CN110751553A - Identification method and device of potential risk object, terminal equipment and storage medium - Google Patents
Identification method and device of potential risk object, terminal equipment and storage medium Download PDFInfo
- Publication number
- CN110751553A CN110751553A CN201911020263.8A CN201911020263A CN110751553A CN 110751553 A CN110751553 A CN 110751553A CN 201911020263 A CN201911020263 A CN 201911020263A CN 110751553 A CN110751553 A CN 110751553A
- Authority
- CN
- China
- Prior art keywords
- risk
- preset
- potential
- voice
- mining model
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 101
- 238000005065 mining Methods 0.000 claims abstract description 102
- 238000011156 evaluation Methods 0.000 claims description 66
- 238000012549 training Methods 0.000 claims description 21
- 238000000605 extraction Methods 0.000 claims description 16
- 230000004044 response Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 230000036651 mood Effects 0.000 claims description 8
- 238000013136 deep learning model Methods 0.000 claims description 6
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 2
- 238000004422 calculation algorithm Methods 0.000 description 9
- 230000006854 communication Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
-
- 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
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
-
- 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
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
- G10L15/063—Training
-
- 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
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a method, a device, terminal equipment and a computer readable storage medium for identifying a potential risk object, wherein each voice feature in voice data of an object to be identified is extracted and is input into a preset risk mining model; outputting warning information for warning the object to be identified of potential risks according to the output result of the preset risk mining model; and continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object. The method and the device avoid the situation that the potential risk is difficult to find or easy to pass by the staff, and improve the identification accuracy and identification efficiency of the object with the potential risk.
Description
Technical Field
The present invention relates to the field of Fintech (financial technology), and in particular, to a method and an apparatus for identifying a potential risk object, a terminal device, and a computer-readable storage medium.
Background
Generally, before lending for individuals or other small enterprises, lending enterprises (such as banks, loan companies and the like) check information of lending objects such as individuals or other small enterprises which need to be borrowed currently by manually making a telephone call, and determine to give the lending objects to loan under the condition that the information of the lending objects is verified to be true and accurate, so as to control the lending risk of the lending enterprises and avoid bad account rate.
However, currently, a worker responsible for verifying information of a lending object generally asks a series of questions with standard answers to the lending object, and because the worker lacks the digging ability for lying on the questions of the lending object to cover real information, the worker can only find the risk of borrowing the current lending object when the answer of the lending object based on the questions is obviously inconsistent with the standard answers, so that the lending enterprise is difficult to fully distribute risk points in the process of loan to individuals or other small and micro enterprises, and the accuracy and efficiency of identifying potential risk objects are low.
Disclosure of Invention
The invention mainly aims to provide a method and a device for identifying a potential risk object, a terminal device and a computer readable storage medium, and aims to solve the technical problems of low accuracy and low efficiency of identifying the potential risk object by existing workers.
In order to achieve the above object, the present invention provides a method for identifying a risk potential object, including:
extracting each voice feature in the voice data of the object to be recognized, and inputting each voice feature into a preset risk mining model;
outputting warning information for warning the object to be identified of potential risks according to the output result of the preset risk mining model;
and continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
Further, the speech features include at least: mood characteristics, speech rate characteristics, and coherence characteristics,
the step of extracting each voice feature in the voice data of the object to be recognized and inputting each voice feature into a preset risk mining model comprises the following steps:
extracting tone features, speech speed features and coherence features from the voice data of the object to be recognized, and carrying out vectorization processing on the tone features, the speech speed features and the coherence features;
and inputting the voice characteristics, the speed characteristics and the coherence characteristics after vectorization into the preset risk mining model so that the preset risk mining model can judge the evaluation probability of lie of the object to be identified and output the evaluation probability as a result.
Further, the step of outputting warning information for warning the object to be recognized of the existence of the potential risk according to the output result of the preset risk mining model includes:
acquiring the evaluation probability output by the preset risk mining model, and detecting whether the evaluation probability exceeds a preset threshold value;
and when the evaluation probability is detected to exceed the preset threshold value, outputting warning information for warning the object to be identified of potential risk.
Further, before the step of extracting each speech feature in the speech data of the object to be recognized and inputting each speech feature into the preset risk mining model, the method further comprises:
acquiring voice data of the object to be recognized for answering a preset question;
the step of obtaining the voice data of the object to be recognized answering the preset question comprises the following steps:
monitoring the conversation process of voice conversation between preset staff and the object to be identified in real time;
and acquiring voice data of the object to be recognized in response to the preset questions from the call process.
Further, the step of continuing to detect voice data identifying the object to be identified based on the alert information to determine whether the object to be identified is a potentially risky object includes:
detecting other preset questions related to the preset questions of the warning information identification so that preset workers can follow the questions of the object to be recognized;
acquiring the evaluation probability obtained by judging each voice feature through the preset risk mining model in the voice data of the object to be recognized for answering each other preset question;
when the evaluation probability is detected to exceed the predetermined threshold, determining that the object to be identified is a potential risk object.
Further, before the step of extracting each speech feature in the speech data of the object to be recognized and inputting each speech feature into the preset risk mining model, the method further comprises:
and training and constructing the preset risk mining model according to the recorded data which are determined as the potential risk objects.
Further, the step of training and constructing the preset risk mining model according to the recorded data determined as the potential risk object includes:
acquiring pre-stored recorded data which is determined to be the potential risk object;
extracting corresponding voice features from the obtained recording data to serve as training data;
calling a preset learning model to train based on the training data to obtain the preset risk mining model, wherein the preset learning model comprises: machine learning models and deep learning models.
In addition, to achieve the above object, the present invention provides an apparatus for identifying a risk potential object, including:
the extraction module is used for extracting each voice feature in the voice data of the object to be recognized and inputting each voice feature into a preset risk mining model;
the information output module is used for outputting warning information for warning the object to be recognized of potential risks according to the output result of the preset risk mining model;
and the identification determination module is used for continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
The present invention also provides a terminal device, including: a memory, a processor and a procedure for identifying a risk potential object stored on the memory and executable on the processor, the procedure for identifying a risk potential object when executed by the processor implementing the steps of the method for identifying a risk potential object as described above.
The present invention also provides a computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program, which, when being executed by a processor, carries out the steps of the method for identifying a risk potential object as described above.
According to the identification method, the identification device, the terminal equipment and the computer readable storage medium of the potential risk object, disclosed by the invention, each voice feature in the voice data of the object to be identified is extracted, and each voice feature is input into a preset risk mining model; outputting warning information for warning the object to be identified of potential risks according to the output result of the preset risk mining model; and continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object. According to the method, when a worker needs to judge whether an object to be recognized is a potential risk object and the information of the object to be recognized is verified in a manual calling mode, voice features are extracted from the voice data of the object to be recognized, which are acquired in advance, through an existing voice feature extraction algorithm, and then the voice features are input into a trained preset risk mining model, so that an output result is obtained through judgment according to the voice features of the preset risk mining model, and warning information for warning that the current object to be recognized possibly has potential risks is output according to the output result, so that the worker can further detect and recognize the voice data of the object to be recognized based on the warning information, and whether the object to be recognized is the potential risk object is confirmed. The method and the device have the advantages that the trained risk mining model is used for judging the voice characteristics of the voice data of the lending object in the process of answering the questions asked by the workers according to the object to be recognized, so that the risk points of the object to be recognized in answering the questions asked by the workers are fully mined, the warning information is output in real time according to the condition of the found risk points, the current object to be recognized is detected and recognized based on the emphasis of the warning information, the condition that the potential risks are difficult to find or are easy to put through by the workers is avoided, and the recognition accuracy and the recognition efficiency of the object with the potential risks are improved.
Drawings
FIG. 1 is a schematic diagram of the hardware operation involved in an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first embodiment of a method for identifying potential risk objects according to the present invention;
FIG. 3 is a flowchart illustrating a detailed process of step S300 according to an embodiment of the method for identifying a risk potential object of the present invention;
FIG. 4 is a schematic diagram of an application scenario of an embodiment of a method for identifying a risk potential object according to the present invention;
fig. 5 is a schematic structural diagram of a device for identifying a risk potential object according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic structural diagram of a hardware operating environment according to an embodiment of the present invention.
It should be noted that fig. 1 is a schematic structural diagram of a hardware operating environment of the terminal device. The terminal equipment of the embodiment of the invention can be terminal equipment such as a PC, a portable computer and the like.
As shown in fig. 1, the terminal device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the terminal device configuration shown in fig. 1 is not intended to be limiting of the terminal device and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a distributed task processing program. Among them, the operating system is a program that manages and controls the hardware and software resources of the sample terminal device, a handler that supports distributed tasks, and the execution of other software or programs.
In the terminal apparatus shown in fig. 1, the user interface 1003 is mainly used for data communication with each terminal; the network interface 1004 is mainly used for connecting a background server and performing data communication with the background server; and processor 1001 may be configured to invoke an identification procedure for a potential risk object stored in memory 1005 and perform the following operations:
extracting each voice feature in the voice data of the object to be recognized, and inputting each voice feature into a preset risk mining model;
outputting warning information for warning the object to be identified of potential risks according to the output result of the preset risk mining model;
and continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
Further, processor 1001 may invoke an identification procedure for a potential risk object stored in memory 1005, and also perform the following operations:
extracting tone features, speech speed features and coherence features from the voice data of the object to be recognized, and carrying out vectorization processing on the tone features, the speech speed features and the coherence features;
and inputting the voice characteristics, the speed characteristics and the coherence characteristics after vectorization into the preset risk mining model so that the preset risk mining model can judge the evaluation probability of lie of the object to be identified and output the evaluation probability as a result.
Further, processor 1001 may invoke an identification procedure for a potential risk object stored in memory 1005, and also perform the following operations:
acquiring the evaluation probability output by the preset risk mining model, and detecting whether the evaluation probability exceeds a preset threshold value;
and when the evaluation probability is detected to exceed the preset threshold value, outputting warning information for warning the object to be identified of potential risk.
Further, the processor 1001 may call the recognition program of the risk potential object stored in the memory 1005, and perform the following operations before performing the operations of extracting each speech feature in the speech data of the object to be recognized and inputting each speech feature into the preset risk mining model:
and acquiring voice data of the object to be recognized for answering a preset question.
Further, processor 1001 may invoke an identification procedure for a potential risk object stored in memory 1005, and also perform the following operations:
monitoring the conversation process of voice conversation between preset staff and the object to be identified in real time;
and acquiring voice data of the object to be recognized in response to the preset questions from the call process.
Further, processor 1001 may invoke an identification procedure for a potential risk object stored in memory 1005, and also perform the following operations:
detecting other preset questions related to the preset questions of the warning information identification so that preset workers can follow the questions of the object to be recognized;
acquiring the evaluation probability obtained by judging each voice feature through the preset risk mining model in the voice data of the object to be recognized for answering each other preset question;
when the evaluation probability is detected to exceed the predetermined threshold, determining that the object to be identified is a potential risk object.
Further, the processor 1001 may call the recognition program of the risk potential object stored in the memory 1005, and perform the following operations before performing the operations of extracting each speech feature in the speech data of the object to be recognized and inputting each speech feature into the preset risk mining model:
and training and constructing the preset risk mining model according to the recorded data which are determined as the potential risk objects.
Further, processor 1001 may invoke an identification procedure for a potential risk object stored in memory 1005, and also perform the following operations:
acquiring pre-stored recorded data which is determined to be the potential risk object;
extracting corresponding voice features from the obtained recording data to serve as training data;
calling a preset learning model to train based on the training data to obtain the preset risk mining model, wherein the preset learning model comprises: machine learning models and deep learning models.
Based on the above structure, various embodiments of the method for identifying a risk potential object of the present invention are provided.
Referring to fig. 2, fig. 2 is a flowchart illustrating a method for identifying a risk potential object according to a first embodiment of the present invention.
While a logical order is shown in the flow chart, in some cases, the steps shown or described may be performed in an order different than presented herein.
The method for identifying the potential risk object according to the embodiment of the present invention is applied to the terminal device, and the terminal device according to the embodiment of the present invention may be a terminal device such as a PC or a portable computer, which is not limited specifically herein.
In order to facilitate understanding of the method for identifying a potential risk object of the present invention, in this embodiment, a lending enterprise (for example, a bank, a loan company, etc.) identifies a lending object such as a person or other small enterprise that needs to be borrowed currently by using a manual telephone call before lending the person or other small enterprise, so as to determine whether the lending object is an application scenario of a potential risk object with a potential risk.
The identification method of the potential risk object comprises the following steps:
step S100, extracting each voice feature in the voice data of the object to be recognized, and inputting each voice feature into a preset risk mining model.
After voice data when an acquired loan object (for convenience of description, in the following embodiments, "loan object" is used as an "object to be recognized" in the present invention) answers questions of a worker of a loan enterprise is answered, an existing voice feature extraction algorithm is immediately called to extract various voice features from the voice data, that is, at least a voice characteristic, a voice speed characteristic and a coherence characteristic are extracted, and after the extracted various voice features are subjected to preset processing, the extracted various voice features are input into a preset risk mining model obtained by training voice features in voice data of a loan object which is determined whether to be paid or not by the loan enterprise, so that the preset risk mining model judges the magnitude of evaluation probability that the loan object lies when answers the questions of the worker based on the various voice features in the voice data of the loan object at present, and outputting the probability obtained by the judgment as a result.
It should be noted that, in this embodiment, a speech feature extraction algorithm is invoked, and from speech data of a lending object when answering a question posed by a lending enterprise, extracted speech features at least include a speech characteristic, a speech speed characteristic, and a coherence characteristic, and it should be noted that the method for identifying a potential risk object of the present invention does not limit the type, the number, and the like of the speech features extracted from the speech data.
It should be noted that, in this embodiment, when it is detected that the lending object starts to answer the question of the staff of the lending enterprise, that is, the voice feature extraction algorithm is called to extract the voice feature from the voice data of the lending object, it should be understood that, in the method for identifying a potential risk object of the present invention, the time for collecting the voice feature in the voice data of the lending object is not limited to the time for collecting the voice feature after the voice data is collected or while the voice data is collected, or the voice data is not collected but the voice feature is directly extracted.
In this embodiment, only based on the voice characteristics of the lending object in response to the questions asked by the staff of the lending enterprise, the evaluation probability that the lending object lies in response to the questions asked by the staff is determined, and the difference between whether the content of the lending object in response to the questions and the standard correct answer to the questions is given to the staff for determination, so that the overall operation pressure for determining whether the lending object lies in response is saved, and even under the condition that the current language content similarity determination is not completely mature, the accuracy of determining the consistency or similarity between the response content and the standard correct answer based on the independence of the staff is ensured.
Further, in another embodiment, before extracting each speech feature in the speech data of the object to be recognized and inputting each speech feature into the preset risk mining model in step S100, the method for recognizing a potential risk object of the present invention further includes:
step S400, acquiring voice data of the object to be recognized answering the preset question.
The method comprises the steps that a worker of a lending enterprise asks a lending object based on a voice call, and therefore voice data of the lending object when the lending object answers to preset questions posed by the worker are collected in real time in the process of verifying information of the lending object according to whether the answer of the lending object is the same as or similar to standard answers of the questions.
It should be noted that: the preset questions are a series of standard questions which are provided to the lending object by a worker in charge of information verification based on a voice call, each standard question has a corresponding standard answer, so that when the worker judges that the answer of the lending object to each proposed standard question is close to or consistent with the standard answer, the fact that the information of the lending object identified by the current standard question is true is determined, that is, the loan of the current lending object can be determined, and otherwise, the loan of the current lending object is refused.
Further, step S400 includes:
step S401, monitoring the conversation process of voice conversation between preset staff and the object to be identified in real time.
The method comprises the steps that a worker of a lending enterprise communicates with a lending object based on a voice call mode, so that the whole process of verifying information of the lending object is monitored in real time in the process of carrying out voice call between the worker of the lending enterprise and the lending object.
It should be noted that the preset staff is a staff in charge of verifying the loan object information of the loan enterprise, for example, a manual customer service.
Step S402, collecting voice data of the object to be recognized when answering for each preset question from the communication process.
In the process of a voice call between a staff of a loan enterprise and a loan object, when the loan object is detected to answer a preset question which is provided by the staff and used for verifying the information of the loan object, voice data of the loan object when the loan object answers currently is collected.
Specifically, for example, in the application scenario shown in fig. 4, in the process of performing a voice call between a worker of a lending enterprise and a lending object in order to verify information of the lending object to determine whether to loan the lending object, the worker (i.e., an artificial customer service) presents a standard question with a standard correct answer to the lending object, so that when it is detected that the lending object answers based on the currently presented question, voice data of the lending object at the current time is collected based on an existing voice extraction technology.
And S200, outputting warning information for warning the object to be recognized to have the potential risk according to the output result of the preset risk mining model.
After the preset risk mining model judges the evaluation probability of lying when the lending object answers questions asked by workers of the lending enterprise according to the tone features, the speech speed features and the coherence feature speech features when the lending object answers the questions asked by the workers of the lending enterprise, the preset risk mining model outputs the probability obtained by judgment as a calculation result of the current model, obtains the evaluation probability output by the preset risk mining model, detects whether the evaluation probability exceeds a preset threshold value, and outputs warning information to warn that the lending object possibly has lending risks when the evaluation probability exceeds the preset threshold value.
In the extracted voice data of the loan object when answering the preset questions made by the staff of the loan enterprise, inputting each voice characteristic into a preset risk mining model for the preset risk mining model to judge the evaluation probability of the current loan object when answering the preset questions based on each voice characteristic, outputting the judged probability as a result, judging whether the probability exceeds a preset threshold (namely whether the probability value is greater than the preset threshold), prompting the staff to continue to ask the current loan object when judging that the probability does not exceed the preset threshold (namely whether the probability value is less than or equal to the preset threshold), and determining that the information of the current loan object is true when a series of standard questions are asked completely and the evaluation probability of the current loan object when answering all questions is not greater than the preset threshold, if the evaluation probability of lying when the current lending object answers any question exceeds a preset threshold (namely the probability value is greater than the preset threshold), warning information for reminding the staff that the potential risk of lending may exist in the current lending object is output to the staff immediately.
Further, in step S200, outputting warning information for warning the object to be recognized of the existence of the potential risk according to the output result of the preset risk mining model includes:
step S201, obtaining the evaluation probability output by the preset risk mining model, and detecting whether the evaluation probability exceeds a predetermined threshold.
After the preset risk mining model judges the evaluation probability of lying when the lending object answers questions asked by workers of the lending enterprise according to the tone features, the speech speed features and the coherence feature speech features when the lending object answers the questions asked by the workers of the lending enterprise, the preset risk mining model outputs the judged probability as a calculation result of the current model, and acquires and detects whether the evaluation probability output by the preset risk mining model exceeds a preset threshold value.
It should be noted that the predetermined threshold is a threshold which is set in advance by a worker in charge of verifying the information of the loan object by the loan enterprise according to practical experience in the actual working process of judging the authenticity of the information of the loan object and is used for determining that the loan object lies, and it should be understood that the method for identifying the potential risk object of the present invention does not limit the specific value of the predetermined threshold.
Specifically, for example, in the application scenario shown in fig. 4, when the preset risk mining model judges that the lending object answers the questions asked by the workers of the lending enterprise according to the tone features, the speech speed features and the coherence features of the lending object when answering the questions posed by the workers of the lending enterprise, and after the probability obtained by the judgment is output as a calculation result of the current model, immediately judging whether the evaluation probability of lying when the loan object output by the current preset risk mining model answers questions of staff of the loan enterprise is greater than that of the staff of the loan enterprise who is responsible for verifying the information of the loan object, and presetting a threshold value for determining that the loan object is lying according to practical experience in the actual working process of judging the authenticity of the information of the loan object.
Step S201, when the evaluation probability is detected to exceed the preset threshold value, outputting warning information for warning the object to be identified of potential risk.
Specifically, for example, in the application scenario shown in fig. 4, when it is detected and determined that the evaluation probability of lying when the lending object output by the current preset risk mining model answers the question of the worker of the lending enterprise is greater than the threshold value set by the worker in advance, the worker is output warning information to remind the worker that the lending object is suspected to lie when answering the question currently posed by the worker, so as to continue to ask the lending object according to the currently posed question, and further determine whether the information of the current lending object is true according to the evaluation probability of lying when the current lending object answers the question of the worker, so as to decide that the lending can be performed or the lending should be refused.
Further, in another embodiment, it is determined that the lending object may be credited upon detecting that the evaluation probability does not exceed the predetermined threshold.
Specifically, for example, in the application scenario shown in fig. 4, when it is detected and determined that the evaluation probability of lying when the lending object output by the current preset risk mining model answers the question of the worker of the lending enterprise is less than or equal to the threshold value preset by the worker in charge of verifying the information of the lending object, it is determined that the lending object does not lie when answering the question currently posed by the worker, so as to prompt the worker that the worker can continue to ask the current lending object, and when the worker finishes asking the question by using the prepared series of standard questions and the current preset risk mining model assists the worker to detect and determine that the evaluation probability of lying when the lending object answers all questions does not exceed the threshold value, it is determined that the information of the current lending object is true, that is, there is no potential risk, a deposit may be made.
The embodiment prompts the staff to continuously ask questions of the current loan object by judging whether the probability exceeds a preset threshold value or not, determines the information of the current loan object to be true when a series of standard questions are asked completely and the evaluation probability of lying when the current loan object answers all the questions does not exceed the preset threshold value, can pay, outputs prompt to the staff for the staff to continuously ask questions if the evaluation probability of lying when the current loan object answers any question exceeds the preset threshold value, determines the information of the current loan object not to be true when the evaluation probability of lying when the current loan object answers the questions of the staff is detected to still exceed the preset threshold value, and should refuse to loan, improves the efficiency of identifying the risk of the staff when the current loan object answers the question, the bad account rate of enterprise lending is reduced.
Step S300, voice data for identifying the object to be identified is continuously detected based on the warning information so as to determine whether the object to be identified is a potential risk object.
After warning information for warning that the currently lending object of the worker possibly has lending risks is output based on the evaluation probability output by the preset risk mining model, the worker continuously asks for the currently lending object based on the warning information and obtains voice data when the currently lending object answers the voice data to perform further voice feature detection evaluation on the voice data, and when the evaluation probability that the currently lending object answers the asking of the worker still exceeds a preset threshold value through detection and evaluation, the information of the currently lending object is determined not to be true, namely, potential risks of lending exist, and the worker should refuse to make payment.
In the embodiment, before a lending enterprise lends a personal or other small and micro enterprises and the like, in the process of verifying information of the lending object by a worker in a manual telephone making mode, voice data of the lending object in response to a preset question provided by the worker is collected in real time, and each voice feature is extracted from the collected voice data of the lending object through an existing voice feature extraction algorithm, so that each voice feature is input into a trained preset risk mining model, an output result is obtained by judging according to each voice feature according to the preset risk mining model, and whether the lending enterprise pays the lending object at present is determined. The method and the device have the advantages that the voice characteristics of the voice data of the loan object are judged by using the trained risk mining model in the process of answering the questions posed by the loan enterprise staff according to the loan object, so that the risk points of the loan object when answering questions of the staff are sufficiently found, and the accuracy and the recognition efficiency of recognizing the target object with loan risk in the current loan behavior by the staff are improved.
According to the method for identifying the potential risk object, whether the object to be identified is the potential risk object or not is determined by a worker, and in the process of verifying the information of the object to be identified through a manual calling mode, voice features are extracted from the pre-collected voice data of the object to be identified through an existing voice feature extraction algorithm, so that the voice features are input into a trained preset risk mining model, an output result is obtained through judgment according to the voice features of the preset risk mining model, and warning information for warning that the current object to be identified possibly has the potential risk is output according to the output result, so that the worker can further detect and identify the voice data of the object to be identified based on the warning information, and whether the object to be identified is the potential risk object or not is determined. The method and the device have the advantages that the trained risk mining model is used for judging the voice characteristics of the voice data of the lending object in the process of answering the questions asked by the workers according to the object to be recognized, so that the risk points of the object to be recognized in answering the questions asked by the workers are fully mined, the warning information is output in real time according to the condition of the found risk points, the current object to be recognized is detected and recognized based on the emphasis of the warning information, the condition that the potential risks are difficult to find or are easy to put through by the workers is avoided, and the recognition accuracy and the recognition efficiency of the object with the potential risks are improved.
Further, based on the first embodiment of the method for identifying a risk potential object, a second embodiment of the method for identifying a risk potential object of the present invention is provided.
In a second embodiment of the method for identifying a risk potential object according to the present invention, before extracting each speech feature in the speech data of the object to be identified and inputting each speech feature into the preset risk mining model in step S100, the method for identifying a risk potential object according to the present invention further includes:
and step A, training and constructing the preset risk mining model according to the recording data which are determined to be the potential risk objects.
Based on the pre-stored voice record data corresponding to each loan object determined to be placed by the loan enterprise and the loan object determined to refuse to be placed, voice characteristics of the loan object determined to be placed and voice characteristics of the loan object determined to refuse to be placed are established, and a preset risk mining model for judging the lying probability of the loan object according to the voice characteristics of the loan object when the loan object answers questions posed by workers of the loan enterprise is constructed.
Further, step a, comprises:
step A1, obtaining pre-stored recorded data that has been determined to be the potential risk object.
Recording data of a predetermined number of loan objects for which loan is already determined to be paid and recording data of the same number of loan objects for which loan is already determined to be rejected are extracted from recording data of all loan objects (i.e., potential risk objects and potential risk objects) for which loan is already determined to be paid and for which loan is already determined to be rejected by checking loan object information by a worker from a previously stored loan enterprise.
And A2, extracting corresponding voice features from all the acquired recording data to serve as training data.
And respectively extracting the corresponding voice characteristics of the loan object determined to put money and the voice characteristics of the loan object determined to reject money from the extracted voice data of the loan object determined to put money and the voice data of the loan object determined to reject money by calling a voice characteristic extraction algorithm, and processing and integrating the extracted corresponding voice characteristics into a training data set.
Step A3, calling a preset learning model to train based on the training data to obtain the preset risk mining model.
And calling the existing machine learning model or deep learning model, and training on a training data set obtained by processing and integrating the extracted voice characteristics of the loan object determined to be paid and the voice characteristics of the loan object determined to refuse to be paid so as to obtain the voice characteristics when the loan object answers questions provided by loan enterprise staff and judge the preset risk mining model of the lying probability of the loan object.
It should be noted that the preset learning model at least includes one of a machine learning model and a deep learning model, and the method for identifying the potential risk object of the present invention does not limit the type and model of the learning model.
Further, in the step S100, extracting each speech feature in the speech data of the object to be recognized, and inputting each speech feature into the preset risk mining model includes:
step S101, extracting the tone features, the speech speed features and the coherence features from the voice data of the object to be recognized, and carrying out vectorization processing on the tone features, the speech speed features and the coherence features.
Specifically, for example, in the application scenario shown in fig. 4, after real-time voice data of the lending object in response to a question asked by a worker of the lending enterprise is acquired, an existing voice feature extraction algorithm is called to extract various voice features such as a mood feature, a speech speed feature, and a coherence feature when the lending object answers the question asked by the worker from the voice data, and after the various mood features, the speech speed features, and the coherence feature voice features are extracted, vectorization processing is performed on the various mood features, the speech speed features, and the coherence feature voice features so as to convert the various mood features, the speech speed features, and the coherence feature voice features into a data form that can be recognized by a machine.
Step S102, inputting the voice characteristics, the speed characteristics and the coherence characteristics after vectorization into the preset risk mining model, so that the preset risk mining model judges the evaluation probability of lie of the object to be identified and outputs the evaluation probability as a result.
Specifically, for example, in the application scenario shown in fig. 4, in the voice data obtained when the lending object collected in real time answers to the questions of the workers of the lending enterprise, the extracted tone features, speech rate features and coherence features are vectorized, after the voice characteristics of each tone, the speed characteristics and the coherence characteristics are correspondingly converted into the data form which can be recognized by the machine, inputting each tone feature, speech speed feature and coherence feature speech feature into a pre-set risk mining model which is constructed in advance, so that in the pre-set risk mining model, and judging the evaluation probability of lying when the loan object answers the questions asked by the staff of the loan enterprise according to the tone features, the speech speed features and the coherence feature speech features when the loan object answers the questions asked by the staff of the loan enterprise.
In this embodiment, based on the pre-stored recording data corresponding to the loan object determined to make a loan and the loan object determined to reject the loan, the voice feature of the loan object determined to make a loan and the voice feature of the loan object determined to reject the loan are constructed, a preset risk mining model for determining the lying probability of the loan object according to the voice feature of the loan object when answering the questions posed by the workers of the loan enterprise is constructed, after the real-time voice data of the loan object when answering the questions asked by the workers of the loan enterprise is collected, the voice feature extraction algorithm is immediately invoked to extract various voice features such as the mood feature, the pace feature, the coherence feature and the like when the loan object answers the questions asked by the workers of the loan enterprise, and after the mood feature, the pace feature and the coherence feature are extracted, vectorizing each of the tone features, the speech speed features and the coherence features to convert the tone features, the speech speed features and the coherence features into data forms recognizable by the machine, and inputting the data forms recognizable by the machine into a preset risk mining model constructed in advance, so that in the preset risk mining model, the evaluation probability of lying when the lending object answers questions asked by workers of the lending enterprise is judged according to the tone features, the speech speed features and the coherence features when the lending object answers the questions posed by the workers of the lending enterprise.
The method and the device realize that the evaluation probability of lying of the loan object when answering the question proposed by the worker is judged according to the voice characteristics such as the tone characteristic, the speech speed characteristic and the coherence characteristic when the loan object answers the question proposed by the worker in the process of answering the question proposed by the loan enterprise worker, so that the potential risk points can be sufficiently found in the information verification process before the loan object is paid by the worker, the efficiency of recognizing the risk when the worker debits and credits the current loan object is improved, and the bad account rate of the enterprise loan is reduced.
Further, based on the first and second embodiments of the identification method of a risk potential object, a third embodiment of the identification method of a risk potential object of the present invention is proposed.
In a third embodiment of the method for identifying a risk potential object according to the present invention, in the step S300, the step of continuously detecting the voice data for identifying the object to be identified based on the warning information to determine whether the object to be identified is a risk potential object includes:
step S301, detecting other preset questions related to the preset questions identified by the warning information so that the preset staff can follow the questions of the object to be recognized.
When the evaluation probability of lying when the loan object output by the current preset risk mining model answers questions of workers of the loan enterprise is greater than a threshold value which is set by the workers of the loan enterprise in advance and is responsible for verifying loan object information, warning information is output to the workers to remind the workers that the loan object is suspected to lie when answering the questions currently proposed by the workers, and meanwhile, other questions which are mutually proved or mutually influenced with the currently proposed questions are found out from a series of standard questions which are prepared by the workers to be proposed to the loan object based on voice communication in advance, and the workers are prompted to follow up the loan object based on the found other questions.
Step S302, obtaining the evaluation probability of each voice feature judged by the preset risk mining model in the voice data of the object to be recognized for answering each other preset question.
Step S303, when the evaluation probability is detected to exceed the preset threshold value, determining that the object to be identified is a potential risk object.
Specifically, for example, in the application scenario shown in fig. 4, when it is further detected that the question for the worker to pursue the question according to the lending object based on the preset risk mining model is answered, voice characteristics such as tone characteristics, speech speed characteristics and coherence characteristics in the voice data, when the estimated probability of lying when the judged and output loan object answers to the pursuit of the staff of the loan enterprise still exceeds a threshold value (namely the probability value is greater than the threshold value) preset by the staff, warning information is continuously output to the staff, therefore, the assistant staff determines whether the information of the current lending object is true or not by integrating the evaluation probability output by the current model, the question asked and the content answered by the lending object for the question asked, and refuses to pay the lending object under the condition that the information is determined not to be true.
Further, in another embodiment, when it is detected that the evaluation probability does not exceed the predetermined threshold, the preset question and/or the other preset questions are recorded for the worker to continuously determine whether to put money to the lending object.
Specifically, for example, in the application scenario shown in fig. 4, when it is further detected that the preset risk mining model is used to answer the question asked by the lending object for the worker according to the preset risk mining model, and the voice characteristics such as the voice characteristics, the speed characteristics, and the coherence characteristics in the voice data, and when it is determined that the evaluation probability of lying when the lending object answers the question asked by the worker of the lending enterprise does not exceed the threshold value (i.e. the probability value is less than or equal to the threshold value) set by the worker in advance, it is determined that the lending object does not lie when answering the current question asked by the worker, so as to prompt the worker to ask the current lending object, and record the question asked by the worker and the content of answering the question answered by the lending object, so that after the process of verifying the information of the lending object by using the voice call method is ended, the loan company acquires relevant information through other channels to verify the information of the loan object identified by the question asked by the staff, and determines to put money on the loan object after the verification is correct.
In this embodiment, when the assessment probability of lying when the loan object output by the preset risk mining model answers questions of the staff of the loan enterprise exceeds a threshold value set by the staff in advance, warning information is output to the staff to remind the staff that the loan object currently has a suspicion of lying, other problems which are mutually adjudicated or mutually influenced with the currently proposed problems are found out, the staff is prompted to trace the loan object based on the found other problems, corresponding problems are recorded to be confirmed through other channels, it is ensured that the loan object cannot cover real information through lying, the defect that the staff lacking risk mining ability cannot find the loan object to lie to borrow is avoided, and the bad account rate is further reduced.
In addition, referring to fig. 5, an embodiment of the present invention further provides a device for identifying a potential risk object, where the device for identifying a potential risk object includes:
the extraction module is used for extracting each voice feature in the voice data of the object to be recognized and inputting each voice feature into a preset risk mining model;
the information output module is used for outputting warning information for warning the object to be recognized of potential risks according to the output result of the preset risk mining model;
and the identification determination module is used for continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
Preferably, the extraction module comprises:
the first extraction unit is used for extracting the tone features, the speech speed features and the coherence features from the voice data of the object to be recognized and carrying out vectorization processing on the tone features, the speech speed features and the coherence features;
and the input unit is used for inputting the voice characteristics, the speed characteristics and the coherence characteristics after vectorization processing into the preset risk mining model so that the preset risk mining model can judge the evaluation probability of lie of the object to be identified and output the evaluation probability as a result.
Preferably, the information input module includes:
the first obtaining unit is used for obtaining the evaluation probability output by the preset risk mining model and detecting whether the evaluation probability exceeds a preset threshold value;
and the output unit is used for outputting warning information for warning the object to be identified of potential risk when the evaluation probability is detected to exceed the preset threshold value.
Preferably, the identification apparatus for risk potential object of the present invention further comprises:
and the acquisition module is used for acquiring voice data of the object to be recognized for answering a preset question.
Preferably, the acquisition module comprises:
the monitoring unit is used for monitoring the conversation process of voice conversation between preset workers and the object to be recognized in real time;
and the acquisition unit is used for acquiring voice data of the object to be recognized in response to the preset questions from the call process.
Preferably, the identification determination module comprises:
the detection unit is used for detecting other preset questions related to the preset questions of the warning information identification so that the preset staff can follow up the questions of the object to be recognized;
a second obtaining unit, configured to obtain, from the speech data of the object to be recognized that answers to each of the other preset questions, the evaluation probability that is obtained by determining, by the preset risk mining model, each speech feature;
a determining unit, configured to determine that the object to be identified is a potential risk object when it is detected that the evaluation probability exceeds the predetermined threshold.
Preferably, the identification apparatus for risk potential object of the present invention further comprises:
and the construction module is used for training and constructing the preset risk mining model according to the sound recording data which are determined as the potential risk objects.
Preferably, the building block comprises:
a third acquiring unit configured to acquire pre-stored sound recording data that has been determined to be the potential risk object;
the second extraction unit is used for extracting corresponding voice features from the acquired sound recording data to serve as training data;
a calling unit, configured to call a preset learning model to train based on the training data to obtain the preset risk mining model, where the preset learning model includes: machine learning models and deep learning models.
The steps of the method for identifying a potential risk object described above are implemented when each functional module of the apparatus for identifying a potential risk object provided in this embodiment runs, and are not described herein again.
Furthermore, an embodiment of the present invention further provides a computer-readable storage medium, which is applied to a computer and may be a non-volatile computer-readable storage medium, where an identification program of a risk potential object is stored on the computer-readable storage medium, and when executed by a processor, the identification program of a risk potential object implements the steps of the identification method of a risk potential object as described above.
The steps implemented when the program for identifying a risk potential object running on the processor is executed may refer to various embodiments of the method for identifying a risk potential object of the present invention, and are not described herein again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A method for identifying a risk potential object, the method comprising:
extracting each voice feature in the voice data of the object to be recognized, and inputting each voice feature into a preset risk mining model;
outputting warning information for warning the object to be identified of potential risks according to the output result of the preset risk mining model;
and continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
2. The method for identifying a risk potential object of claim 1, wherein the speech features include at least: mood characteristics, speech rate characteristics, and coherence characteristics,
the step of extracting each voice feature in the voice data of the object to be recognized and inputting each voice feature into a preset risk mining model comprises the following steps:
extracting tone features, speech speed features and coherence features from the voice data of the object to be recognized, and carrying out vectorization processing on the tone features, the speech speed features and the coherence features;
and inputting the voice characteristics, the speed characteristics and the coherence characteristics after vectorization into the preset risk mining model so that the preset risk mining model can judge the evaluation probability of lie of the object to be identified and output the evaluation probability as a result.
3. The method for identifying an object at risk according to claim 2, wherein the step of outputting warning information for warning the object to be identified of risk according to the output result of the preset risk mining model comprises:
acquiring the evaluation probability output by the preset risk mining model, and detecting whether the evaluation probability exceeds a preset threshold value;
and when the evaluation probability is detected to exceed the preset threshold value, outputting warning information for warning the object to be identified of potential risk.
4. The method for identifying objects at risk as set forth in claim 1, wherein before the step of extracting each speech feature in the speech data of the object to be identified and inputting each speech feature into a preset risk mining model, the method further comprises:
acquiring voice data of the object to be recognized for answering a preset question;
the step of obtaining the voice data of the object to be recognized answering the preset question comprises the following steps:
monitoring the conversation process of voice conversation between preset staff and the object to be identified in real time;
and acquiring voice data of the object to be recognized in response to the preset questions from the call process.
5. The method for identifying an object at risk according to any one of claims 1 to 4, wherein the step of continuing to detect the voice data for identifying the object to be identified based on the warning information to determine whether the object to be identified is an object at risk comprises:
detecting other preset questions related to the preset questions of the warning information identification so that preset workers can follow the questions of the object to be recognized;
acquiring the evaluation probability obtained by judging each voice feature through the preset risk mining model in the voice data of the object to be recognized for answering each other preset question;
when the evaluation probability is detected to exceed the predetermined threshold, determining that the object to be identified is a potential risk object.
6. The method for identifying a risk potential object as claimed in claim 1, wherein before the step of extracting each speech feature in the speech data of the object to be identified and inputting each speech feature into a preset risk mining model, the method further comprises:
and training and constructing the preset risk mining model according to the recorded data which are determined as the potential risk objects.
7. The method for identifying risk potential objects of claim 6, wherein the step of training and constructing the pre-set risk mining model based on the recorded data that has been determined to be the risk potential object comprises:
acquiring pre-stored recorded data which is determined to be the potential risk object;
extracting corresponding voice features from the obtained recording data to serve as training data;
calling a preset learning model to train based on the training data to obtain the preset risk mining model, wherein the preset learning model comprises: machine learning models and deep learning models.
8. An apparatus for identifying a risk potential object, the apparatus comprising:
the extraction module is used for extracting each voice feature in the voice data of the object to be recognized and inputting each voice feature into a preset risk mining model;
the information output module is used for outputting warning information for warning the object to be recognized of potential risks according to the output result of the preset risk mining model;
and the identification determination module is used for continuously detecting voice data for identifying the object to be identified based on the warning information so as to determine whether the object to be identified is a potential risk object.
9. A terminal device, characterized in that the terminal device comprises: memory, a processor and a procedure for identifying a risk potential object stored on the memory and executable on the processor, the procedure for identifying a risk potential object, when executed by the processor, implementing the steps of the method for identifying a risk potential object according to any one of claims 1 to 7.
10. A storage medium, characterized in that the storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the method of identification of a risk potential object according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911020263.8A CN110751553A (en) | 2019-10-24 | 2019-10-24 | Identification method and device of potential risk object, terminal equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911020263.8A CN110751553A (en) | 2019-10-24 | 2019-10-24 | Identification method and device of potential risk object, terminal equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110751553A true CN110751553A (en) | 2020-02-04 |
Family
ID=69279860
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911020263.8A Pending CN110751553A (en) | 2019-10-24 | 2019-10-24 | Identification method and device of potential risk object, terminal equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110751553A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111539741A (en) * | 2020-05-15 | 2020-08-14 | 支付宝(杭州)信息技术有限公司 | Method, system and non-transitory storage medium for identifying operational risk |
| CN112348520A (en) * | 2020-10-21 | 2021-02-09 | 上海淇玥信息技术有限公司 | XGboost-based risk assessment method and device and electronic equipment |
| CN113807953A (en) * | 2021-09-24 | 2021-12-17 | 重庆富民银行股份有限公司 | Wind control management method and system based on telephone return visit |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20120131923A (en) * | 2011-05-27 | 2012-12-05 | 권용득 | Method for voice-recognition and method for analysing risk using it |
| WO2017185862A1 (en) * | 2016-04-28 | 2017-11-02 | 腾讯科技（深圳）有限公司 | Method, apparatus and device for identifying malicious call and establishing identification model |
| CN107680602A (en) * | 2017-08-24 | 2018-02-09 | 平安科技（深圳）有限公司 | Voice fraud recognition methods, device, terminal device and storage medium |
| CN109543516A (en) * | 2018-10-16 | 2019-03-29 | 深圳壹账通智能科技有限公司 | Signing intention judgment method, device, computer equipment and storage medium |
| CN109544166A (en) * | 2018-11-05 | 2019-03-29 | 阿里巴巴集团控股有限公司 | A kind of Risk Identification Method and device |
| CN109559761A (en) * | 2018-12-21 | 2019-04-02 | 广东工业大学 | A kind of risk of stroke prediction technique based on depth phonetic feature |
| CN110263326A (en) * | 2019-05-21 | 2019-09-20 | 平安科技（深圳）有限公司 | A kind of user's behavior prediction method, prediction meanss, storage medium and terminal device |
-
2019
- 2019-10-24 CN CN201911020263.8A patent/CN110751553A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20120131923A (en) * | 2011-05-27 | 2012-12-05 | 권용득 | Method for voice-recognition and method for analysing risk using it |
| WO2017185862A1 (en) * | 2016-04-28 | 2017-11-02 | 腾讯科技（深圳）有限公司 | Method, apparatus and device for identifying malicious call and establishing identification model |
| CN107680602A (en) * | 2017-08-24 | 2018-02-09 | 平安科技（深圳）有限公司 | Voice fraud recognition methods, device, terminal device and storage medium |
| CN109543516A (en) * | 2018-10-16 | 2019-03-29 | 深圳壹账通智能科技有限公司 | Signing intention judgment method, device, computer equipment and storage medium |
| CN109544166A (en) * | 2018-11-05 | 2019-03-29 | 阿里巴巴集团控股有限公司 | A kind of Risk Identification Method and device |
| CN109559761A (en) * | 2018-12-21 | 2019-04-02 | 广东工业大学 | A kind of risk of stroke prediction technique based on depth phonetic feature |
| CN110263326A (en) * | 2019-05-21 | 2019-09-20 | 平安科技（深圳）有限公司 | A kind of user's behavior prediction method, prediction meanss, storage medium and terminal device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111539741A (en) * | 2020-05-15 | 2020-08-14 | 支付宝(杭州)信息技术有限公司 | Method, system and non-transitory storage medium for identifying operational risk |
| CN112348520A (en) * | 2020-10-21 | 2021-02-09 | 上海淇玥信息技术有限公司 | XGboost-based risk assessment method and device and electronic equipment |
| CN113807953A (en) * | 2021-09-24 | 2021-12-17 | 重庆富民银行股份有限公司 | Wind control management method and system based on telephone return visit |
| CN113807953B (en) * | 2021-09-24 | 2023-11-03 | 重庆富民银行股份有限公司 | Wind control management method and system based on telephone return visit |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110751553A (en) | Identification method and device of potential risk object, terminal equipment and storage medium | |
| CN107566358A (en) | A kind of Risk-warning reminding method, device, medium and equipment | |
| CN105975514A (en) | Automatic quality testing method and system | |
| CN111553701A (en) | Session-based risk transaction determination method and device | |
| CN113191787A (en) | Telecommunication data processing method, device electronic equipment and storage medium | |
| CN113596844A (en) | Early warning method, device, medium and electronic equipment based on data information | |
| CN107958215A (en) | A kind of antifraud recognition methods, device, server and storage medium | |
| CN111445928A (en) | Voice quality inspection method, device, equipment and storage medium | |
| CN110796553A (en) | Service request processing method, device, terminal and storage medium | |
| CN110166634A (en) | User identity assists in identifying method, apparatus, equipment and readable storage medium storing program for executing | |
| CN111160928A (en) | Identity verification method and device | |
| CN110689443A (en) | Insurance data processing method and device, storage medium and server | |
| CN114978474A (en) | Method and system for automatically handling user chat risk level | |
| CN115034886A (en) | Default risk prediction method and device | |
| CN114971017A (en) | Bank transaction data processing method and device | |
| CN113642462A (en) | Driving behavior assessment method and device, terminal equipment and storage medium | |
| CN112000862B (en) | Data processing method and device | |
| CN115393077B (en) | Data processing method based on loan transaction man-machine conversation system and related device | |
| CN109711984B (en) | Pre-loan risk monitoring method and device based on collection urging | |
| CN107154996B (en) | Incoming call interception method and device, storage medium and terminal | |
| CN115271932A (en) | Outbound risk identification method and device | |
| CN111601000B (en) | Communication network fraud identification method and device and electronic equipment | |
| CN113222303A (en) | Method and device for predicting out-of-band risk of bank signature card | |
| CN111080444A (en) | Information auditing method and device | |
| CN113011971B (en) | Risk measure method, apparatus, system and computer storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |