WO2020253109A1 - Procédé et appareil de recommandation d'informations de ressources basés sur la reconnaissance vocale, et terminal et support - Google Patents

Procédé et appareil de recommandation d'informations de ressources basés sur la reconnaissance vocale, et terminal et support Download PDF

Info

Publication number
WO2020253109A1
WO2020253109A1 PCT/CN2019/120891 CN2019120891W WO2020253109A1 WO 2020253109 A1 WO2020253109 A1 WO 2020253109A1 CN 2019120891 W CN2019120891 W CN 2019120891W WO 2020253109 A1 WO2020253109 A1 WO 2020253109A1
Authority
WO
WIPO (PCT)
Prior art keywords
resource
recommended
user
data information
genetic algorithm
Prior art date
Application number
PCT/CN2019/120891
Other languages
English (en)
Chinese (zh)
Inventor
燕如
Original Assignee
深圳壹账通智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳壹账通智能科技有限公司 filed Critical 深圳壹账通智能科技有限公司
Publication of WO2020253109A1 publication Critical patent/WO2020253109A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/26—Speech to text systems
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
    • G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use

Definitions

  • This application relates to the field of artificial intelligence technology, and in particular to a method, device, terminal and medium for recommending resource information based on speech recognition.
  • recommenders In the promotion process of resource services, recommenders are often used for promotion, and recommenders usually only recommend resource names to users based on experience values. However, due to the different personal situations of social personnel, the variety of resource names, and the different product terms and application scenarios between resources, it is difficult for recommenders to analyze the resources that are more suitable for users in a short time. Resource recommendation accuracy is low.
  • the main purpose of this application is to provide a method, device, terminal and medium for recommending resource information based on speech recognition, aiming to solve the problem of difficulty in determining resources suitable for users in a short time in the resource recommendation process of the prior art, resulting in low accuracy of resource recommendation Technical issues.
  • this application provides a method for recommending resource information based on speech recognition, which includes the following steps:
  • the resource type determine each resource name corresponding to the resource type and the fitness function corresponding to the resource type;
  • the resource information corresponding to the name of the resource to be recommended is sent to the user to be recommended.
  • this application also provides a device for recommending resource information based on speech recognition, including:
  • An obtaining module configured to obtain an audio clip of the user to be recommended, and analyze the data information of the user to be recommended according to the audio clip;
  • the determining module is configured to determine the resource type applicable to the user to be recommended according to the data information
  • the selection module is configured to determine, according to the resource type, each resource name corresponding to the resource type and a fitness function corresponding to the resource type;
  • the update module is used to update the preset genetic algorithm model according to the determined fitness function
  • a calculation module configured to use the name of each resource as an input parameter of the genetic algorithm model and run the genetic algorithm model to obtain the name of the resource to be recommended;
  • the recommendation module is used to send the resource information corresponding to the name of the resource to be recommended to the user to be recommended.
  • this application also provides a terminal, the terminal comprising: a memory, a processor, and computer-readable instructions stored on the memory and capable of running on the processor, the computer-readable instructions It is configured to implement the steps of the resource information recommendation method based on voice recognition as described above.
  • the present application also provides a storage medium storing computer-readable instructions, and when the computer-readable instructions are executed by a processor, the above-mentioned method for recommending resource information based on speech recognition is implemented. step.
  • FIG. 1 is a schematic structural diagram of a terminal of a hardware operating environment involved in a solution of an embodiment of the present application
  • FIG. 2 is a schematic flowchart of a first embodiment of a method for recommending resource information based on speech recognition in this application;
  • FIG. 3 is a schematic flowchart of a second embodiment of a method for recommending resource information based on speech recognition in this application;
  • FIG. 4 is a schematic flowchart of a third embodiment of a method for recommending resource information based on speech recognition in this application;
  • FIG. 5 is a schematic flowchart of a fourth embodiment of a method for recommending resource information based on speech recognition in this application;
  • FIG. 6 is a schematic flowchart of a fifth embodiment of a method for recommending resource information based on speech recognition in this application;
  • Fig. 7 is a structural block diagram of a first embodiment of a device for recommending resource information based on speech recognition in this application.
  • FIG. 1 is a schematic diagram of a terminal structure of a hardware running environment involved in a solution of an embodiment of the application.
  • the terminal may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005.
  • the communication bus 1002 is used to implement connection and communication between these components.
  • the user interface 1003 may include a display screen (Display) and an input module such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.
  • the network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (WI-FI) interface).
  • WI-FI wireless fidelity
  • the memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory.
  • RAM Random Access Memory
  • NVM Non-Volatile Memory
  • the memory 1005 may also be a storage device independent of the foregoing processor 1001.
  • the structure shown in FIG. 1 does not constitute a limitation on the terminal, and may include more or less components than shown in the figure, or combine some components, or arrange different components.
  • the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module, and computer readable instructions.
  • the network interface 1004 is mainly used for data communication with a network server; the user interface 1003 is mainly used for data interaction with users; the processor 1001 and the memory 1005 in the terminal of this application can be set in the terminal
  • the terminal invokes the computer-readable instructions stored in the memory 1005 through the processor 1001, and executes the method for recommending resource information based on voice recognition provided in the embodiment of the present application.
  • FIG. 2 is a schematic flowchart of a first embodiment of a method for recommending resource information based on voice recognition in this application.
  • the method for recommending resource information based on speech recognition includes the following steps:
  • Step S10 Obtain an audio clip of the user to be recommended, and analyze the data information of the user to be recommended according to the audio clip; it should be understood that the subject of the method in this embodiment is the terminal, and the user to be recommended is the resource recommender recommendation
  • the data information of the user to be recommended may usually include the age and the region to which the user is to be recommended; it may also include information such as eating habits, current health index, and reaction speed. Among them, by analyzing the audio clips of the user to be recommended, the age of the user to be recommended is determined.
  • the specific analysis method can use conventional methods in the field, or for example: voiced voice segment detection is performed on the audio of each age group in the training set, and then High frequency boost, wavelet packet decomposition, frequency band splicing, and then WPMFC extraction, and then GMM model training, to obtain 8 Gaussian mixture models; for the audio of the recommended user, the voiced sound segment detection is also used, and then the high frequency boost is performed. Decomposition, band splicing, and WPMFC extraction, and then the extracted features are calculated with 8 Gaussian mixture models to obtain the maximum likelihood.
  • the age group corresponding to the GMM model with the maximum likelihood is the recognition result.
  • the age group is usually divided into four stages: children, youth, middle age and old age.
  • the region to which the user to be recommended belongs is determined.
  • the specific analysis method can use conventional methods in the field, or for example: construct a voice model and collect voice data of dialects in various regions, including each type The unique words of the dialect and their pronunciation, and then through the recognition system, various variants of the words are recognized, and the region to which the user to be recommended belongs is determined according to the recognized dialect.
  • the data information includes the region to which the user to be recommended belongs; the acquisition of audio clips of the user to be recommended, and the analysis of the data information of the user to be recommended according to the audio clips Before the steps, include:
  • the step of the user’s data information includes obtaining the audio clips of the user to be recommended and various variants of words; according to the voice model, the obtained audio clips and various variants of the words, it is determined that the user to be recommended belongs to Area.
  • Step S20 Determine the resource type applicable to the user to be recommended according to the data information
  • the resource types include return-type resources and consumption-type resources. In other embodiments, they may also be classified according to other rules.
  • the resources can be insurance products. Take insurance products as an example below. Returned resources are also called savings resources, that is, after the resource has survived for the agreed period, the resource company has to return the premium paid or the expected conversion value of the resource usage specified in the contract; Consumption resources are a kind of consumption resources, that is, the user (insurant) signs a contract with the resource company (resource owner). If a resource accident as stipulated in the contract occurs within the agreed time, the resource company will compensate or compensate according to the originally agreed amount. If there is no resource accident within the agreed time, the resource company will not refund the premium paid.
  • the resource type of the user to be recommended is determined.
  • the user to be recommended belongs to the single stage (18-25 years old, this stage has just entered the society and is in the early stage of business development)
  • this type of user is usually suitable for purchase; if Determine that the user to be recommended belongs to the stage of starting a family (25-45 years old, this stage will face raising children and elderly people).
  • Such users are usually young and rich in income, and they are suitable for buying back resources; if it is determined that the user to be recommended belongs to retirement planning Stage (45-60 years old) or asset inheritance stage (60+), these users are suitable for buying consumer resources because they are older; they can also be adjusted according to the user characteristics targeted by the resources company's related products.
  • Step S30 Determine each resource name corresponding to the resource type and the fitness function corresponding to the resource type according to the resource type;
  • consumer resources usually include consumer accident insurance, consumer medical insurance, consumer critical illness insurance, and consumer life insurance, which can be adjusted according to the specific product launches of the resource company; return resources It usually includes life insurance, pension, education fund, etc., which can be adjusted according to the specific products launched by the resource company.
  • the resource as an insurance product as an example.
  • the resource names corresponding to the consumer resource include consumer accident insurance, consumer medical resources, consumer critical illness insurance, and consumer life insurance
  • the resource name corresponding to the return type resource includes products such as life insurance, pension, education fund, etc.
  • Genetic Algorithm is a computational model that simulates the biological evolution process of natural selection and genetic mechanism of Darwin's biological evolution theory, and is a method of searching for the optimal solution by simulating the natural evolution process.
  • the genetic algorithm starts from a population that represents the possible potential solution set of the problem, and a population is composed of a certain number of individuals coded by genes. Each individual is actually an entity with chromosome characteristics.
  • the fitness function is a guarantee for solving the optimal solution or the suboptimal solution, and according to the resource type, the fitness function corresponding to the resource type and the resource type is determined.
  • Step S40 update the preset genetic algorithm model according to the determined fitness function
  • resource types correspond to different fitness functions, and correspondingly, different resource types correspond to different genetic algorithm models.
  • Step S50 Use the name of each resource as an input parameter of the genetic algorithm model and run the genetic algorithm model to obtain the name of the resource to be recommended;
  • the solving steps of the genetic algorithm are as follows: 1) Initialize the population; 2) Calculate the fitness value of each individual in the population; 3) Select according to a certain rule determined by the individual fitness value to enter the next generation Individual; 4) Crossover operation according to probability Pc; 5) Mutation operation according to probability Pc; 6) If a certain stopping condition is not met, then switch to no switch 2), otherwise go to the next step; 7) Output the fitness value in the population
  • the optimal chromosome is used as the satisfactory or optimal solution of the problem.
  • the determined resource names are used as the input parameters of the genetic algorithm model and the genetic algorithm model is run to obtain the resource names to be recommended;
  • the resource names can be coded using conventional binary codes or The corresponding number can be formulated for each resource name;
  • the initial population can be randomly generated initial chromosomes based on the determined resource name and the premium of the resource name; then genetic operations (selection operations, crossover operations, mutation operations) are performed, and the optimal solution is finally obtained , The optimal resource plan.
  • Step S60 Send the resource information corresponding to the name of the resource to be recommended to the user to be recommended.
  • the resource information corresponding to the resource name to be recommended is sent to the user to be recommended.
  • the resource information corresponding to the resource name to be recommended may also be Send to the person who implements the recommended resource, such as a salesperson.
  • This application obtains the audio clips of the users to be recommended, analyzes the data information of the users to be recommended according to the audio clips, and then determines the resource types applicable to the users to be recommended according to the data information, and then according to the resources Type, determine each resource name corresponding to the resource type and the fitness function corresponding to the resource type, update a preset genetic algorithm model according to the determined fitness function, and use the resource name as the Input parameters of the genetic algorithm model and run the genetic algorithm model to obtain the name of the resource to be recommended; sending the resource information corresponding to the resource name to be recommended to the user to be recommended can quickly determine the resource suitable for the user, thereby improving resource information recommendation Accuracy.
  • FIG. 3 is a schematic flowchart of a second embodiment of a method for recommending resource information based on speech recognition in this application.
  • the data information includes the region and age of the user to be recommended.
  • the step S20 includes:
  • Step S201 Determine the resource type applicable to the user to be recommended according to the region and age of the user to be recommended, as well as the preset corresponding relationship between the region, age and the resource type.
  • FIG. 4 is a schematic flowchart of a third embodiment of a method for recommending resource information based on speech recognition in this application.
  • the data information includes the region and age of the user to be recommended.
  • the step S20 includes:
  • Step S201' Determine the rich-poor index of the region according to the region to which the user to be recommended belongs;
  • the wealth index of a region usually corresponds to its spending power. According to the region to which the user to be recommended belongs, determining the rich-poor index of the region may be obtained by obtaining the financial index of the region, or may be determined by other methods.
  • Step S202' Determine the life stage of the user to be recommended according to the wealth index and the age of the user to be recommended;
  • Step S203' Determine the resource type applicable to the user to be recommended according to the determined life stage and the preset correspondence between the life stage and the resource type.
  • the corresponding relationship between life stages and resource types is as follows: Single stage (18-25 years old, this stage has just entered the society and is in the early stage of career development), this type of user is usually suitable for purchase and consumption Type; the stage of starting a family and establishing a business (25-45 years old, this stage will face raising children, elderly people), these users are usually young and rich in income, suitable for buying and returning resources; retirement planning stage (45-60 years old) or assets In the inheritance stage (60+), these users are suitable for buying consumer resources due to their older age.
  • FIG. 5 is a schematic flowchart of a fourth embodiment of a method for recommending resource information based on speech recognition in this application.
  • the resource type is a consumer resource; in this embodiment, before step S10, it further includes:
  • Step S01 Establish a genetic algorithm model, where the fitness function of the genetic algorithm model is:
  • M i is the expected conversion value of resource use (in the case of an insurance product, it can be the insurance amount)
  • T i the effective period of resource use (in the case of an insurance product, it can be the insurance period)
  • N i is the conversion value of the resource acquisition (take insurance products, for example, be)
  • the parameters in the fitness function of the genetic algorithm model are usually determined according to the resource name corresponding to the resource type.
  • consumption-type resources have consumption-type accident insurance
  • consumption-type accident insurance corresponds to resource usage expectations
  • the conversion value, the effective period of the resource use, the conversion value of the resource acquisition, etc. may be set according to the specific provisions of the resource company for the specific resource name.
  • FIG. 6 is a schematic flowchart of a fifth embodiment of a method for recommending resource information based on speech recognition in this application.
  • the resource type is a return type resource; in this embodiment, before the step S10, the method further includes: step S01': establishing a genetic algorithm model, wherein the fitness of the genetic algorithm model The function is:
  • M'i is the expected conversion value of resource usage (in the case of an insurance product, it can be the insurance amount)
  • T i ' is the effective period of resource usage (in the case of an insurance product, it can be the insurance period)
  • N i ' is the resource Obtain the conversion value
  • I'(x) is the actual conversion value of the resource use (taking insurance products as an example, it can be compensation Quota)
  • Rs' is the probability of resource return.
  • the parameters in the fitness function of the genetic algorithm model are usually determined according to the resource name corresponding to the resource type.
  • the resource is an insurance product.
  • the return type resource has education fund insurance, and the education
  • the expected conversion value of resource use, the effective period of resource use, and the conversion value of resource acquisition corresponding to golden insurance can be set according to the specific regulations of the resource company for specific insurance products. The following methods can be used to determine the probability of resource return:
  • the data information includes the eating habits of the user to be recommended and the current health index; after the step of determining each resource name corresponding to the resource type and the fitness function corresponding to the resource type according to the resource type , The method further includes the following steps:
  • the method further includes:
  • the life prediction model may be:
  • X represents the eating habits of the user to be recommended (in this embodiment, vegetarians, Y is 1, and carnivores, Y is 0.5), Y represents the current health index, and ⁇ 1 and ⁇ 2 are X and Y respectively Weight. According to the life prediction value, the resource return probability of each resource name is determined.
  • the data information also includes the reaction speed of the user to be recommended
  • the step of determining the life prediction value of the user to be recommended according to the eating habits, current health index and a preset life prediction model of the user to be recommended includes:
  • human response speed is a measure of intelligence
  • intelligence is an indicator of the "system integrity" of the human body.
  • People with higher IQ tend to live longer. The eating habits are light, the nutrition is reasonable, and the higher the current health index, the longer life will be.
  • the method before the step of determining the life prediction value of the user to be recommended according to the eating habits, reaction speed, current health index, and a preset life prediction model of the user to be recommended, the method further includes: establishing a life prediction Model; wherein, the life prediction model may be:
  • X represents the eating habits of the user to be recommended (in this embodiment, vegetarians, Y is 1, and carnivores, Y is 0.5), Y represents the current health index, and Z represents the response speed (usually the response rate Speed normalization processing, usually 0 to 1), ⁇ 1, ⁇ 2, ⁇ 3 are the weights of X, Y, Z respectively.
  • the embodiment of the present application also proposes a storage medium, and the computer-readable storage medium may be a non-volatile readable storage medium.
  • the storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by a processor, the steps of the method for recommending resource information based on voice recognition as described above are realized.
  • Fig. 7 is a structural block diagram of a first embodiment of an apparatus for recommending resource information based on speech recognition in this application.
  • the device for recommending resource information based on voice recognition proposed in the embodiment of the present application includes:
  • the obtaining module 701 obtains an audio clip of the user to be recommended, and analyzes the data information of the user to be recommended according to the audio clip;
  • the user to be recommended is the user to whom the resource recommender recommends the resource name.
  • the data information of the user to be recommended can usually include the age of the user to be recommended, the region to which they belong, etc.; it can also include eating habits, current health index, and response speed. And other information. Among them, by analyzing the audio clips of the user to be recommended, the age of the user to be recommended is judged.
  • the specific analysis method can use conventional methods in the field, or for example:
  • voiced voice segment detection, high frequency boost, wavelet packet decomposition, frequency band splicing, WPMFC extraction, and GMM model training are used to obtain 8 Gaussian mixture models; the same for the audio of the recommended user Firstly, voiced voice segment detection, then high frequency boost, wavelet packet decomposition, frequency band splicing, WPMFC extraction, and then the likelihood of the extracted features and 8 Gaussian mixture models is calculated to obtain the maximum likelihood, with the maximum likelihood
  • the age group corresponding to Randu's GMM model is the recognition result, and the age group is usually divided into four stages: children, youth, middle-aged and old.
  • the region to which the user to be recommended belongs is determined.
  • the specific analysis method can use conventional methods in the field, or for example: construct a voice model and collect voice data of dialects in various regions, including each type The unique words of the dialect and their pronunciation, and then through the recognition system, various variants of the words are recognized, and the region to which the user to be recommended belongs is determined according to the recognized dialect.
  • the determining module 702 is configured to determine the resource type applicable to the user to be recommended according to the data information
  • the resource types include return-type resources and consumption-type resources. In other embodiments, they may also be classified according to other rules.
  • the resources can be insurance products. Take insurance products as an example below. Returned resources are also called savings resources, that is, after the resource has survived for the agreed period, the resource company has to return the premium paid or the expected conversion value of the resource usage specified in the contract; Consumption resources are a kind of consumption resources, that is, the user (insurant) signs a contract with the resource company (resource owner). If a resource accident as stipulated in the contract occurs within the agreed time, the resource company will compensate or compensate according to the originally agreed amount. If there is no resource accident within the agreed time, the resource company will not refund the premium paid.
  • the resource type of the user to be recommended is determined.
  • the user to be recommended belongs to the single stage (18-25 years old, this stage has just entered the society and is in the early stage of business development)
  • this type of user is usually suitable for purchase; if Determine that the user to be recommended belongs to the stage of starting a family (25-45 years old, this stage will face raising children and elderly people).
  • Such users are usually young and rich in income, and they are suitable for buying back resources; if it is determined that the user to be recommended belongs to retirement planning Stage (45-60 years old) or asset inheritance stage (60+), these users are suitable for buying consumer resources because they are older; they can also be adjusted according to the user characteristics targeted by the resources company's related products.
  • the selection module 703 is configured to determine each resource name corresponding to the resource type and the fitness function corresponding to the resource type according to the resource type;
  • consumer resources usually include consumer accident insurance, consumer medical insurance, consumer critical illness insurance, and consumer life insurance, which can be adjusted according to the specific products launched by the resource company; return resources It usually includes life insurance, pension, education fund, etc., which can be adjusted according to the specific products launched by the resource company.
  • the resource types corresponding to the consumer resource include consumer accident insurance, consumer medical resources, consumer critical illness insurance, consumer life insurance and other products ;
  • the resource name corresponding to the return type resource includes products such as life insurance, pension, education fund resources, etc.
  • Genetic Algorithm is a computational model that simulates the biological evolution process of natural selection and genetic mechanism of Darwin's biological evolution theory, and is a method of searching for the optimal solution by simulating the natural evolution process.
  • the genetic algorithm starts from a population that represents the possible potential solution set of the problem, and a population is composed of a certain number of individuals coded by genes. Each individual is actually an entity with chromosome characteristics.
  • the fitness function is a guarantee for solving the optimal solution or the suboptimal solution, and according to the resource type, the fitness function corresponding to the resource type and the resource type is determined.
  • the update module 704 updates the preset genetic algorithm model according to the determined fitness function
  • resource types correspond to different fitness functions, and correspondingly, different resource types correspond to different genetic algorithm models.
  • the calculation module 705 is configured to use the name of each resource as an input parameter of the genetic algorithm model and run the genetic algorithm model to obtain the name of the resource to be recommended;
  • the solving steps of the genetic algorithm are as follows: 1) Initialize the population; 2) Calculate the fitness value of each individual in the population; 3) Select according to a certain rule determined by the individual fitness value to enter the next generation Individual; 4) Crossover operation according to probability Pc; 5) Mutation operation according to probability Pc; 6) If a certain stopping condition is not met, then switch to no switch 2), otherwise go to the next step; 7) Output the fitness value in the population
  • the optimal chromosome is regarded as the satisfactory solution or optimal solution of the problem.
  • the determined resource names are used as the input parameters of the genetic algorithm model and the genetic algorithm model is run to obtain the resource names to be recommended;
  • the resource names can be coded using conventional binary codes or The corresponding number can be formulated for each resource name;
  • the initial population can be randomly generated initial chromosomes based on the determined resource name and the premium of the resource name; then genetic operations (selection operations, crossover operations, mutation operations) are performed, and the optimal solution is finally obtained , The optimal resource plan.
  • the recommendation module 706 sends the resource information corresponding to the name of the resource to be recommended to the user to be recommended. It should be understood that, in this embodiment, the resource information corresponding to the resource name to be recommended is sent to the user to be recommended. In other embodiments, the resource information corresponding to the resource name to be recommended may also be Send to the person who implements the recommended resource, such as a salesperson.
  • This application obtains the audio clips of the users to be recommended, analyzes the data information of the users to be recommended according to the audio clips, and then determines the resource types applicable to the users to be recommended according to the data information, and then according to the resources Type, determine each resource name corresponding to the resource type and the fitness function corresponding to the resource type, update a preset genetic algorithm model according to the determined fitness function, and use the resource name as the Input parameters of the genetic algorithm model and run the genetic algorithm model to obtain the name of the resource to be recommended; sending the resource information corresponding to the resource name to be recommended to the user to be recommended can quickly determine the resource suitable for the user, thereby improving resource information recommendation Accuracy.
  • the device for recommending resource information based on speech recognition in the present application reference may be made to the foregoing method embodiments, which will not be repeated here.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

La présente invention relève du domaine technique de l'intelligence artificielle. L'invention concerne un procédé et un appareil de recommandation d'informations de ressources et un appareil basé sur la reconnaissance vocale, un terminal et un support. Le procédé consiste à : acquérir un clip audio d'un utilisateur auquel recommander, et analyser, selon le clip audio, des informations de données de l'utilisateur auquel recommander ; ensuite, déterminer, en fonction des informations de données, un type de ressource approprié à l'utilisateur auquel recommander ; puis déterminer, selon le type de ressource, chaque nom de ressource correspondant au type de ressource et une fonction de conformité correspondant au type de ressource ; mettre à jour un modèle d'algorithme génétique prédéfini en fonction de la fonction de conformité déterminée ; prendre chaque nom de ressource en tant que paramètre d'entrée du modèle d'algorithme génétique et exécuter le modèle d'algorithme génétique de façon à obtenir le nom d'une ressource à recommander ; et envoyer des informations de ressource correspondant au nom de la ressource à recommander à l'utilisateur auquel recommander. Des ressources qui sont appropriées pour un utilisateur peuvent être rapidement déterminées, ce qui permet d'améliorer la précision de recommandation d'informations de ressources.
PCT/CN2019/120891 2019-06-19 2019-11-26 Procédé et appareil de recommandation d'informations de ressources basés sur la reconnaissance vocale, et terminal et support WO2020253109A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910532919.8 2019-06-19
CN201910532919.8A CN110379427A (zh) 2019-06-19 2019-06-19 基于语音识别的资源信息推荐方法、装置、终端及介质

Publications (1)

Publication Number Publication Date
WO2020253109A1 true WO2020253109A1 (fr) 2020-12-24

Family

ID=68249312

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/120891 WO2020253109A1 (fr) 2019-06-19 2019-11-26 Procédé et appareil de recommandation d'informations de ressources basés sur la reconnaissance vocale, et terminal et support

Country Status (2)

Country Link
CN (1) CN110379427A (fr)
WO (1) WO2020253109A1 (fr)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110379427A (zh) * 2019-06-19 2019-10-25 深圳壹账通智能科技有限公司 基于语音识别的资源信息推荐方法、装置、终端及介质
CN110390047A (zh) * 2019-06-19 2019-10-29 深圳壹账通智能科技有限公司 基于遗传算法的资源信息推荐方法、装置、终端及介质
CN110992928A (zh) * 2019-11-26 2020-04-10 维沃移动通信有限公司 音频处理方法及终端设备
CN113516533A (zh) * 2021-06-24 2021-10-19 平安科技（深圳）有限公司 基于改进bert模型的产品推荐方法、装置、设备及介质
CN114117236A (zh) * 2021-12-07 2022-03-01 广州道然信息科技有限公司 基于智能终端的用户交互方法、装置、设备和存储介质

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1991976A (zh) * 2005-12-31 2007-07-04 潘建强 基于音素的语音识别方法与系统
US20110161119A1 (en) * 2009-12-24 2011-06-30 The Travelers Companies, Inc. Risk assessment and control, insurance premium determinations, and other applications using busyness
CN105096938A (zh) * 2015-06-30 2015-11-25 百度在线网络技术（北京）有限公司 一种用于获取用户的用户特征信息的方法和装置
CN108665369A (zh) * 2018-03-30 2018-10-16 北京有保无险科技有限公司 一种基于用户的保险评估方法和系统
CN108959618A (zh) * 2018-07-18 2018-12-07 北京欣欣苹果网络科技有限公司 互联网信息收集及处理方法和装置
CN109300054A (zh) * 2018-11-27 2019-02-01 泰康保险集团股份有限公司 保险产品推荐方法、装置、服务器及存储介质
CN109559221A (zh) * 2018-11-20 2019-04-02 中国银行股份有限公司 基于用户数据的催收方法、装置和存储介质
CN110379427A (zh) * 2019-06-19 2019-10-25 深圳壹账通智能科技有限公司 基于语音识别的资源信息推荐方法、装置、终端及介质
CN110390047A (zh) * 2019-06-19 2019-10-29 深圳壹账通智能科技有限公司 基于遗传算法的资源信息推荐方法、装置、终端及介质

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1991976A (zh) * 2005-12-31 2007-07-04 潘建强 基于音素的语音识别方法与系统
US20110161119A1 (en) * 2009-12-24 2011-06-30 The Travelers Companies, Inc. Risk assessment and control, insurance premium determinations, and other applications using busyness
CN105096938A (zh) * 2015-06-30 2015-11-25 百度在线网络技术（北京）有限公司 一种用于获取用户的用户特征信息的方法和装置
CN108665369A (zh) * 2018-03-30 2018-10-16 北京有保无险科技有限公司 一种基于用户的保险评估方法和系统
CN108959618A (zh) * 2018-07-18 2018-12-07 北京欣欣苹果网络科技有限公司 互联网信息收集及处理方法和装置
CN109559221A (zh) * 2018-11-20 2019-04-02 中国银行股份有限公司 基于用户数据的催收方法、装置和存储介质
CN109300054A (zh) * 2018-11-27 2019-02-01 泰康保险集团股份有限公司 保险产品推荐方法、装置、服务器及存储介质
CN110379427A (zh) * 2019-06-19 2019-10-25 深圳壹账通智能科技有限公司 基于语音识别的资源信息推荐方法、装置、终端及介质
CN110390047A (zh) * 2019-06-19 2019-10-29 深圳壹账通智能科技有限公司 基于遗传算法的资源信息推荐方法、装置、终端及介质

Also Published As

Publication number Publication date
CN110379427A (zh) 2019-10-25

Similar Documents

Publication Publication Date Title
WO2020253109A1 (fr) Procédé et appareil de recommandation d'informations de ressources basés sur la reconnaissance vocale, et terminal et support
Rios-Avila Recentered influence functions (RIFs) in Stata: RIF regression and RIF decomposition
Chen et al. Uncovering the importance of relationship characteristics in social networks: Implications for seeding strategies
Plieninger et al. Validity of multiprocess IRT models for separating content and response styles
US8775332B1 (en) Adaptive user interfaces
González-Val et al. New evidence on Gibrat’s law for cities
Ballings et al. Social media optimization: Identifying an optimal strategy for increasing network size on Facebook
Goldfarb et al. Are all managers created equal?
WO2020253354A1 (fr) Procédé et appareil de recommandation d'informations de ressource basés sur un algorithme génétique, terminal et support
US20110258045A1 (en) Inventory management
Garzia et al. Party placement in supranational elections: An introduction to the euandi 2014 dataset
Bursztyn et al. When product markets become collective traps: The case of social media
Nguyen Ethnic gaps in child education outcomes in Vietnam: an investigation using Young Lives data
Qi et al. Subjective well-being measurement based on Chinese grassroots blog text sentiment analysis
McFowland III et al. A prescriptive analytics framework for optimal policy deployment using heterogeneous treatment effects.
CN115375177A (zh) 用户价值评估方法、装置、电子设备及存储介质
CN115577316A (zh) 一种基于多模态数据融合的用户人格预测方法与应用
US20240009575A1 (en) Ethical ai development platform and methods for use therewith
Deng et al. An empirical analysis of factors influencing users' adoption and use of mobile services in China
EP2613275B1 (fr) Dispositif de recherche, procédé de recherche, programme de recherche et support de mémoire lisible par un ordinateur destiné à contenir un programme de recherche
Méndez-Prado et al. An Assessment Tool to Identify the Financial Literacy Level of Financial Education Programs Participants’ Executed by Ecuadorian Financial Institutions
CN116091113A (zh) 营销模型数据处理方法、系统及计算机可读存储介质
CN111538909A (zh) 一种信息推荐方法及装置
Mahalingam et al. Predicting financial savings decisions using sigmoid function and information gain ratio
Richardson et al. Age weights for health services derived from the relative social willingness-to-pay instrument

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19933507

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19933507

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 30.03.2022)

122 Ep: pct application non-entry in european phase

Ref document number: 19933507

Country of ref document: EP

Kind code of ref document: A1