CN111241262A - Loan qualification auditing method based on artificial intelligence and related equipment - Google Patents
Loan qualification auditing method based on artificial intelligence and related equipment Download PDFInfo
- Publication number
- CN111241262A CN111241262A CN202010063830.4A CN202010063830A CN111241262A CN 111241262 A CN111241262 A CN 111241262A CN 202010063830 A CN202010063830 A CN 202010063830A CN 111241262 A CN111241262 A CN 111241262A
- Authority
- CN
- China
- Prior art keywords
- user
- parameter
- information table
- loan
- parameter information
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3343—Query execution using phonetics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/334—Query execution
- G06F16/3344—Query execution using natural language analysis
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- General Business, Economics & Management (AREA)
- Human Computer Interaction (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Health & Medical Sciences (AREA)
- Acoustics & Sound (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention discloses a loan qualification auditing method based on artificial intelligence and related equipment, and relates to the field of artificial intelligence, wherein the method comprises the following steps: establishing a parameter information table; carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process; if the identified intention is the informing intention, filling the extracted parameters into the parameter information table; standardizing the parameters in the parameter information table; and evaluating the loan qualification of the user according to the parameter information table. The method has the advantages that the conversation is carried out with the user through the conversation model, the labor cost is reduced, the auditing efficiency is improved, meanwhile, the intention recognition and the parameter extraction are automatically carried out on the user words on the basis, the loan qualification is automatically evaluated according to the parameter information table of the user, the problems of misreading and forgetting of the user information are avoided, and the auditing accuracy is improved.
Description
Technical Field
The invention relates to the field of artificial intelligence, in particular to a loan qualification auditing method and device based on artificial intelligence, computer equipment and a storage medium.
Background
Generally, when a loan application of a customer is received, a credit auditing specialist of a bank needs to communicate with the loan customer by telephone to know the loan intention, property condition and credit condition of the customer, so as to screen the intention customer and evaluate the loan qualification of the customer, thereby helping a back-end seat to make a proper loan scheme for the customer.
The manual review and evaluation method consumes a large amount of labor cost, is low in efficiency, and can affect the effect of communication with users due to the fact that the quality of the personnel is uneven. In addition, in the process of the communication between the credit auditing specialist and the user, the credit auditing specialist needs to quickly record information from multiple aspects of the user, and the situations of error record, omission record and the like of the user information inevitably occur, so that the auditing accuracy is low.
Disclosure of Invention
The embodiment of the invention provides a loan qualification auditing method, a loan qualification auditing device, computer equipment and a storage medium, and aims to solve the problems that in the prior art, the loan qualification auditing method needs to consume a large amount of manpower, and has low efficiency and low auditing accuracy.
In a first aspect, an embodiment of the present invention provides a loan qualification auditing method, including:
establishing a parameter information table;
carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process;
if the identified intention is the informing intention, filling the extracted parameters into the parameter information table;
standardizing the parameters in the parameter information table; and
and evaluating the loan qualification of the user according to the parameter information table.
In a second aspect, an embodiment of the present invention provides a loan qualification auditing apparatus, including:
a parameter information table establishing unit for establishing a parameter information table;
the dialogue unit is used for carrying out dialogue with the user by utilizing the dialogue model and carrying out intention recognition and parameter extraction on the user utterance by utilizing the deep learning model in the dialogue process;
a filling unit for filling the extracted parameter into the parameter information table if the identified intention is the informing intention;
the standardization unit is used for standardizing the parameters in the parameter information table; and
and the evaluation unit is used for evaluating the loan qualification of the user according to the parameter information table.
In a third aspect, an embodiment of the present invention provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the loan qualification auditing method according to the first aspect.
In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the computer program implements the loan qualification auditing method according to the first aspect.
The embodiment of the invention provides a loan qualification auditing method, a loan qualification auditing device, computer equipment and a storage medium, wherein the method comprises the following steps: establishing a parameter information table; carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process; if the identified intention is the informing intention, filling the extracted parameters into the parameter information table; standardizing the parameters in the parameter information table; and evaluating the loan qualification of the user according to the parameter information table. The method has the advantages that the conversation is carried out with the user through the conversation model, the labor cost is reduced, the auditing efficiency is improved, meanwhile, the intention recognition and the parameter extraction are automatically carried out on the user words on the basis, the loan qualification is automatically evaluated according to the parameter information table of the user, the problems of misreading and forgetting of the user information are avoided, and the auditing accuracy is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flow chart of a loan qualification auditing method according to an embodiment of the present invention;
fig. 2 is a sub-flow diagram of a loan qualification auditing method according to an embodiment of the invention;
fig. 3 is another schematic flow chart of a loan qualification auditing method according to an embodiment of the present invention;
fig. 4 is a schematic block diagram of a loan qualification auditing apparatus provided by an embodiment of the invention;
fig. 5 is a schematic block diagram of sub-units of a loan qualification auditing apparatus provided by an embodiment of the invention;
fig. 6 is a schematic block diagram of another sub-unit of the loan qualification auditing device provided by the embodiment of the invention;
FIG. 7 is a schematic block diagram of a computer device provided by an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
Referring to fig. 1, fig. 1 is a schematic flow chart of a loan qualification auditing method according to an embodiment of the present invention, where the method may include steps S101 to S105:
s101, establishing a parameter information table;
in the embodiment of the invention, before the conversation with the user is started, a blank parameter information table is established for the user to record the relevant information of the user and serve as a loan qualification evaluation basis, and the parameter information table can comprise a plurality of parameter name fields and corresponding parameter value fields, and can also comprise standardized parameter value fields. It should be noted that, a standard template may be set for the parameter information table in advance, and then, only the parameter information table needs to be established for each user according to the standard template. Obviously, in the embodiment of the present invention, the mentioned parameter information table is not limited to the conventional table, and may be any form of carrier capable of recording user information, and the form of the carrier is various.
The "parameter name" field contains parameter names related to personal information of the user, loan intention, property condition, credit condition, etc., and is used to indicate the meaning of the parameter, for example, the "name" field is the "parameter name" field, which represents the name of the user. The "parameter value" field corresponds to the "parameter name" field and is used to indicate the specific content of the parameter, for example, the content filled in the "parameter value" field corresponding to the "name" field may be "zhang san".
The 'standardized parameter value' field refers to a parameter value obtained by standardizing the 'parameter value', so that the parameter value can be unified, and subsequent loan qualification evaluation can be conveniently carried out.
It should be noted that all user information is unknown before the session begins, so the "parameter value" and "normalized parameter value" fields are initially empty.
An example of a blank parameter information table is the following table one:
watch 1
In one embodiment, the parameter information table includes a padding parameter and a non-padding parameter. The mandatory filling parameter belongs to the parameter which must be filled, and the parameter needs to be acquired from the user utterance and filled in a 'parameter value' field in a parameter information table; the non-essential filling parameters belong to parameters which are not required to be filled, and can be filled or not filled. In a specific application scenario, the attribute of the "parameter name" field may be preset, for example, the "name" field is a mandatory fill attribute, the "gender" field is a mandatory fill attribute, and the "household" field is an unnecessary fill attribute. The corresponding 'parameter values' can be divided into the required parameters and the non-required parameters, for example, the 'parameter value' field corresponding to the 'name' field is the required parameters.
S102, carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process;
in the step, a dialogue is firstly conducted with a user by using a dialogue model, and the dialogue process can be that the dialogue model asks questions, the user answers, or the user asks, answers or communicates with the user in other modes.
During the conversation process, the deep learning model is continuously utilized to perform intention recognition and parameter extraction on the user utterance so as to timely acquire the real idea of the user and timely acquire the relevant information of the user.
In an embodiment, as shown in fig. 2, the step S102 may include steps S201 to S203:
s201, acquiring a user utterance, and converting the user utterance into text information;
s202, preprocessing the text information;
s203, inputting the preprocessed text information into the deep learning model, identifying intentions in the text information and extracting parameters in the text information.
In this embodiment, after the blank parameter information table is established, a dialog with the user may be started. After the conversation starts, the user utterance is obtained in real time, and then the user utterance is converted into text information by using a speech recognition method. And then preprocessing the text information, specifically, preprocessing the text information includes performing operations such as special character filtering, sentence breaking and word segmentation, and automatic correction of wrong words and word diseases occurring in voice recognition on the text information, for example, in a specific application scenario, preprocessing the text information sequentially includes: the method comprises the steps of filtering special characters, full-angle space replacement, sentence breaking and word segmentation, wherein the step of filtering the special characters refers to the step of filtering some special symbols in text information, and the regular expression can be used for filtering. The full-angle blank replacement means that the half-angle blank can be replaced by the full-angle blank, the punctuation means that the text information is divided into a plurality of short sentences, and the segmentation means that each short sentence is further divided into a plurality of short words. And then inputting the preprocessed text information into a deep learning model, and respectively performing intention identification and parameter extraction, namely identifying the intention in the text information and extracting the parameters in the text information.
The intention recognition refers to recognizing whether the user's intention is to "inform" user information or "ask" loan product information, i.e., confirming whether the user's intention is to "inform" intention or "ask" intention. Specifically, whether the user provides information or inquires about product information can be judged through keywords in the text information, whether corresponding parameter values exist and the integrity of the corresponding parameter values, and if the user's speech is ambiguous, the user can directly ask a question to obtain a positive answer of the user, so that the user's intention is recognized.
The parameter extraction refers to extracting parameters contained in the user words, such as extracting two parameters of 'monthly income' and 'expected loan amount' in the user words.
S103, if the identified intention is the informing intention, filling the extracted parameters into the parameter information table;
when the recognized intention is a "notification" intention, it indicates that the extracted parameter is what the user needs to be notified, so the extracted parameter can be filled in the parameter information table. It should be noted that even in the case that the identified intention is the "inform" intention, the extracted parameters are not necessarily filled in the parameter information table, because the parameters spoken by the user may not belong to any parameter in the parameter information table, in which case the extracted parameters can be directly ignored, but more preferably, a remark column is set in the parameter information table, the parameters are recorded and marked in the remark column, and the parameters are interpreted according to the marked contents, so that the analysis can be performed by adding manual intervention later.
In one embodiment, the loan qualification auditing method further includes: and if the identified intention is an inquiry intention, inquiring in the database according to the extracted parameters, and making an answer according to the inquiry result. When the identified user intent is a "query" intent, indicating that the extracted parameter is the parameter the user is querying, a corresponding answer may be made after querying in the database.
In one embodiment, as mentioned above, the parameter information table includes mandatory parameters and non-mandatory parameters, and the loan qualification auditing method further includes: before the conversation is finished, detecting whether the necessary filling parameters in the parameter information table are filled; if not, the question related to the filling-necessary parameter is provided for the user so as to obtain a corresponding answer and fill the answer into the parameter information table.
In the embodiment, in the multi-turn conversation process with the user, the parameters in the user words are continuously extracted, the parameter information table is updated, and the corresponding parameter values are filled. Specifically, for the "parameter value" field corresponding to the "parameter name" field, if the user does not mention the field in the process of dialogue with the user, the field may be ignored, that is, the "parameter value" field corresponding to the "parameter name" field may be empty. If the user does not mention the field of the 'parameter value' corresponding to the field of the 'parameter name' which must be filled in, the user needs to put forward a corresponding question, extract relevant information from the answer (answer content) of the user, and fill the relevant information into the field of the 'parameter value' corresponding to the field. Until the parameter value field corresponding to the parameter name field is filled completely.
In specific implementation, some fields corresponding to important and critical information may be set as the "parameter name" field to be filled, for example: age, marital status, presence of a house, presence of overdue credit cards, expected loan amount, presence of proper occupation, monthly income, presence of explicit loan use, etc. Fields corresponding to non-essential and reference-only information are set to the non-essential "parameter name" fields, for example: whether a vehicle exists, the place where the house is located, the bank running water in about 6 months, and the like. When the user mentions the expected loan and the relevant conditions of the user, the words of the user can be preprocessed, then the intention of the user is recognized by using a deep learning model, parameters contained in the words of the user are extracted, and meanwhile, a parameter information table is continuously updated by using the extracted parameters. When the user stops the dialogue, according to the completion condition of the parameter field corresponding to the parameter name field, the user asks questions and perfects the parameter field corresponding to the parameter name field until all the parameter values corresponding to the parameter name field are recorded in the parameter information table.
In one embodiment, the loan qualification auditing method further includes: detecting the user utterance, and confirming whether the user has emotion change or not based on the detection result; if yes, judging whether the emotion change reaches a change threshold value; if yes, the manual seat is turned to continue the conversation with the user.
In the processing process (such as preprocessing process) of the user utterance, when the utterance of 'malicious attack' such as dirty words is recognized, the dialogue is stopped, and simultaneously, the dialogue record is saved and is stored in association with the parameter information table of the user or is directly stored in the parameter information table. After the dialogue between the dialogue model and the user is stopped, the dialogue can be transferred to an artificial seat, meanwhile, the parameter information table and the dialogue records are provided for the artificial seat, the artificial seat analyzes service problems in the dialogue process through the parameter information table and the dialogue records, judges the reasons of the problems, further processes the problems, and continues to complete the dialogue with the user. If the user attitude is not good and the conversation is stopped, the manual agent has the right to decide to refuse to provide the loan according to the situation, if the user is in error, the conversation is not stopped, but the interactive conversation is continued, and if the conversation can be smoothly carried out, the conversation is ignored. If the problem is caused by a dialogue model, for example, a user speech recognition error, the dialogue model may be optimized for the problem to improve the dialogue quality of the dialogue model.
In one embodiment, the loan qualification auditing method further includes: and saving the conversation record with the user and associating the conversation record with the parameter information table. Even if the emotion change is not detected or does not reach the change threshold, the conversation record can be still stored after the conversation is finished, and the follow-up tracing is facilitated.
S104, standardizing the parameters in the parameter information table;
while updating the parameter information table, the parameter values in the parameter information table may be normalized and filled in the corresponding "normalized parameter value" field. Specifically, if the parameter value of different parameters is too spoken, it is not beneficial to further analysis and evaluation, so it is necessary to standardize the "parameter value" field of the parameter information table according to the service requirement, and fill the standardized parameter into the "standardized parameter value" field. Specifically, the spoken fields, such as approximate, left-right, top-bottom, and more points in the utterance, may be modified into standardized mathematical symbols. For example, the monthly income "about ten thousand" is standardized as "10,000 ±; one year running water "hundred thousand more" was standardized as "100000 +". Of course, in the present embodiment, the parameter values in the parameter information table may be normalized after the session is completed, that is, the normalization operation may be performed in real time, or may be performed after the session is completed.
And S105, evaluating the loan qualification of the user according to the parameter information table.
In this embodiment, after the session process is finished, the parameter information table is saved, and then the loan qualification of the user is evaluated according to the parameter information table.
In one embodiment, as shown in fig. 3, the step S105 includes steps S301 to S303:
s301, screening out users with loan intention according to the parameter information table;
s302, comparing each parameter of the user with the loan intention with the loan requirement;
and S303, evaluating the loan qualification of the user according to the comparison result, and outputting a loan qualification evaluation result.
In this embodiment, users with loan intentions are screened out according to the parameter information table, and then parameter comparison is performed, so that the loan qualification of the users is evaluated. Specifically, each parameter in the parameter information table is used to confirm whether the user has loan intention, and for example, the parameter information table has a "parameter name" field including: whether the users have loan intention or not and the corresponding field of the parameter value can be yes or no, so that the parameter information table with the field of the parameter value being yes can be screened out, and the corresponding users can be screened out and are all users with loan intention. Then, the parameter information table of the user with loan intention can be compared with the loan requirements one by one, for example, the loan type which can be applied by the user is determined according to each parameter in the parameter information table, the loan requirement of the corresponding loan type is obtained, then, each parameter of the user, such as property information and working conditions, is compared with the conditions in the loan requirement, the loan qualification of the user can be evaluated according to the comparison result, for example, each condition of the user meets the condition of the loan requirement, the user can be determined to apply for loan, and a higher loan is given, or most conditions of the user meet the condition of the loan requirement, the user can be determined to apply for loan, and a medium loan limit is given; or the user can determine that the user can apply for the loan and give a lower loan amount if few conditions of the user meet the conditions of the loan requirement; or the user condition does not meet the condition of the loan requirement, the user can be determined not to apply for the loan, that is, whether the user has loan qualification or not and the user loan amount of the user under the condition that the user has the loan qualification can be determined according to the specific parameter condition.
After the loan qualification of the customer is evaluated, the result of the loan qualification evaluation may be represented in a table form, for example, a customer information table is generated, and various information of the customer and the evaluation result are recorded. An example of a customer information table is the following table two:
watch two
| Personal information | Name: zhang III, sex: for male |
| Asset information | The method comprises the following steps: the car age is: 4 years, vehicle number: 3 ten thousand |
| Loan intention | Expectation limit: 30 ten thousand, loan purpose: fitment house |
| Loan qualification | Can apply for 'car owner loan', highest loan amount: 10 ten thousand |
Namely, according to the property information and the loan intention of the user, the loan qualification of the client can be evaluated as ' applicable ' owner loan ', the highest loan amount: '10 ten thousand'.
The method has the advantages that the conversation is carried out with the user through the conversation model, the labor cost is reduced, the auditing efficiency is improved, meanwhile, the intention recognition and the parameter extraction are automatically carried out on the user words on the basis, the loan qualification is automatically evaluated according to the parameter information table of the user, the problems of misreading and forgetting of the user information are avoided, and the auditing accuracy is improved.
The embodiment of the invention also provides a loan qualification auditing device which is used for executing any embodiment of the loan qualification auditing method. Specifically, referring to fig. 4, fig. 4 is a schematic block diagram of a loan qualification auditing apparatus according to an embodiment of the present invention.
As shown in fig. 4, the loan qualification approval apparatus 400 includes a parameter information table creation unit 401, a dialogue unit 402, a filling unit 403, a normalization unit 404, and an evaluation unit 405.
A parameter information table establishing unit 401, configured to establish a parameter information table;
the parameter information table may include several "parameter name" fields and their corresponding "parameter value" fields, and may also include "standardized parameter value" fields. It should be noted that, a standard template may be set for the parameter information table in advance, and then, only the parameter information table needs to be established for each user according to the standard template. Obviously, in the embodiment of the present invention, the mentioned parameter information table is not limited to the conventional table, and may be any form of carrier capable of recording user information, and the form of the carrier is various.
The "parameter name" field contains parameter names related to personal information of the user, loan intention, property condition, credit condition, etc., and is used to indicate the meaning of the parameter, for example, the "name" field is the "parameter name" field, which represents the name of the user. The "parameter value" field corresponds to the "parameter name" field and is used to indicate the specific content of the parameter, for example, the content filled in the "parameter value" field corresponding to the "name" field may be "zhang san". The 'standardized parameter value' field refers to a parameter value obtained by standardizing the 'parameter value', so that the parameter value can be unified, and subsequent loan qualification evaluation can be conveniently carried out.
A dialogue unit 402, configured to perform dialogue with a user by using a dialogue model, and perform intent recognition and parameter extraction on a user utterance by using a deep learning model during the dialogue process;
the dialog unit 402 uses the dialog model to perform a dialog with the user, and the dialog process may be a question asked by the dialog model, a response answered by the user, a query asked by the user, a response answered by the dialog model, or another communication between the dialog model and the user. During the conversation process, the deep learning model is continuously utilized to perform intention recognition and parameter extraction on the user utterance so as to timely acquire the real idea of the user and timely acquire the relevant information of the user.
In one embodiment, as shown in fig. 5, the dialog unit 402 includes:
a conversion unit 4021 for acquiring a user utterance and converting the user utterance into text information;
the preprocessing unit 4022 is used for preprocessing the text information;
the recognition and extraction unit 4023 is configured to input the preprocessed text information to the deep learning model to recognize an intention in the text information and extract parameters in the text information.
In this embodiment, after the blank parameter information table is established, a dialog with the user may be started. After the conversation starts, the user utterance is obtained in real time, and then the user utterance is converted into text information by using a speech recognition method. And preprocessing the text information, specifically, preprocessing the text information comprises the operations of filtering special characters, segmenting sentences and dividing words, automatically correcting wrong words and word diseases caused by voice recognition and the like. And then inputting the preprocessed text information into a deep learning model, and respectively performing intention identification and parameter extraction, namely identifying the intention in the text information and extracting the parameters in the text information. The intention recognition refers to recognizing whether the user's intention is to "inform" user information or "ask" loan product information, i.e., confirming whether the user's intention is to "inform" intention or "ask" intention. Specifically, whether the user provides information or inquires about product information can be judged through keywords in the text information, whether corresponding parameter values exist and the integrity of the corresponding parameter values. The parameter extraction refers to extracting parameters contained in the user words, such as extracting two parameters of 'monthly income' and 'expected loan amount' in the user words.
A filling unit 403, configured to fill the extracted parameter into the parameter information table if the identified intention is an informing intention;
when the recognized intention is a "notification" intention, it indicates that the extracted parameter is what the user needs to be notified, so the extracted parameter can be filled in the parameter information table.
In one embodiment, the parameter information table includes mandatory parameters and non-mandatory parameters, and the loan qualification auditing device 400 further includes:
the parameter detection unit is used for detecting whether the necessary filling parameters in the parameter information table are filled before the conversation is finished;
and the questioning unit is used for proposing the question associated with the filling-necessary parameters to the user so as to obtain the corresponding answer and fill the answer into the parameter information table.
In the embodiment, in the multi-turn conversation process with the user, the parameters in the user words are continuously extracted, the parameter information table is updated, and the corresponding parameter values are filled. Specifically, for the "parameter value" field corresponding to the "parameter name" field, if the user does not mention the field in the process of dialogue with the user, the field may be ignored, that is, the "parameter value" field corresponding to the "parameter name" field may be empty. If the user does not mention the field of the 'parameter value' corresponding to the field of the 'parameter name' which must be filled in, the user needs to put forward a corresponding question, extract relevant information from the answer (answer content) of the user, and fill the relevant information into the field of the 'parameter value' corresponding to the field. Until the parameter value field corresponding to the parameter name field is filled completely.
In one embodiment, the loan qualification auditing apparatus 400 further comprises:
the emotion detection unit is used for detecting the user utterance and confirming whether the emotion change occurs to the user or not based on the detection result;
the judging unit is used for judging whether the emotion change reaches a change threshold value;
and the manual seat transferring unit is used for transferring the manual seat to continue the conversation with the user.
In the processing process (such as preprocessing process) of the user utterance, when the utterance of 'malicious attack' such as dirty words is recognized, the dialogue is stopped, and simultaneously, the dialogue record is saved and is stored in association with the parameter information table of the user or is directly stored in the parameter information table. After the dialogue between the dialogue model and the user is stopped, the dialogue can be transferred to an artificial seat, meanwhile, the parameter information table and the dialogue records are provided for the artificial seat, the artificial seat analyzes service problems in the dialogue process through the parameter information table and the dialogue records, judges the reasons of the problems, further processes the problems, and continues to complete the dialogue with the user. If the user attitude is not good and the conversation is stopped, the manual agent has the right to decide to refuse to provide the loan according to the situation, if the user is in error, the conversation is not stopped, but the interactive conversation is continued, and if the conversation can be smoothly carried out, the conversation is ignored. If the problem is caused by a dialogue model, for example, a user speech recognition error, the dialogue model may be optimized for the problem.
In one embodiment, the loan qualification auditing apparatus 400 further comprises:
and the association storage unit is used for storing the conversation record between the conversation model and the user and associating the conversation record with the parameter information table. Even if the emotion change is not detected or does not reach the change threshold, the conversation record can be still stored after the conversation is finished, and the follow-up tracing is facilitated.
In one embodiment, the loan qualification auditing apparatus 400 further comprises:
and the answer unit is used for inquiring in the database according to the extracted parameters and making an answer according to an inquiry result if the identified intention is an inquiry intention. When the identified user intent is a "query" intent, indicating that the extracted parameter is the parameter the user is querying, a corresponding answer may be made after querying in the database.
A normalizing unit 404, configured to normalize a parameter in the parameter information table;
while updating the parameter information table, the parameter values in the parameter information table may be normalized and filled in the corresponding "normalized parameter value" field. Specifically, if the parameter value of different parameters is too spoken, it is not beneficial to further analysis and evaluation, so it is necessary to standardize the "parameter value" field of the parameter information table according to the service requirement, and fill the standardized parameter into the "standardized parameter value" field. Specifically, the spoken fields, such as approximate, left-right, top-bottom, and more points in the utterance, may be modified into standardized mathematical symbols. For example, the monthly income "about ten thousand" is standardized as "10,000 ±; one year running water "hundred thousand more" was standardized as "100000 +". Of course, in the present embodiment, the parameter values in the parameter information table may be normalized after the session is completed, that is, the normalization operation may be performed in real time, or may be performed after the session is completed.
And the evaluation unit 405 is used for evaluating the loan qualification of the user according to the parameter information table.
In the evaluation unit 405, after the session process is finished, the parameter information table is saved, and then the loan qualification of the user is evaluated according to the parameter information table.
In one embodiment, as shown in fig. 6, the evaluation unit 405 includes:
a screening unit 4051 for screening users with loan intention according to the parameter information table;
a comparison unit 4052, configured to compare each parameter of the user with the loan intention with the loan requirement;
the output unit 4053 is configured to evaluate the loan qualification of the user according to the comparison result, and output a loan qualification evaluation result.
In this embodiment, users with loan intentions are screened out according to the parameter information table, and then parameter comparison is performed, so that the loan qualification of the users is evaluated. Specifically, each parameter in the parameter information table is used to confirm whether the user has loan intention, and for example, the parameter information table has a "parameter name" field including: whether the users have loan intention or not and the corresponding field of the parameter value can be yes or no, so that the parameter information table with the field of the parameter value being yes can be screened out, and the corresponding users can be screened out and are all users with loan intention. Then, the parameter information table of the user with loan intention can be compared with the loan requirements one by one, for example, the loan type which can be applied by the user is determined according to each parameter in the parameter information table, the loan requirement of the corresponding loan type is obtained, then, each parameter of the user, such as property information and working conditions, is compared with the conditions in the loan requirement, the loan qualification of the user can be evaluated according to the comparison result, for example, each condition of the user meets the condition of the loan requirement, the user can be determined to apply for loan, and a higher loan is given, or most conditions of the user meet the condition of the loan requirement, the user can be determined to apply for loan, and a medium loan limit is given; or the user can determine that the user can apply for the loan and give a lower loan amount if few conditions of the user meet the conditions of the loan requirement; or the user condition does not meet the condition of the loan requirement, the user can be determined not to apply for the loan, that is, whether the user has loan qualification or not and the user loan amount of the user under the condition that the user has the loan qualification can be determined according to the specific parameter condition.
The device carries out conversation with the user through the conversation model, reduces the labor cost, improves the auditing efficiency, automatically carries out intention identification and parameter extraction on the user utterance on the basis, automatically evaluates loan qualification according to the parameter information table of the user, avoids the problems of misreading and forgetting of the user information and the like, and improves the auditing accuracy.
It should be noted that, as will be clear to those skilled in the art, the detailed implementation process of the loan qualification auditing apparatus 400 and each unit may refer to the corresponding description in the foregoing method embodiments, and for convenience and brevity of description, no further description is provided herein.
The loan qualification auditing apparatus 400 may be implemented in the form of a computer program that is executable on a computer device such as that shown in fig. 7.
Referring to fig. 7, fig. 7 is a schematic block diagram of a computer device according to an embodiment of the present invention. The computer device 500 includes a processor 502, memory, and a network interface 505 connected by a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.
The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, causes the processor 502 to perform a loan qualification review method.
The processor 502 is used to provide computing and control capabilities that support the operation of the overall computer device 500.
The memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503, and when the computer program 5032 is executed by the processor 502, the processor 502 may be enabled to execute the loan qualification auditing method.
The network interface 505 is used for network communication, such as providing transmission of data information. Those skilled in the art will appreciate that the configuration shown in fig. 7 is a block diagram of only a portion of the configuration associated with aspects of the present invention and is not intended to limit the computing device 500 to which aspects of the present invention may be applied, and that a particular computing device 500 may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
Wherein the processor 502 is configured to run the computer program 5032 stored in the memory to implement the following functions: establishing a parameter information table; carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process; if the identified intention is the informing intention, filling the extracted parameters into the parameter information table; standardizing the parameters in the parameter information table; and evaluating the loan qualification of the user according to the parameter information table.
In one embodiment, the processor 502 performs the following operations in performing the steps of performing a dialogue with a user using a dialogue model, and performing intent recognition and parameter extraction on a user utterance using a deep learning model during the dialogue: acquiring user words, and converting the user words into text information; preprocessing the text information; and inputting the preprocessed text information into the deep learning model so as to identify intentions in the text information and extract parameters in the text information.
In one embodiment, the processor 502 performs the following operations when performing the step of evaluating the loan qualification of the user according to the parameter information table: screening out users with loan intention according to the parameter information table; comparing each parameter of the user with the loan intention with the loan requirement; and evaluating the loan qualification of the user according to the comparison result, and outputting a loan qualification evaluation result.
Those skilled in the art will appreciate that the embodiment of a computer device illustrated in fig. 7 does not constitute a limitation on the specific construction of the computer device, and that in other embodiments a computer device may include more or fewer components than those illustrated, or some components may be combined, or a different arrangement of components. For example, in some embodiments, the computer device may only include a memory and a processor, and in such embodiments, the structures and functions of the memory and the processor are consistent with those of the embodiment shown in fig. 7, and are not described herein again.
It should be understood that, in the embodiment of the present invention, the Processor 502 may be a Central Processing Unit (CPU), and the Processor 502 may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable gate arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
In another embodiment of the invention, a computer-readable storage medium is provided. The computer readable storage medium may be a non-volatile computer readable storage medium. The computer-readable storage medium stores a computer program, wherein the computer program when executed by a processor implements the steps of: establishing a parameter information table; carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process; if the identified intention is the informing intention, filling the extracted parameters into the parameter information table; standardizing the parameters in the parameter information table; and evaluating the loan qualification of the user according to the parameter information table.
In one embodiment, the performing dialog with the user by using the dialog model, and performing intention recognition and parameter extraction on the user utterance by using the deep learning model in the dialog process comprises the following steps: acquiring user words, and converting the user words into text information; preprocessing the text information; and inputting the preprocessed text information into the deep learning model so as to identify intentions in the text information and extract parameters in the text information.
In one embodiment, the evaluating the loan qualification of the user according to the parameter information table includes: screening out users with loan intention according to the parameter information table; comparing each parameter of the user with the loan intention with the loan requirement; and evaluating the loan qualification of the user according to the comparison result, and outputting a loan qualification evaluation result.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In the embodiments provided by the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only a logical division, and there may be other divisions when the actual implementation is performed, or units having the same function may be grouped into one unit, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may also be an electric, mechanical or other form of connection.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment of the present invention.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a storage medium. Based on such understanding, the technical solution of the present invention essentially or partially contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a magnetic disk, or an optical disk.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and various equivalent modifications and substitutions can be easily made by those skilled in the art within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A loan qualification auditing method is characterized by comprising the following steps:
establishing a parameter information table;
carrying out dialogue with a user by using a dialogue model, and carrying out intention recognition and parameter extraction on a user utterance by using a deep learning model in the dialogue process;
if the identified intention is the informing intention, filling the extracted parameters into the parameter information table;
standardizing the parameters in the parameter information table; and
and evaluating the loan qualification of the user according to the parameter information table.
2. The loan qualification auditing method of claim 1, wherein the dialogue with the user by using the dialogue model and the intention recognition and parameter extraction of the user utterance by using the deep learning model in the dialogue process comprises:
acquiring user words, and converting the user words into text information;
preprocessing the text information;
and inputting the preprocessed text information into the deep learning model so as to identify intentions in the text information and extract parameters in the text information.
3. The loan qualification auditing method of claim 1, wherein the parameter information table contains mandatory parameters and non-mandatory parameters, the loan qualification auditing method further comprising:
before the conversation is finished, detecting whether the necessary filling parameters in the parameter information table are filled;
if not, the question related to the filling-necessary parameter is provided for the user so as to obtain a corresponding answer and fill the answer into the parameter information table.
4. The loan qualification auditing method of claim 1, further comprising:
detecting the user utterance, and confirming whether the user has emotion change or not based on the detection result;
if yes, judging whether the emotion change reaches a change threshold value;
if yes, the manual seat is turned to continue the conversation with the user.
5. The loan qualification auditing method of claim 1, further comprising:
and storing a conversation record between the conversation model and the user, and associating the conversation record with the parameter information table.
6. The loan qualification auditing method of claim 1, further comprising:
and if the identified intention is an inquiry intention, inquiring in the database according to the extracted parameters, and making an answer according to the inquiry result.
7. The loan qualification auditing method of claim 1, wherein the evaluating the loan qualification of the user according to the parameter information table comprises:
screening out users with loan intention according to the parameter information table;
comparing each parameter of the user with the loan intention with the loan requirement;
and evaluating the loan qualification of the user according to the comparison result, and outputting a loan qualification evaluation result.
8. A loan qualification auditing device, comprising:
a parameter information table establishing unit for establishing a parameter information table;
the dialogue unit is used for carrying out dialogue with the user by utilizing the dialogue model and carrying out intention recognition and parameter extraction on the user utterance by utilizing the deep learning model in the dialogue process;
a filling unit for filling the extracted parameter into the parameter information table if the identified intention is the informing intention;
the standardization unit is used for standardizing the parameters in the parameter information table; and
and the evaluation unit is used for evaluating the loan qualification of the user according to the parameter information table.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program implements a loan qualification review method as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when executed by a processor, implements a loan qualification auditing method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010063830.4A CN111241262A (en) | 2020-01-20 | 2020-01-20 | Loan qualification auditing method based on artificial intelligence and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010063830.4A CN111241262A (en) | 2020-01-20 | 2020-01-20 | Loan qualification auditing method based on artificial intelligence and related equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111241262A true CN111241262A (en) | 2020-06-05 |
Family
ID=70878067
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010063830.4A Pending CN111241262A (en) | 2020-01-20 | 2020-01-20 | Loan qualification auditing method based on artificial intelligence and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111241262A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113239679A (en) * | 2021-04-17 | 2021-08-10 | 上海快确信息科技有限公司 | Bond position automatic detection technology based on deep learning |
| CN113362072A (en) * | 2021-06-30 | 2021-09-07 | 平安普惠企业管理有限公司 | Wind control data processing method and device, electronic equipment and storage medium |
| CN117312583A (en) * | 2023-09-28 | 2023-12-29 | 日照市住房保障管理服务中心 | House user data auxiliary management method and system |
-
2020
- 2020-01-20 CN CN202010063830.4A patent/CN111241262A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113239679A (en) * | 2021-04-17 | 2021-08-10 | 上海快确信息科技有限公司 | Bond position automatic detection technology based on deep learning |
| CN113362072A (en) * | 2021-06-30 | 2021-09-07 | 平安普惠企业管理有限公司 | Wind control data processing method and device, electronic equipment and storage medium |
| CN113362072B (en) * | 2021-06-30 | 2023-09-08 | 成都一蟹科技有限公司 | Wind control data processing method and device, electronic equipment and storage medium |
| CN117312583A (en) * | 2023-09-28 | 2023-12-29 | 日照市住房保障管理服务中心 | House user data auxiliary management method and system |
| CN117312583B (en) * | 2023-09-28 | 2024-04-26 | 日照市住房保障管理服务中心 | House user data auxiliary management method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109767787B (en) | Emotion recognition method, device and readable storage medium | |
| CN110659318B (en) | Big data-based policy pushing method, system and computer equipment | |
| CN111241262A (en) | Loan qualification auditing method based on artificial intelligence and related equipment | |
| CN110890088B (en) | Voice information feedback method and device, computer equipment and storage medium | |
| CN110781280A (en) | Knowledge graph-based voice assisting method and device | |
| CN111597818A (en) | Call quality inspection method, call quality inspection device, computer equipment and computer readable storage medium | |
| CN110399473A (en) | The method and apparatus for determining answer for customer problem | |
| CN111047092A (en) | Dispute case victory rate prediction method and device, computer equipment and storage medium | |
| WO2020042164A1 (en) | Artificial intelligence systems and methods based on hierarchical clustering | |
| CN111309882B (en) | Method and device for realizing intelligent customer service question and answer | |
| CN112288584B (en) | Insurance report processing method and device, computer readable medium and electronic equipment | |
| CN111723559B (en) | Real-time information extraction method and device | |
| CN115983285A (en) | Questionnaire auditing method, device, electronic equipment and storage medium | |
| CN111737543A (en) | Question and answer pair extraction method, device, equipment and storage medium | |
| CN114356982A (en) | Marketing compliance checking method and device, computer equipment and storage medium | |
| CN115952260A (en) | Question and answer method and device in automobile field, electronic equipment and storage medium | |
| CN113111157B (en) | Question-answer processing method, device, computer equipment and storage medium | |
| CN109787784B (en) | Group recommendation method and device, storage medium and computer equipment | |
| CN112367494A (en) | AI-based online conference communication method and device and computer equipment | |
| CN112115248A (en) | Method and system for extracting conversation strategy structure from conversation corpus | |
| CN112989021A (en) | Method, device and equipment for judging violation of counselor | |
| CN111309990A (en) | Statement response method and device | |
| CN113420125B (en) | Question-answer pair determining method, system, storage medium and equipment based on industry type | |
| CN117591669A (en) | Customer service data mining method, customer service data mining device, customer service data mining equipment and customer service data mining medium | |
| CN109960725B (en) | Text classification processing method and device based on emotion and computer equipment |
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 |