mentioned interactive application scenarios includes: determining the scenario processing elements from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenarios; combining the above-mentioned scenario processing elements and the element values ​​corresponding to the above-mentioned scenario processing elements to obtain a state element set; searching for the above-mentioned human-computer interaction rules that meet the element combination conditions according to the above-mentioned state element set; and executing the predicate actions in the above-mentioned human-computer interaction rules.

[0009] Optionally, the above-mentioned execution scenario selection rule includes: collecting and organizing all historical conversation records to obtain a total conversation; combining all intentions in the above-mentioned total conversation in pairs to obtain an intention set; using the association rules in the above-mentioned scenario selection rules to generate reasoning relationships based on the above-mentioned intention set, wherein the above-mentioned association rules are used to derive the occurrence probability between the above-mentioned intentions and the above-mentioned interactive application scenarios based on the relationship between the respective intentions in the above-mentioned intention set; when selecting a scenario candidate set, determining the probability of occurrence of a second interactive intention related to the above-mentioned first interactive intention after the first interactive intention appears based on the above-mentioned reasoning relationship; combining the above-mentioned occurrence probability with user feature information, calculating the feature value of each to-be-selected scene in the above-mentioned scenario candidate set, and arranging them in reverse order according to the feature value to obtain a sorting result; and selecting the above-mentioned interactive application scenario from the above-mentioned scenario candidate set according to the above-mentioned sorting result.

[0010] According to another aspect of an embodiment of the present invention, a human-computer interaction processing device is also provided, including: an acquisition module, used to acquire initial interaction information of the target object when an interaction request of the target object is detected, wherein the initial interaction information is obtained based on a summary of historical industry issues; a first execution module, used to execute human-computer interaction rules corresponding to the interaction application scenario if there is an interaction application scenario associated with the initial interaction information; a second execution module, used to execute scenario selection rules if there is no interaction application scenario associated with the initial interaction information, wherein the scenario selection rules are used to locate the interaction application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; a third execution module, used to execute the human-computer interaction rules corresponding to the interaction application scenario if the scenario selection rules locate the interaction application scenario.

[0011] Optionally, the above-mentioned device also includes: a judgment module, which is used to judge whether the above-mentioned initial interaction information is mapped to the historical interaction scene if the above-mentioned scene selection rule cannot locate the above-mentioned interactive application scene; a first determination module, which is used to establish an association relationship between the above-mentioned initial interaction information and the above-mentioned historical interaction scene if the above-mentioned initial interaction information has been mapped to the above-mentioned historical interaction scene, and determine the corresponding above-mentioned human-computer interaction rule based on the above-mentioned association relationship; an indication module, which is used to instruct the staff to use the human-computer interaction system to configure new human-computer interaction rules for the above-mentioned initial interaction information if the above-mentioned initial interaction information is not mapped to the above-mentioned historical interaction scene.

[0012] Optionally, the above-mentioned first execution module includes: a second determination module, used to determine the scenario processing elements from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenario; a first acquisition sub-module, used to combine the above-mentioned scenario processing elements and the element values ​​corresponding to the above-mentioned scenario processing elements to obtain a state element set; a search module, used to search for the above-mentioned human-computer interaction rules that meet the element combination conditions according to the above-mentioned state element set; and a first execution sub-module, used to execute the predicate action in the above-mentioned human-computer interaction rules.

[0013] Optionally, the second execution module includes: a second acquisition submodule, used to collect and organize all historical conversation records to obtain a total conversation; a third acquisition submodule, used to combine all intentions in the total conversation in pairs to obtain an intention set; a generation module, used to adopt the association rules in the scene selection rules to generate an inference relationship based on the intention set, wherein the association rules are used to derive the occurrence probability between the intention and the interactive application scenario based on the relationship between the intentions in the intention set; a third determination module, used to determine the probability of occurrence of a second interaction intention related to the first interaction intention after the first interaction intention appears when selecting a scene candidate set based on the inference relationship; a calculation module, used to combine the occurrence probability with user feature information, calculate the feature value of each scene to be selected in the scene candidate set, and arrange them in reverse order according to the feature value to obtain a sorting result; a selection module, used to select the interactive application scenario from the scene candidate set according to the sorting result.

[0014] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing any one of the above-mentioned human-computer interaction processing methods.

[0015] According to another aspect of an embodiment of the present invention, there is further provided an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the above-mentioned human-computer interaction processing methods.

[0016] In an embodiment of the present invention, a human-computer interaction processing method is adopted. When an interaction request of the target object is detected, initial interaction information of the target object is obtained, wherein the initial interaction information is obtained based on a summary of historical industry issues. If there is an interaction application scenario associated with the initial interaction information, the human-computer interaction rules corresponding to the interaction application scenario are executed. If there is no interaction application scenario associated with the initial interaction information, the scenario selection rules are executed, wherein the scenario selection rules are used to locate the interaction application scenario to be executed according to the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues. If the scenario selection rules locate the interaction application scenario, the human-computer interaction rules corresponding to the interaction application scenario are executed, thereby achieving the purpose of inferring customer intent according to the interaction request of the target object and determining the human-computer interaction processing method based on the inference result, thereby achieving the technical effect of improving the accuracy of customer intent inference and user satisfaction, and further solving the technical problem of inaccurate information push and poor user satisfaction caused by the inability of the human-computer interaction processing method in the prior art to accurately infer customer intent. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1a It is a schematic diagram of a human-computer interaction display interface according to the prior art;

[0019] Figure 1b is a schematic diagram of another human-computer interaction display interface according to the prior art;

[0020] Figure 2 is a flow chart of a human-computer interaction processing method according to an embodiment of the present invention;

[0021] Figure 3 is a schematic diagram of an optional human-computer interaction display interface according to an embodiment of the present invention;

[0022] Figure 4 is a schematic diagram of another optional human-computer interaction display interface according to an embodiment of the present invention;

[0023] Figure 5 is a flow chart of an optional human-computer interaction processing method according to an embodiment of the present invention;

[0024] Figure 6 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention;

[0025] Figure 7 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention;

[0026] Figure 8 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention;

[0027] Fig. 9 It is a structural schematic diagram of a human-computer interaction processing device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] Example 1

[0031] There are two common communication methods between artificial intelligence robots and users: one is question-and-answer, where the user gives a query with clear semantics and complete information, and the robot gives an accurate answer; the other is topic-based, where the user communicates with the robot multiple times. The information contained in each query is not complete, but there is a connection between them. The robot needs to merge and understand multiple queries, which is also called context understanding. Its characteristics are that the intelligent robot and the user will communicate multiple times. During the multiple human-machine communications, the machine first confirms the topic content of the user relationship. We abstract the topic content and design a dialogue scene (abbreviated as: scene). Once a scene is determined, the system will reason and answer the user based on the rules corresponding to the scene and the facts in the knowledge base. The entire interaction process is specifically divided into: 1) Semantic understanding: Using natural language understanding technology, the user's natural form of questions is mapped to a preliminary intention; 2) Scenario selection: Based on the preliminary intention, the system automatically analyzes the user's possible true intention and confirms a scenario to communicate with the user; 3) Scenario execution: After entering the scene, interact with the user according to the designed dialogue logic. During the interaction, the user can select by clicking or directly reply to the text content.

[0032] Figure 1a-1b is a schematic diagram of a human-computer interaction display interface according to the prior art, such as Figure 1a-1b As shown in the figure, the existing human-computer interaction methods generally have two defects in the multi-round dialogue process:

[0033] 1) Lack of scene selection. Existing human-computer interaction technologies often design scenes directly based on the initial intentions after semantically understanding the customer's initial intentions, without automatically reasoning to see if the system has a suitable scene to interact with the customer. In fact, in the actual communication process between people, both parties will guess each other's true intentions, which is the "scene selection" mentioned above.

[0034] For example, in the financial industry, a customer says, “My card is broken, what do I do?”

[0035] The initial intention of semantic understanding is: how to deal with card (bank card) damage.

[0036] The system knowledge base already has scenarios: bank card replacement, card loss and settlement method change, customer emergency handling, and customer bank card loss application. If the "scenario selection" step is missing, the customer's initial intention cannot be automatically mapped to the existing scenario, that is, the system cannot actively guess the customer's true intention. There are only two ways to solve this problem: a. Manually specify or design a new scenario, and b. Give options for customers to choose: For a, because the customer's initial intention is very large, the workload of manual specification is very large, and once the business changes, the initial intention needs to be sorted out and re-specified; for b, the interaction process will appear too redundant. Generally, it is best not to exceed 5 sessions in one round of interaction to solve the customer's problem. Saving the number of interactions as much as possible is an important indicator of the quality of the system. After the "scenario selection", the system will automatically select an optimal scenario to interact with the customer based on indicators such as the relevance of the intention and the user's characteristic value.

[0037] 2) The configuration of scenario execution rules is too complex. For example, expert systems use triple definitions or IF-THEN definitions, which are difficult for business personnel without professional knowledge to configure.

[0038] Based on the above problems, an embodiment of the present invention provides a method embodiment of human-computer interaction processing. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0039] In an embodiment of the present invention, a human-computer interaction processing method is adopted. When an interaction request of the target object is detected, initial interaction information of the target object is obtained, wherein the initial interaction information is obtained based on a summary of historical industry issues. If there is an interaction application scenario associated with the initial interaction information, the human-computer interaction rules corresponding to the interaction application scenario are executed. If there is no interaction application scenario associated with the initial interaction information, the scenario selection rules are executed, wherein the scenario selection rules are used to locate the interaction application scenario to be executed according to the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues. If the scenario selection rules locate the interaction application scenario, the human-computer interaction rules corresponding to the interaction application scenario are executed, thereby achieving the purpose of inferring customer intent according to the interaction request of the target object and determining the human-computer interaction processing method based on the inference result, thereby achieving the technical effect of improving the accuracy of customer intent inference and user satisfaction, and further solving the technical problem of inaccurate information push and poor user satisfaction caused by the inability of the human-computer interaction processing method in the prior art to accurately infer customer intent.

[0040] Figure 2 is a flow chart of a human-computer interaction processing method according to an embodiment of the present invention. Figure 2 As shown, the method comprises the following steps:

[0041] Step S102, when an interaction request of a target object is detected, initial interaction information of the target object is obtained;

[0042] Step S104: if there is an interactive application scenario associated with the initial interactive information, executing a human-computer interaction rule corresponding to the interactive application scenario;

[0043] Step S106: if there is no interactive application scenario associated with the initial interactive information, executing a scenario selection rule;

[0044] Step S108: If the above scenario selection rule locates the above interactive application scenario, the human-computer interaction rule corresponding to the above interactive application scenario is executed.

[0045] Optionally, the above initial interaction information is obtained based on a summary of historical industry issues, and the above initial exchange information includes the initial intention of the target object.

[0046] Optionally, the above scenario selection rules are used to locate the interactive application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogue.

[0047] Optionally, when an interactive application scenario associated with the initial interactive information of the target object is detected, the human-computer interaction rules corresponding to the interactive application scenario are directly called and sent to the terminal device of the target object, such as Figure 3 As shown, when the initial interaction information of the target object is "upgrade 5G package", there is an interactive application scenario associated with "upgrade 5G package", and the human-computer interaction rules are determined to push relevant content for the 5G package, for example, 5G single mobile phone enjoyment package (mobile phone), 5G family enjoyment package (mobile phone + broadband), 5G upgrade package, and so on.

[0048] Optionally, when it is detected that there is no interactive application scenario associated with the initial interactive information of the target object, the scenario selection rule is called and sent to the terminal device of the target object, such as Figure 4 As shown, when the initial interaction information of the target object is "The mobile phone package fee is too high this month", there is no interaction scenario associated with "The mobile phone package fee is too high this month", and the scenario selection rule-related connection is sent to the terminal device of the target object.

[0049] As an optional embodiment, Figure 5 is a flowchart of an optional human-computer interaction processing method according to an embodiment of the present invention. Figure 5 As shown, the above method also includes:

[0050] Step S202: if the scene selection rule cannot locate the interactive application scene, determine whether the initial interaction information is mapped to a historical interaction scene;

[0051] Step S204, if the initial interaction information has been mapped to the historical interaction scene, establishing an association relationship between the initial interaction information and the historical interaction scene, and determining the corresponding human-computer interaction rule based on the association relationship;

[0052] Step S206: If the initial interaction information is not mapped to the historical interaction scenario, the staff is instructed to use the human-computer interaction system to configure new human-computer interaction rules for the initial interaction information.

[0053] Optionally, when the above-mentioned scene selection rules cannot locate the above-mentioned interactive application scenarios, the business personnel determine whether the above-mentioned initial interaction information is mapped to the historical interaction scenarios. If the above-mentioned initial interaction information can be mapped to the historical interaction scenarios, the business personnel manually establishes the association relationship between the above-mentioned initial interaction information and the above-mentioned historical interaction scenarios through the interactive interface, and determines the corresponding above-mentioned human-computer interaction rules based on the above-mentioned association relationship.

[0054] Optionally, when the above initial interaction information is not mapped to the above historical interaction scenario, the staff uses the human-computer interaction system to configure new human-computer interaction rules for the above initial interaction information through an interactive interface.

[0055] As an optional embodiment, Figure 6 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention. Figure 6 As shown, the above-mentioned execution of human-computer interaction rules corresponding to the above-mentioned interactive application scenarios includes:

[0056] Step S302, determining scenario processing elements from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenario;

[0057] Step S304, combining the scene processing elements and the element values ​​corresponding to the scene processing elements to obtain a state element set;

[0058] Step S306, searching the above human-computer interaction rules that meet the element combination conditions according to the above state element set;

[0059] Step S308: executing the predicate action in the above human-computer interaction rule.

[0060] Optionally, when the system automatically replies to customers, it needs to infer specific answers based on certain conditions (such as customer information, previous consultation information, system operation information, etc.) and reply to the target object. Because these conditions are bound to specific scenarios, we call them scenario elements. For example, the scenario elements in the "What discount packages are there" scenario include: package type (prepaid package, postpaid package), package name, package level, customer location, customer type, etc. After collecting the scenario elements and the element values ​​corresponding to the scenario elements, a state element set is obtained.

[0061] Optionally, searching for the above-mentioned human-computer interaction rules that meet the element combination conditions according to the above-mentioned state element set includes: judging whether the current human-computer interaction rules meet the element combination conditions; if the judgment result is yes, determining that the current human-computer interaction rules are the above-mentioned human-computer interaction rules that meet the element combination conditions; otherwise, continuing to judge whether the human-computer interaction rules meet the element combination conditions until the above-mentioned human-computer interaction rules that meet the element combination conditions are obtained.

[0062] Optionally, after obtaining the above human-computer interaction rules that meet the element combination conditions, the predicate action in the above human-computer interaction rules is executed, wherein the above predicate action corresponds to the specific operation to be performed by the system, that is, the system starts the above human-computer interaction rules to perform the corresponding specific operation. The above specific operations may include but are not limited to: replying a text to the customer, replying a menu to the customer, playing a recording, sending a text message, etc. Some examples of action predicates are as follows:

[0063] Prompt the customer (<prompt content>, <format>): The system displays the <prompt content> in the format specified by <format>; <format> includes: text, tts broadcast,

[0064] For example: Prompt the customer ("Hello, you currently owe 58 yuan.", <text>);

[0065] Ask the customer (<inquiry content>, <format>): the system displays the <inquiry content> in the format specified by <format>; <format> includes: text, tts broadcast;

[0066] Menu inquiry customer (<menu option content>, <interaction template>): The system fills <menu option content> into <interaction template> in the form of a menu and displays it to the customer. The customer interacts by clicking or entering a serial number.

[0067] For example: The menu asks the customer ("CCB Card||BOCOM Card||ICBC Card", "Please enter your bank card type: <?>");

[0068] Transfer to manual (<Reply Script>): The system displays <Reply Script> to prompt the user to click to transfer to manual;

[0069] For example: Transfer to manual service ("You can enter manual service by clicking on the customer service MM!");

[0070] Is it a telecom number (<mobile phone number>): The system determines whether <mobile phone number> is a telecom number. If not, it returns "0";

[0071] Find the customer's current latitude and longitude(): Find the longitude and latitude information of the current customer;

[0072] Find customer account information(): Find the customer's account information by calling the interface;

[0073] Find customer type(): Check whether the current customer is a VIP customer or a blacklist customer;

[0074] Enter sub-scene (<scene name>): The system supports scenes By calling sub-scenes, richer multi-round dialogues can be achieved, which is also conducive to scene reuse.

[0075] As an optional embodiment, Figure 7 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention. Figure 7As shown, after determining the interactive application scenario, the scenario processing elements are determined from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenario; and according to the above-mentioned scenario processing elements, a state element set is determined; the human-computer interaction rule search process is triggered according to the above-mentioned state element set; it is judged whether the current human-computer interaction rule meets the element combination condition, and after searching for the human-computer interaction rule that meets the element combination condition, the predicate action in the above-mentioned human-computer interaction rule is executed.

[0076] As an optional embodiment, Figure 8 is a flowchart of another optional human-computer interaction processing method according to an embodiment of the present invention. Figure 8 As shown, the above execution scenario selection rules include:

[0077] Step S402, collecting and arranging all historical conversation records to obtain a total conversation;

[0078] Step S404, combining all the intents in the total conversation in pairs to obtain an intent set;

[0079] Step S406, using the association rule in the scenario selection rule to generate an inference relationship based on the intention set, wherein the association rule is used to derive the probability of occurrence between the intention and the interactive application scenario based on the relationship between the intentions in the intention set;

[0080] Step S408, when selecting a scene candidate set, determining the probability of a second interaction intention related to the first interaction intention appearing successively after the first interaction intention appears based on the above reasoning relationship;

[0081] Step S410, combining the occurrence probability with the user feature information, calculating the feature value of each scene to be selected in the scene candidate set, and arranging them in reverse order according to the feature value to obtain a sorting result;

[0082] Step S412: selecting the above-mentioned interactive application scenario from the above-mentioned scenario candidate set according to the above-mentioned sorting result.

[0083] Optionally, the above scenario selection rules are used to locate the interactive application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogue.

[0084] Optionally, the first interaction intention is an initial intention determined according to the interaction request of the target object, and the second interaction intention is the next intention that may appear based on the first interaction intention and summarized according to historical conversation records.

[0085] Optionally, a second interaction intention whose occurrence probability is greater than a preset value is included in the above-mentioned scenario candidate set.

[0086] Optionally, the above user characteristic information may include but is not limited to: age, region, membership level, business behavior, etc.

[0087] It should be noted that the scene selection based on association rules is an inference relationship formed by statistically combining intentions based on historical conversation records (i.e. historical data), such as: A->B, which indicates the probability of occurrence of the second interaction intention B related to the first interaction intention A mentioned above. Its purpose is to find the relationship between the intentions in the intention set, so as to derive the probability of occurrence between the intention and the scene. For example, if the first interaction intention is "lost phone card", there is an 80% chance of applying for a replacement card service and entering the "replacement card scenario"; if the first interaction intention is "suspended due to arrears", there is a 60% chance of recharging and a 20% chance of changing the package.

[0088] Optionally, the correlation degree is used as a measurement indicator of the above association rule, that is, the correlation degree is used to characterize the probability of the second interaction intention related to the above first interaction intention appearing successively after the first interaction intention appears.

[0089] In an optional embodiment, all historical conversation records in the system are collected and sorted to obtain the total conversation n, and the sorting results are shown in Table 1; all intentions in the conversation are combined in pairs to form a new sample set, and the combination results are shown in Table 2; when selecting the scene candidate set, the second interaction intention related to the above first interaction intention is first found from the statistical data through the first interaction intention yx and the correlation (that is, the probability of occurrence of the above second interaction intention) is calculated. The calculation formula is as follows:

[0090]

[0091] Among them, r x,m represents the correlation between intention X and intention M, that is, the proportion of samples with consecutive appearances of intention X and intention M to samples with intention X, ∑y x y m Indicates the total number of samples where intention X and intention M appear consecutively; ∑y x Indicates the total number of samples with intent X.

[0092] Table 1

[0093] Session number Conversation Record intention 1 <![CDATA[t 1.1 、t 1.2 、t 1.3 ...]]> <![CDATA[y1、y2、y3....]]> …… …… …… n <![CDATA[t n.1 、t n.2 、t n.3 ...]]> <![CDATA[y1、y2、y3....]]>

[0094] Table 2

[0095]

[0096]

[0097] Optionally, a threshold ɑ (such as 0.6-0.8) is set according to the actual business environment, and the intentions with correlation higher than the threshold ɑ are selected, and the corresponding scenes are included in the scene candidate set. If the above scene candidate set is not empty, the feature value of each scene to be selected in the above scene candidate set is calculated in combination with the user feature information, and the scenes are arranged in reverse order according to the feature values; if the above scene candidate set is empty, the initial interaction information is further determined by business personnel access to determine whether it is mapped to the historical interaction scene.

[0098] In an optional embodiment, the above occurrence probability is combined with user feature information, the feature value of each to-be-selected scene in the above scene candidate set is calculated, and the scene candidate set is arranged in reverse order according to the feature value to obtain a sorting result, including: firstly collecting historical conversation records n and user feature information, and summarizing and statistically analyzing the statistical results as shown in Table 3; combining multiple interaction intentions in the historical conversation records in pairs to obtain interaction intentions. Figure 2 Tuple, the interaction Figure 2 The tuple and user feature information are cross-combined to form user feature information and interaction intention. Figure 2 The associated sample data of the tuple, the combination results are shown in Table 4; based on the above combination results, the feature value of each scene to be selected in the above scene candidate set is calculated, and the calculation formula is as follows:

[0099]

[0100] in, Indicates the current user features (f1...f n ),Intention y x and intention m Continuous occurrence of eigenvalues; represents user features (f1…f n ),Intention y x and intention m The number of consecutive samples; represents user features (f1…f n ) is the total number of samples under .

[0101] Table 3

[0102] Sample No. user feature intention 1 <![CDATA[u1]]> <![CDATA[f 1.1 、f 1.2 、f 1.3 ……]]> <![CDATA[y1、y2、y3....]]> …… …… …… …… n <![CDATA[u n ]]> <![CDATA[f n.1 、f n.2 、f n.3 ……]]> <![CDATA[y1、y2、y3....]]>

[0103] Table 4

[0104]

[0105]

[0106] The embodiments of the present invention can at least achieve the following technical effects: accurately understand the customer's initial intention through an automatic multi-round dialogue method, and at the same time infer the customer's true intention; business personnel can easily configure a variety of scenarios to ensure the simplicity of operation of the interaction process, improve the efficiency of human-computer interaction, and realize effective intelligent interaction; determine the set of candidate scenarios according to the intention correlation calculation formula to achieve accurate and effective push scenarios; use the eigenvalue calculation formula to select the optimal solution for the candidate scenarios, and further optimize the results of the scenario selection; abstract the interaction scenarios into scenario elements, scenario predicate actions, and scenario rules, and the scenario rules are composed of the values ​​of the scenario elements and the scenario predicate actions, and the scenario rules are stored and represented in a table form to facilitate the operation of business personnel.

[0107] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0108] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0109] Example 2

[0110] In this embodiment, a human-computer interaction processing device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the terms "unit" and "device" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0111] According to an embodiment of the present invention, a device embodiment for implementing the above-mentioned human-computer interaction processing method is also provided. Fig. 9 is a schematic diagram of the structure of a human-computer interaction processing device according to an embodiment of the present invention. Fig. 9As shown, the above-mentioned human-computer interaction processing device includes: an acquisition module 50, a first execution module 52, a second execution module 54, and a third execution module 56, wherein:

[0112] The acquisition module 50 is used to acquire the initial interaction information of the target object when an interaction request of the target object is detected, wherein the initial interaction information is obtained based on a summary of historical industry issues; the first execution module 52 is used to execute the human-computer interaction rules corresponding to the interaction application scenario if there is an interaction application scenario associated with the initial interaction information; the second execution module 54 is used to execute the scenario selection rules if there is no interaction application scenario associated with the initial interaction information, wherein the scenario selection rules are used to locate the interaction application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; the third execution module 56 is used to execute the human-computer interaction rules corresponding to the interaction application scenario if the scenario selection rules locate the interaction application scenario.

[0113] It should be noted that the above modules can be implemented by software or hardware. For example, for the latter, it can be implemented in the following ways: the above modules can be located in the same processor; or the above modules can be located in different processors in any combination.

[0114] It should be noted that the acquisition module 50, the first execution module 52, the second execution module 54, and the third execution module 56 correspond to steps S102 to S108 in Example 1, and the examples and application scenarios implemented by the modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned Example 1. It should be noted that the modules as part of the device can be run in a computer terminal.

[0115] It should be noted that the optional or preferred implementation of this embodiment can refer to the relevant description in Example 1, which will not be repeated here.

[0116] The above-mentioned human-computer interaction processing device may also include a processor and a memory. The above-mentioned acquisition module 50, the first execution module 52, the second execution module 54, the third execution module 56, etc. are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize corresponding functions.

[0117] In an optional embodiment, the above-mentioned device also includes: a judgment module, which is used to judge whether the above-mentioned initial interaction information is mapped to the historical interaction scene if the above-mentioned scene selection rule cannot locate the above-mentioned interactive application scene; a first determination module, which is used to establish an association relationship between the above-mentioned initial interaction information and the above-mentioned historical interaction scene if the above-mentioned initial interaction information has been mapped to the above-mentioned historical interaction scene, and determine the corresponding above-mentioned human-computer interaction rules based on the above-mentioned association relationship; an indication module, which is used to instruct the staff to use the human-computer interaction system to configure new human-computer interaction rules for the above-mentioned initial interaction information if the above-mentioned initial interaction information is not mapped to the above-mentioned historical interaction scene.

[0118] In an optional embodiment, the above-mentioned first execution module includes: a second determination module, which is used to determine the scenario processing elements from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenario; a first acquisition sub-module, which is used to combine the above-mentioned scenario processing elements and the element values ​​corresponding to the above-mentioned scenario processing elements to obtain a state element set; a search module, which is used to search for the above-mentioned human-computer interaction rules that meet the element combination conditions according to the above-mentioned state element set; and a first execution sub-module, which is used to execute the predicate action in the above-mentioned human-computer interaction rules.

[0119] In an optional embodiment, the second execution module includes: a second acquisition submodule, which is used to collect and organize all historical conversation records to obtain a total conversation; a third acquisition submodule, which is used to combine all intentions in the total conversation in pairs to obtain an intention set; a generation module, which is used to adopt the association rules in the scene selection rules to generate an inference relationship based on the intention set, wherein the association rules are used to derive the occurrence probability between the intention and the interactive application scenario based on the relationship between the intentions in the intention set; a third determination module, which is used to determine the probability of occurrence of a second interaction intention related to the first interaction intention after the first interaction intention appears when selecting a scene candidate set based on the inference relationship; a calculation module, which is used to combine the occurrence probability with user feature information, calculate the feature value of each scene to be selected in the scene candidate set, and arrange them in reverse order according to the feature value to obtain a sorting result; a selection module, which is used to select the interactive application scenario from the scene candidate set according to the sorting result.

[0120] The processor includes a kernel, which retrieves the corresponding program unit from the memory. The kernel may be one or more. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. The memory includes at least one memory chip.

[0121] According to an embodiment of the present application, an embodiment of a non-volatile storage medium is also provided. Optionally, in this embodiment, the non-volatile storage medium includes a stored program, wherein when the program is executed, the device where the non-volatile storage medium is located is controlled to execute any one of the human-computer interaction processing methods.

[0122] Optionally, in this embodiment, the non-volatile storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group, and the non-volatile storage medium includes a stored program.

[0123] Optionally, when the program is running, the device where the non-volatile storage medium is located is controlled to perform the following functions: when an interaction request of a target object is detected, initial interaction information of the target object is obtained, wherein the initial interaction information is obtained based on a summary of historical industry issues; if there is an interaction application scenario associated with the initial interaction information, the human-computer interaction rules corresponding to the interaction application scenario are executed; if there is no interaction application scenario associated with the initial interaction information, the scenario selection rules are executed, wherein the scenario selection rules are used to locate the interaction application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; if the scenario selection rules locate the interaction application scenario, the human-computer interaction rules corresponding to the interaction application scenario are executed.

[0124] Optionally, when the program is running, the device where the non-volatile storage medium is located is controlled to perform the following functions: if the above-mentioned scene selection rule cannot locate the above-mentioned interactive application scenario, determine whether the above-mentioned initial interaction information is mapped to the historical interaction scenario; if the above-mentioned initial interaction information has been mapped to the above-mentioned historical interaction scenario, establish an association relationship between the above-mentioned initial interaction information and the above-mentioned historical interaction scenario, and determine the corresponding above-mentioned human-computer interaction rule based on the above-mentioned association relationship; if the above-mentioned initial interaction information is not mapped to the above-mentioned historical interaction scenario, instruct the staff to use the human-computer interaction system to configure new human-computer interaction rules for the above-mentioned initial interaction information.

[0125] Optionally, when the program is running, the device where the non-volatile storage medium is located is controlled to perform the following functions: determine the scenario processing elements from the customer information, the above-mentioned consultation information, and the system operation information according to the above-mentioned interactive application scenario; combine the above-mentioned scenario processing elements and the element values ​​corresponding to the above-mentioned scenario processing elements to obtain a state element set; search for the above-mentioned human-computer interaction rules that meet the element combination conditions according to the above-mentioned state element set; and execute the predicate actions in the above-mentioned human-computer interaction rules.

[0126] Optionally, when the program is running, the device where the non-volatile storage medium is located is controlled to perform the following functions: collect and organize all historical conversation records to obtain a total conversation; combine all intentions in the above total conversation in pairs to obtain an intention set; adopt the association rules in the above scenario selection rules to generate an inference relationship based on the above intention set, wherein the above association rules are used to derive the occurrence probability between the above intention and the above interactive application scenario based on the relationship between the individual intentions in the above intention set; when selecting a scenario candidate set, determine the probability of occurrence of a second interactive intention related to the above first interactive intention after the first interactive intention appears based on the above inference relationship; combine the above occurrence probability with user feature information, calculate the feature value of each to-be-selected scene in the above scenario candidate set, and arrange them in reverse order according to the feature value to obtain a sorting result; select the above interactive application scenario from the above scenario candidate set according to the above sorting result.

[0127] According to an embodiment of the present application, an embodiment of a processor is also provided. Optionally, in this embodiment, the processor is used to run a program, wherein the program executes any one of the human-computer interaction processing methods when it is run.

[0128] According to an embodiment of the present application, an embodiment of a computer program product is also provided, which, when executed on a data processing device, is suitable for executing a program that initializes any one of the steps of the above-mentioned human-computer interaction processing method.

[0129] Optionally, the above-mentioned computer program product, when executed on a data processing device, is suitable for executing a program initialized with the following method steps: when an interaction request of a target object is detected, obtaining initial interaction information of the above-mentioned target object, wherein the above-mentioned initial interaction information is obtained based on a summary of historical industry issues; if there is an interaction application scenario associated with the above-mentioned initial interaction information, executing human-computer interaction rules corresponding to the above-mentioned interaction application scenario; if there is no interaction application scenario associated with the above-mentioned initial interaction information, executing scenario selection rules, wherein the above-mentioned scenario selection rules are used to locate the interaction application scenario to be executed based on the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; if the above-mentioned scenario selection rules locate the above-mentioned interaction application scenario, executing the human-computer interaction rules corresponding to the above-mentioned interaction application scenario.

[0130] According to an embodiment of the present application, an embodiment of an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute any one of the above-mentioned human-computer interaction processing methods.

[0131] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0132] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0134] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0135] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0136] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable non-volatile storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a non-volatile storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned non-volatile storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0137] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A human-computer interaction processing method, characterized in that: include: When an interaction request of a target object is detected, initial interaction information of the target object is obtained, wherein the initial interaction information is obtained based on a summary of historical industry issues; If there is an interactive application scenario associated with the initial interaction information, executing a human-computer interaction rule corresponding to the interactive application scenario; If there is no interactive application scenario associated with the initial interactive information, the scenario selection rule is executed, wherein the scenario selection rule is used to locate the interactive application scenario to be executed according to the speed of processing things, user satisfaction, and the number of multiple rounds of dialogues; the execution of the scenario selection rule includes: collecting and collating all historical dialogue records to obtain a total conversation; combining all intentions in the total conversation in pairs to obtain an intention set; using the association rule in the scenario selection rule to generate an inference relationship based on the intention set, wherein the association rule is used to derive the occurrence probability between the intention and the interactive application scenario based on the relationship between each intention in the intention set; when selecting the scenario candidate set, the probability of the occurrence of a second interactive intention related to the first interactive intention after the first interactive intention appears is determined based on the inference relationship; the occurrence probability is combined with user feature information to calculate the feature value of each scene to be selected in the scenario candidate set, and the scene is arranged in reverse order according to the feature value to obtain a sorting result; the interactive application scenario is selected from the scenario candidate set according to the sorting result; If the scene selection rule locates the interactive application scene, the human-computer interaction rule corresponding to the interactive application scene is executed.

2. The method according to claim 1, characterized in that: The method further comprises: If the scene selection rule cannot locate the interactive application scene, determining whether the initial interaction information is mapped to a historical interaction scene; If the initial interaction information has been mapped to the historical interaction scene, establishing an association relationship between the initial interaction information and the historical interaction scene, and determining the corresponding human-computer interaction rule based on the association relationship; If the initial interaction information is not mapped to the historical interaction scenario, the staff is instructed to use the human-computer interaction system to configure a new human-computer interaction rule for the initial interaction information.

3. The method according to claim 1, characterized in that The executing of the human-computer interaction rules corresponding to the interactive application scenario includes: Determine scenario processing elements from customer information, above-mentioned consultation information, and system operation information according to the interactive application scenario; The scene processing elements and the element values ​​corresponding to the scene processing elements are combined to obtain a state element set; Searching the human-computer interaction rule that satisfies the element combination condition according to the state element set; Execute the predicate action in the human-computer interaction rule.

4. A human-computer interaction processing device, characterized in that: include: An acquisition module, configured to acquire initial interaction information of a target object when an interaction request of the target object is detected, wherein the initial interaction information is obtained based on a summary of historical issues of the industry; A first execution module, configured to execute a human-computer interaction rule corresponding to an interaction application scenario if there is an interaction application scenario associated with the initial interaction information; a second execution module, configured to execute a scene selection rule if there is no interactive application scene associated with the initial interaction information, wherein the scene selection rule is used to locate the interactive application scene to be executed according to the speed of processing things, user satisfaction, and the number of multi-round dialogues; and a third execution module, configured to execute a human-computer interaction rule corresponding to the interactive application scene if the scene selection rule locates the interactive application scene; The second execution module includes: a second acquisition submodule, which is used to collect and organize all historical conversation records to obtain a total conversation; a third acquisition submodule, which is used to combine all intentions in the total conversation in pairs to obtain an intention set; a generation module, which is used to adopt the association rules in the scene selection rules to generate an inference relationship based on the intention set, wherein the association rule is used to derive the occurrence probability between the intention and the interactive application scenario based on the relationship between the intentions in the intention set; a third determination module, which is used to determine the probability of occurrence of a second interactive intention related to the first interactive intention after the first interactive intention appears when selecting a scene candidate set based on the inference relationship; a calculation module, which is used to combine the occurrence probability with user feature information, calculate the feature value of each scene to be selected in the scene candidate set, and arrange them in reverse order according to the feature value to obtain a sorting result; a selection module, which is used to select the interactive application scenario from the scene candidate set according to the sorting result.

5. The device according to claim 4, characterized in that The device also includes: A judgment module, configured to judge whether the initial interaction information is mapped to a historical interaction scene if the scene selection rule cannot locate the interaction application scene; A first determination module, configured to establish an association relationship between the initial interaction information and the historical interaction scene if the initial interaction information has been mapped to the historical interaction scene, and determine the corresponding human-computer interaction rule based on the association relationship; The instruction module is used to instruct the staff to use the human-computer interaction system to configure new human-computer interaction rules for the initial interaction information if the initial interaction information is not mapped to the historical interaction scene.

6. The device according to claim 4, characterized in that The first execution module includes: A second determination module is used to determine a scenario processing element from the customer information, the above consultation information, and the system operation information according to the interactive application scenario; A first acquisition submodule is used to combine the scene processing elements and the element values ​​corresponding to the scene processing elements to obtain a state element set; A search module, used for searching the human-computer interaction rule satisfying the element combination condition according to the state element set; The first execution submodule is used to execute the predicate action in the human-computer interaction rule.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the human-computer interaction processing method according to any one of claims 1 to 3.

8. An electronic device, comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the human-computer interaction processing method according to any one of claims 1 to 3.

Images