An intelligent reply method and device, a storage medium, and an electronic device

Abstract

In the method, device, storage medium and electronic equipment provided in the specification, game data from at least two sources is collected in real time, entities in Chinese text data are extracted, an entity relationship graph is determined, input text of a user is obtained, entities in the input text are determined, reply text of the input text of the user is generated according to the entities in the input text, the entity relationship graph and the game data, and the reply text is returned. The entity relationship graph of each entity is determined according to game data from multiple sources, which facilitates the identification of game terms used by game users by an intelligent customer service. Meanwhile, game data from multiple sources is obtained in real time, so that the intelligent customer service can better understand emerging game content and provide answers for the user. According to the input text of the user, the answer is generated in real time based on the game data and the entity relationship graph, so that the problem of mismatch between the reply text and the input text of the user is reduced.

Description

TECHNICAL FIELD

[0001] The present specification relates to the field of computers, and in particular, to an intelligent reply method and device, a storage medium, and an electronic device. BACKGROUND

[0002] With the development of computer technology, the game experience of users affects the game life. Intelligent customer service has become an important tool for enhancing user experience.

[0003] In the prior art, intelligent customer service usually performs keyword matching on the input text of the user and the text stored in the database, and returns the reply corresponding to the text with the highest matching degree to the user as output. However, this method relies on the pre-stored text in the database. The intelligent customer service based on the pre-set database lacks understanding of emerging game content, so that the intelligent customer service does not match the reply text of some users with the input text of the user, and the user needs to adjust the input input text multiple times to obtain the reply text output by the intelligent customer service related to the input text, which affects the game experience of the user and the use experience of the intelligent customer service.

[0004] Therefore, the present specification provides an intelligent reply method, device, storage medium, and electronic device. SUMMARY

[0005] The present specification provides an intelligent reply method, device, storage medium, and electronic device to partially solve the above problems existing in the prior art.

[0006] The present specification adopts the following technical solutions:

[0007] An intelligent reply method comprises:

[0008] Real-time collection of game data, word segmentation processing of text data in the game data, determination of each data entity, the game data comprising at least two of feedback data of a user, running data of a game, and evaluation data of a game;

[0009] According to the game data, the relationship between the data entities is determined, a plurality of entity relationship graphs are constructed, and the database is updated according to the game data and the entity relationship graphs;

[0010] Obtaining input text input by a user, extracting each input entity in the input text, generating reply text of the input text according to the each input entity and the game data in the database and the entity relationship graphs, and returning the reply text to the user.

[0011] Optionally, updating the database according to the game data and the entity relationship graphs comprises:

[0012] Classify the game data, and determine the data corresponding to each classification according to the classification results respectively;

[0013] De-duplicate the data corresponding to each classification with the game data stored in the database, and store the de-duplicated data to the address corresponding to the database respectively;

[0014] Determine the similarity between the entity relationship graph and the entity relationship graph stored in the database, and store the entity relationship graph with a similarity less than a preset value to the database.

[0015] Optionally, before obtaining the input text input by the user, the method further comprises:

[0016] Obtaining the feedback data in the game data in real time;

[0017] Inputting the feedback data into the trained text generation model, so that the text generation model generates at least one shortcut text;

[0018] Generating the corresponding reply text of the shortcut text according to the game data, and updating the database according to the shortcut text and the corresponding reply text.

[0019] Optionally, the method further comprises:

[0020] Sending a reply satisfaction degree evaluation request to the user, and obtaining the satisfaction degree returned by the user;

[0021] When the satisfaction degree reaches a preset value, updating the database according to the input text of the user and the corresponding reply text.

[0022] Optionally, generating the reply text of the input text according to the input entity, the game data in the database and the entity relationship graph and returning it to the user, specifically comprising:

[0023] Matching the input text in the database, and calculating the matching degree of the input text and the text stored in the database;

[0024] Judging whether the matching degree reaches a preset value, if yes, returning the reply text corresponding to the text stored in the database to the user;

[0025] If not, generating the reply text of the input text according to the input entity, the game data in the database and the entity relationship graph and returning it to the user.

[0026] Optionally, before generating the reply text of the input text according to the input entities, the game data in the database and the entity relationship graph and returning the reply text to the user, the method further comprises:

[0027] According to the matching of the input entities and the entities corresponding to the preset text type, the type of the input text is determined.

[0028] According to the type, a preset game optimization direction is determined, and the optimization direction comprises at least one of technical optimization, rule optimization, numerical value optimization and guidance optimization.

[0029] Optionally, generating the reply text of the input text according to the input entities, the game data in the database and the entity relationship graph and returning the reply text to the user specifically comprises:

[0030] The historical input input by the user and the corresponding historical reply are obtained.

[0031] The historical input is subjected to entity extraction to determine historical entities, and the matching of the historical entities and the entities corresponding to the preset text type is performed to determine the type of the historical input.

[0032] It is judged whether the type of the historical input is consistent with the type of the input text.

[0033] If yes, the reply text of the input text is generated according to the historical reply corresponding to the historical input, the input entities and the game data in the database and the entity relationship graph and is returned to the user.

[0034] If no, the reply text of the input text is generated according to the input entities, the game data in the database and the entity relationship graph and is returned to the user.

[0035] The present specification provides an intelligent reply device, which is used to implement the above intelligent reply method, comprising:

[0036] The collection module is used to collect game data in real time, perform word segmentation processing on text data in the game data, and determine each data entity, wherein the game data comprises at least two of feedback data of a user, running data of a game and evaluation data of a game.

[0037] The update module is used to determine the relationship between the data entities according to the game data, construct a plurality of entity relationship graphs, and update a database according to the game data and the entity relationship graphs.

[0038] The reply module is configured to acquire input text input by a user, extract each input entity in the input text, and generate reply text of the input text based on the each input entity, the game data in the database, and the entity relationship graph, and return the reply text to the user.

[0039] The present specification provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the intelligent reply method.

[0040] The present specification provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the intelligent reply method when executing the program.

[0041] The above at least one technical solution adopted by the present specification can achieve the following beneficial effects:

[0042] In the intelligent reply method provided by the present specification, game data from at least two sources is collected in real time, each entity in the text data is extracted to determine an entity relationship graph, input text of a user is acquired, entities in the input text are determined, reply text of the input text of the user is generated based on the entities in the input text, the entity relationship graph, and the game data, and the reply text is returned.

[0043] As can be seen from the above method, the relationship graph of each entity is determined based on game data from multiple sources, which facilitates the intelligent customer service to recognize game terms used by game users. At the same time, game data from multiple sources is acquired in real time, which enables the intelligent customer service to better understand emerging game content so as to help users solve problems related to the emerging game content. The answer is generated in real time based on the game data and the entity relationship graph according to the input text of the user, which reduces the problem of mismatch between the reply text and the input text of the user. BRIEF DESCRIPTION OF DRAWINGS

[0044] The accompanying drawings, which are included to provide a further understanding of the present specification, constitute a part of the present specification, and the illustrative embodiments of the present specification and their description serve to explain the present specification, and do not constitute an improper limitation on the present specification. In the drawings:

[0045] Figure 1 A schematic diagram of an intelligent reply method provided by an embodiment of the present specification;

[0046] Figure 2 A schematic diagram of an entity relationship graph provided by an embodiment of the present specification;

[0047] Figure 3 A schematic diagram of an entity relationship graph provided by an embodiment of the present specification;

[0048] Figure 4 A schematic diagram of an intelligent reply device provided by an embodiment of the present specification;

[0049] Figure 5 An electronic device schematic diagram corresponding to Figure 1 provided by the present specification. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the present specification clearer, the technical solutions of the present specification will be described in detail below in combination with specific embodiments of the present specification and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present specification, not all the embodiments. Based on the embodiments in the present specification, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present invention.

[0051] The technical solutions provided by the embodiments of the present specification will be described in detail below in combination with the drawings.

[0052] Figure 1 A schematic diagram of an intelligent reply method provided by the present specification, specifically comprising the following steps:

[0053] S100: Real-time acquisition of game data, word segmentation processing of text data in the game data, determination of each data entity, the game data including at least two of user feedback data, game running data and game evaluation data.

[0054] In one or more embodiments of the present specification, the intelligent reply method is not limited to be implemented by a specific device, and can be implemented by any electronic device, which can be a terminal, a server, a single-chip microcomputer, etc. Since the subsequent steps of the method include entity recognition, user question answering, etc., these steps are usually executed by a server, and the intelligent reply method will be described in the subsequent one or more embodiments of the present specification as an example of the server.

[0055] In order to realize the understanding of emerging game content by intelligent customer service, the server should acquire different kinds of game data in real time, so as to generate reply text corresponding to user input text according to different game data.

[0056] Specifically, the server can collect game data, wherein the game data includes at least two of feedback data of a user, running data, and evaluation data, the feedback data of the user includes intelligent customer service interaction history, manual customer service interaction history, or feedback of the user in an official published questionnaire, the running data of the game includes game logs, game interaction data, etc., and the evaluation data of the game includes evaluation, help-seeking, suggestion, etc. related to the game published by the user on a social software or a website. Of course, the server can also obtain other related data to enrich the game database, so that the intelligent customer service can better understand the emerging game content and the feedback of each user. The emerging game content includes the latest update content of the game, homo culture, user's nickname for various items, equipment, and playing methods in the game, etc.

[0057] It should be noted that in one or more embodiments of the present specification, the intelligent customer service interaction history can include the customer service interaction history of all users. Moreover, the source of the game data collected by the server is not limited, the game data can come from the official data of the game, can be data searched based on a search engine, or can be data fed back by the user. It is also not limited that the server specifically adopts which way to realize the acquisition of various game data, such as searching the contents related to the game through a third-party search engine, can be through issuing a questionnaire to the user to determine the user's feedback on the game, or can be through program code or other tools to obtain the contents related to the game from a third-party social software or communication software as game data.

[0058] Then, in order to realize the understanding of the intelligent customer service to the emerging game content and the understanding of the specific game terms input by the user, the server should also perform entity extraction on the text data in the real-time collected game data. Specifically, the server performs word segmentation processing on the text data in the game data, and performs part-of-speech tagging, identifies entities according to the tagged part-of-speech, and identifies entities with the part-of-speech of noun, verb, or adjective.

[0059] It should be noted that in one or more embodiments of the present specification, the specific type of the game data is not limited, which can be unstructured data or structured data. The game data at least includes text data.

[0060] S102: According to the game data, determine the relationship between each data entity, construct a plurality of entity relationship graphs, and update the database according to the game data and the entity relationship graph.

[0061] After determining the entities in the game data, the server can determine the relationship between the entities according to the game data to better understand the game terminology, construct an entity relationship graph, and then update the database according to the entity relationship graph and the real-time acquired game data. In order to generate the reply text of the user input text according to the updated database subsequently.

[0062] Specifically, the server determines the relationship between the entities in the game data, such as synonym, mutual restraint, sub-equipment, synthesis, etc. For each entity, the relationship between the entity and other entities is determined, and an entity relationship graph is constructed. Then, the database is updated according to the constructed entity relationship graph and the real-time collected game data.

[0063] It should be noted that in one or more embodiments of the present specification, the relationship between the entities is not limited to being determined in a specific way. It can be extracted based on the text data in the game data, it can also be predicted by a large language model, or other methods can be used. Of course, the type of the determined entity relationship graph is not limited. The two entities can be connected by a line representing the relationship between the two entities to form an image, or the relationship between the entities can be represented in the form of a knowledge graph. Of course, the relationship between the entities can also be represented in other forms, which are not limited in the present specification.

[0064] Of course, in one or more embodiments of the present specification, the number and size of the constructed entity relationship graph are not limited. Each entity relationship graph can only represent the relationship between two entities, such as Figure 2 as shown in Figure 2 a schematic diagram of an entity relationship graph provided by the present specification. Each circle represents an entity, and the connection line between the circles represents the relationship between the entities. Among them, entity a includes entity b, entity c, and entity e, entity b and entity e are aliases of each other, and entity d can be obtained through entity b or entity e. Figure 3 as shown in Figure 3 a schematic diagram of an entity relationship graph provided by the present specification. Each circle represents an entity, and the connection line between the circles represents the relationship between the entities. Among them, entity a includes entity b and entity c and entity e, entity b and entity e are aliases of each other, and entity d can be obtained through entity b or entity e.

[0065] S104: Acquire the question input by the user, extract each question entity in the question, and generate an answer to the question according to the database and the each question entity and return the answer to the user.

[0066] After updating the database, when the user input question is obtained, the server should first perform entity extraction on the user input question to generate an answer to the question, then match the extracted entities in the question with the entities in the database, and determine the corresponding data according to the relationship between the matched entities and other entities, and generate the corresponding answer.

[0067] Specifically, the user input question is processed by word segmentation, and each word is taken as an entity. Then for each entity, the related entities of the entity are determined according to the entity relationship graph in the database, and the game data related to the entity and other entities are determined, and then the answer is generated based on the game data related to the entity and other entities in the database by training the model reply generation model or large language model and returned to the user.

[0068] Based on Figure 1 In an intelligent reply method, at least two sources of game data are collected in real time, and each entity in the text data is extracted to determine the entity relationship graph, then the user input text is obtained, and the entities in the input text are determined, and the reply text of the user input text is generated based on the entities in the input text, the entity relationship graph and the game data and returned.

[0069] As can be seen from the above method, the relationship graph of each entity is determined based on multiple sources of game data, which facilitates the intelligent customer service to recognize the game terms used by game users, at the same time, multiple sources of game data are obtained in real time, which enables the intelligent customer service to better understand emerging game content, so as to answer the user's questions about emerging game content. And based on the game data and the entity relationship graph, the answer is generated in real time according to the user's input text, which reduces the problem of mismatch between the reply text and the user's input text.

[0070] In addition, in step S102, in order to reduce the processing amount of the server on the game data in the answer generation process, the server can also store the game data classified when building the server, so that when generating the answer, the corresponding type of game data can be obtained.

[0071] Specifically, the server can classify the game data, and determine the game data corresponding to each classification according to the classification result, then store each classification corresponding game data to the corresponding database address after deduplication with the game data stored in the database. Determine the similarity between each entity relationship graph and the entity relationship graph stored in the database, and store the entity relationship graph with a similarity less than a preset value to the database.

[0072] Of course, in one or more embodiments of the present specification, no limitation is imposed on the specific classification of game data, and the classification of game data can be implemented according to the structure of game data, such as structured data, unstructured data, and graph data. The classification of game data can be implemented according to the source of game data, such as game internal data and game feedback data. The classification can also be implemented according to other methods.

[0073] Then, in order to further improve the reply efficiency of the input text of the user, the server can also acquire the feedback data of each user in real time, and then generate common shortcut texts according to the feedback data of the user, determine the reply texts corresponding to the shortcut texts, and store them in the database, so that if the input text of the user is a shortcut text, the intelligent customer service can quickly respond and directly return the corresponding reply text to the user.

[0074] Specifically, the server acquires the feedback data in the game data in step S100, inputs the feedback data into the trained text generation model, and makes the text generation model generate at least one shortcut text. Then, the reply text of the shortcut text is generated according to the game data, and the database is updated according to the shortcut text and the corresponding reply text. The shortcut text and the corresponding reply text can be stored in the database in the form of a text pair. When matching the input text of the user with the shortcut text in the database, the server can also match each input entity in the input text with the entity in the shortcut text, so as to realize fast matching.

[0075] It should be noted that whether the server receives the input text input by the user or not, the server continuously generates shortcut texts and corresponding reply texts according to the acquired feedback data, so as to update the database, so that when the input text input by the user is received, fast matching can be performed according to the database, and the efficiency of intelligent reply is improved.

[0076] In addition, in one or more embodiments of the present specification, the server can also classify and store each shortcut text. Then, after acquiring the input text of the user, the input text is classified to determine the type of the input text, and then the type of the input text and the type of the shortcut text are matched to determine the shortcut texts of the corresponding type, and then the input text is matched with the shortcut texts in the classification, so as to further ensure the accuracy of the reply text and the reply efficiency.

[0077] It should be noted that in one or more embodiments of the present specification, no limitation is imposed on the specific method used by the server to generate shortcut texts, which can be a trained text generation model or a fine-tuned large language model.

[0078] Further, in one or more embodiments of the present specification, the text generation model is not limited to a specific model trained by the text generation model. It can be a neural network model or other model. During training, a plurality of feedback texts can be predetermined, and the shortcut questions in the feedback texts can be extracted as labels. Then the feedback texts are input into the text generation model, and the text generation model generates shortcut texts based on the feedback texts. The difference between the shortcut texts generated by the text generation model and the labels is the optimization target, and the text generation model is trained. Similarly, the fine-tuning of the large language model can also be as described above.

[0079] In addition, in order to further improve the user experience of the intelligent customer service, the server can also return a reply text to the user, and at the same time return a reply satisfaction evaluation request to the user, so as to update the database according to the user's evaluation.

[0080] Specifically, the server can send a reply satisfaction evaluation request to the user and obtain the user's return satisfaction. When the satisfaction reaches a preset value, the user's input text and the corresponding reply text are updated to the database. So that when other users raise the same or similar text, the server can directly return the corresponding reply text. If the user returns a low satisfaction, it means that the intelligent customer service has a low accuracy in generating reply texts for the input text, and can return to manual generation. It can also be based on the type of input text of the user to determine that the intelligent customer service has insufficient accuracy in generating answers to this type of question, and needs to be adjusted and returned to manual processing.

[0081] In addition, based on the real-time updated database and the input text input by the user, the server can also determine the optimization direction of the game. Specifically, the server can match each input entity with the preset entity corresponding to the text type to determine the text type, and then determine the preset game optimization direction according to the text type. The optimization direction includes at least one of technical optimization, rule optimization, numerical optimization, and guidance optimization. For example, if the user inputs the text "Why do I always fail when logging in to the game?", it means that the user's question type is a technical problem, which corresponds to technical optimization. When this type of question is raised frequently, it means that the official should test the stability and pressure of the server, or update the server to maintain the stability of the game server and improve the user's game experience. If the user inputs the text "Why can't I pass a certain level?", it means that the game's numerical value or rules may need to be optimized, especially when a large number of users raise the problem of not being able to pass a certain level. It means that the rules or numerical value of the game need to be optimized to improve the user's game experience. When the user inputs the text "How to do XX task", it means that the guidance of the task in the game has a problem, and the guidance of the task should be strengthened.

[0082] Further, in order to make the answer generated for the user input text more accurate, the server can also obtain the historical text input by the user when generating the answer, so as to identify the intention of the user according to the historical reply, and thus generate a more accurate answer.

[0083] Specifically, the server obtains the historical text input by the user and the corresponding reply text. Then, the historical text is subjected to entity extraction, the historical entities are determined, and the types of the historical texts are determined according to the matching between the historical entities and the entities corresponding to the preset text types. Then, it is judged whether the types of the historical texts are consistent with the type of the input text. If yes, the reply text is generated according to the reply text corresponding to the historical text, the input entities and the database, and is returned to the user. If no, it is indicated that the input text proposed by the user this time is irrelevant to the historical text, and the reply text is generated according to the input entities and the database and is returned to the user.

[0084] Of course, when determining the historical text of the user, the server should extract the historical text within a preset time, so as to ensure the relevance of the historical text and the input text of the user this time.

[0085] The above is an intelligent reply method provided by one or more embodiments of the present specification. Based on the same idea, the present specification also provides a corresponding intelligent reply device, as shown in Figure 4 .

[0086] Figure 4 A schematic diagram of an intelligent reply device provided by the present specification, which is used to execute the intelligent reply method provided by the present specification. Figure 1 .

[0087] The acquisition module 400 is configured to acquire game data in real time, perform word segmentation processing on text data in the game data, and determine data entities. The game data includes at least two of feedback data of a user, running data of a game, and evaluation data of the game.

[0088] The update module 401 is configured to determine relationships between the data entities according to the game data, construct a plurality of entity relationship graphs, and update a database according to the game data and the entity relationship graphs.

[0089] The reply module 402 is configured to obtain input text input by a user, extract input entities in the input text, generate reply text of the input text according to the input entities, the game data in the database, and the entity relationship graphs, and return the reply text to the user.

[0090] Optionally, the updating module 401 is specifically configured to classify the game data, and determine data corresponding to each classification according to a classification result; remove duplicate data corresponding to each classification from the game data stored in the database, and store the removed data to addresses corresponding to the database respectively; determine a similarity between the entity relationship graph and an entity relationship graph stored in the database, and store an entity relationship graph with a similarity less than a preset value to the database.

[0091] Optionally, the updating module 401 is specifically configured to acquire feedback data in the game data in real time; input the feedback data into the trained text generation model, so that the text generation model generates at least one shortcut text; generate a corresponding reply text of the shortcut text according to the game data, and update the database according to the shortcut text and the corresponding reply text.

[0092] Optionally, the apparatus further comprises an evaluation module 403 configured to send a reply satisfaction degree evaluation request to the user, and acquire a satisfaction degree returned by the user; when the satisfaction degree reaches a preset value, update the database according to the input text of the user and the corresponding reply text.

[0093] Optionally, the reply module 402 is specifically configured to perform matching in the database according to the input text, calculate a matching degree of the input text and the text stored in the database, and determine whether the matching degree reaches a preset value; if yes, return the reply text corresponding to the text stored in the database to the user; if no, generate the reply text of the input text according to the input entity, the game data in the database and the entity relationship graph, and return the reply text to the user.

[0094] Optionally, the apparatus further comprises an optimization module 404 configured to perform matching according to the input entity and a preset entity corresponding to a text type, determine a type of the input text, determine a preset game optimization direction according to the type, and the optimization direction comprises at least one of a technical optimization, a rule optimization, a numerical value optimization and a guide optimization.

[0095] Optionally, the reply module 402 is specifically configured to acquire historical input and corresponding historical reply of the user input; perform entity extraction on the historical input, determine historical entities, and match the historical entities with entities corresponding to a preset text type to determine a type of the historical input; determine whether the type of the historical input is consistent with the type of the input text; if yes, generate the reply text of the input text according to the historical reply corresponding to the historical input, the input entities, and the game data and the entity relationship graph in the database and return the reply text to the user; and if no, generate the reply text of the input text according to the input entities and the game data and the entity relationship graph in the database and return the reply text to the user.

[0096] The present specification also provides Figure 5 The schematic structural diagram of the electronic device is shown. As Figure 5 described above, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and of course can also include other hardware required by a business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs to implement the above Figure 1 intelligent reply method. Of course, in addition to the software implementation manner, the present specification does not exclude other implementation manners, such as a logic device or a combination of software and hardware, and so on, that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or a logic device.

[0097] In the 1990s, it was quite obvious to distinguish whether an improvement in a technology was in hardware (e.g., improvement in circuit structures of diodes, transistors, switches, etc.) or in software (improvement in method flow). However, as technology has evolved, many improvements in method flow today can be considered as direct improvements in hardware circuit structures. Designers almost always obtain the corresponding hardware circuit structures by programming the improved method flow into hardware circuits. Therefore, it cannot be said that an improvement in a method flow cannot be implemented by hardware entity modules. For example, a programmable logic device (PLD) such as a field programmable gate array (FPGA) is an integrated circuit whose logic function is determined by user programming of the device. A digital system is "integrated" on a piece of PLD by the designer programming it by himself, without having to ask a chip manufacturer to design and manufacture a special integrated circuit chip. Moreover, instead of manually fabricating an integrated circuit chip, this programming is now mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing a program, and the original code before compilation also has to be written in a specific programming language, which is called a hardware description language (HDL), and there are many types of HDL, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc., and the most commonly used are VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should be aware that, as long as the method flow is logically programmed in the above-mentioned hardware description languages and programmed into an integrated circuit, a hardware circuit that implements the logical method flow can be easily obtained.

[0098] The controller can be implemented in any suitable way, for example, the controller can take the form of a microprocessor or processor and a computer readable medium storing computer readable program code, such as software or firmware, executable by the (micro)processor, logic gates, switches, an application specific integrated circuit (ASIC), a programmable logic controller and an embedded microcontroller, examples of which include but are not limited to the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20 and Silicone Labs C8051F320, the memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also know that, in addition to being implemented in pure computer readable program code, the controller can equally well be implemented to perform the same functions using logic gates, switches, an application specific integrated circuit, a programmable logic controller and an embedded microcontroller, etc. by means of a logical programming of the method steps. The controller can thus be considered as a hardware component, and the means comprised therein for performing the various functions can be considered as structures within the hardware component. Alternatively, the means for performing the various functions can even be considered as both a software module implementing the method and a structure within the hardware component.

[0099] The systems, apparatuses, modules or units illustrated by the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0100] For the sake of description, the above apparatuses are described in various units by functions respectively. Of course, the functions of each unit can be implemented in the same or multiple software and / or hardware in implementing the present specification.

[0101] Those skilled in the art will understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0102] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0103] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0104] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks in the flowcharts and / or combination thereof. ​ one or more flowcharts and / or blocks in the flowcharts and / or combination thereof.

[0105] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0106] The memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) and / or cache memory, non-volatile memory, such as read-only memory (ROM), EPROM, and / or flash memory. The memory is an example of computer-readable media.

[0107] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0108] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or apparatus that comprises a list of elements does not only include those elements, but can also include other elements not expressly listed or inherent to such process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0109] Those skilled in the art will appreciate that embodiments of the present specification can be provided as methods, systems or computer program products. Therefore, the present specification can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] The present specification can be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The present specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices that are connected through a communication network. In a distributed computing environment, program modules can be located in both local and remote computer storage media including storage devices.

[0111] The various embodiments described in this specification are described using a numbering of embodiments approach: these are each individually integrated contributions pertaining to different but related aspects of the description. Each of the various embodiments can stand on its own, and each can be used to advantage in a wide variety of applications. The various embodiments can be used in combination with each other in any suitable way. Not all components of each embodiment are necessarily described with respect to each illustrated embodiment. Methods of the various embodiments can be performed in any suitable order.

[0112] The above description is embodied in the form of embodiments only and is not intended to limit the present specification. The present specification can be variously changed and modified by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present specification should be included in the scope of the claims of the present specification.

Claims

1. An intelligent reply method, characterized by, The method comprises: real-time acquisition of game data, word segmentation processing of text data in the game data, determination of each data entity, the game data including at least two of user feedback data, game running data and game evaluation data; determining the relationship between each data entity according to the game data, constructing a plurality of entity relationship graphs, and updating the database according to the game data and the entity relationship graph; obtaining input text input by a user, extracting each input entity in the input text, generating reply text of the input text according to the each input entity and the game data in the database and the entity relationship graph, and returning the reply text to the user, and determining the type of the input text according to the each input entity and the entity corresponding to the preset text type, determining a preset game optimization direction according to the type, the optimization direction including at least one of technical optimization, rule optimization, numerical optimization and guidance optimization; The method further comprises: real-time acquisition of feedback data in the game data; inputting the feedback data into a trained text generation model to make the text generation model generate at least one shortcut text; generating a corresponding reply text of the shortcut text according to the game data, and updating the database according to the shortcut text and the corresponding reply text.

2. The method of claim 1, wherein, updating the database according to the game data and the entity relationship graph, specifically comprising: classifying the game data, and determining the data corresponding to each classification according to the classification result; de-duplicating the data corresponding to each classification and the game data stored in the database, and storing the de-duplicated data to the address corresponding to the database; determining the similarity between the entity relationship graph and the entity relationship graph stored in the database, and storing the entity relationship graph with a similarity less than a preset value to the database.

3. The method of claim 1, wherein, The method further comprises: sending a reply satisfaction degree evaluation request to the user and obtaining the satisfaction degree returned by the user; when the satisfaction degree reaches a preset value, updating the database according to the input text and the corresponding reply text of the user.

4. The method of claim 1, wherein, generating the reply text of the input text according to the each input entity and the game data in the database and the entity relationship graph, and returning the reply text to the user, specifically comprising: matching the input text in the database, calculating the matching degree of the input text and the text stored in the database; determining whether the matching degree reaches a preset value, if yes, returning the reply text corresponding to the text stored in the database to the user; if not, generating the reply text of the input text according to the each input entity and the game data in the database and the entity relationship graph, and returning the reply text to the user.

5. The method of claim 1, wherein, generating the reply text of the input text according to the each input entity and the game data in the database and the entity relationship graph, and returning the reply text to the user, specifically comprising: obtaining historical input and corresponding historical reply input by the user; perform entity extraction on the historical input to determine historical entities, and match the historical entities with entities corresponding to a preset text type to determine a type of the historical input; determine whether the type of the historical input is consistent with a type of the input text; if yes, generate a reply text of the input text based on a historical reply corresponding to the historical input, the input entities, and the game data and the entity relationship graph in the database, and return the reply text to the user; if no, generate a reply text of the input text based on the input entities and the game data and the entity relationship graph in the database, and return the reply text to the user.

6. An intelligent reply device characterized by, The method comprises: collecting game data in real time, performing word segmentation on text data in the game data, and determining data entities, wherein the game data comprises at least two of feedback data of a user, running data of a game, and evaluation data of the game; updating the database based on the game data and the entity relationship graph; a reply module configured to obtain an input text input by a user, extract input entities in the input text, generate a reply text of the input text based on the input entities and the game data and the entity relationship graph in the database, and return the reply text to the user, and determine a type of the input text based on the input entities and entities corresponding to a preset text type, determine a preset game optimization direction based on the type, wherein the optimization direction comprises at least one of technical optimization, rule optimization, numerical optimization, and guidance optimization; The updating module is further configured to obtain feedback data in the game data in real time, input the feedback data into a trained text generation model, make the text generation model generate at least one shortcut text, generate a corresponding reply text of the shortcut text based on the game data, and update the database based on the shortcut text and the corresponding reply text.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and the computer program is executed by the processor to implement the method of any one of claims 1-5.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the program to implement the method of any one of claims 1-5.

Images