CN116401355B - Consultation business decision management method and system based on digital intelligence interaction - Google Patents
Consultation business decision management method and system based on digital intelligence interaction Download PDFInfo
- Publication number
- CN116401355B CN116401355B CN202310672533.3A CN202310672533A CN116401355B CN 116401355 B CN116401355 B CN 116401355B CN 202310672533 A CN202310672533 A CN 202310672533A CN 116401355 B CN116401355 B CN 116401355B
- Authority
- CN
- China
- Prior art keywords
- data
- consultation
- service
- features
- business
- 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.)
- Active
Links
- 238000007726 management method Methods 0.000 title claims abstract description 61
- 230000003993 interaction Effects 0.000 title claims abstract description 31
- 238000004458 analytical method Methods 0.000 claims abstract description 97
- 238000000034 method Methods 0.000 claims abstract description 40
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000012549 training Methods 0.000 claims description 17
- 238000003066 decision tree Methods 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 14
- 230000002085 persistent effect Effects 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 13
- 238000010276 construction Methods 0.000 claims description 11
- 238000007621 cluster analysis Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 230000002688 persistence Effects 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 5
- 230000004044 response Effects 0.000 abstract description 3
- 238000009223 counseling Methods 0.000 description 7
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 230000004075 alteration Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
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
- G06F16/3344—Query execution using natural language analysis
-
- 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
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- 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
- G06F16/35—Clustering; Classification
- G06F16/353—Clustering; Classification into predefined classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
- G06F40/35—Discourse or dialogue representation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a consultation business decision management method and a consultation business decision management system based on digital intelligence interaction, which relate to the technical field of data processing, and comprise the following steps: p consultation window data of P target users in a preset time window are obtained; the method comprises the steps of calling P historical business consultation data based on a data interaction device, inputting the data into a semantic understanding analysis model, and obtaining P semantic understanding categories; generating Q interesting service feature sets and M service depth features; inputting the Q interesting service feature sets, the M service depth features, the P target users and the P semantic understanding categories into a service decision management model, and outputting P service decision schemes; and carrying out consultation business decision management on the P target users. The application solves the technical problems of low processing efficiency and poor response quality of the user consultation service in the prior art, and achieves the technical effects of intelligently making the consultation service decision and improving the management quality.
Description
Technical Field
The application relates to the technical field of data processing, in particular to a consultation service decision management method and system based on digital intelligence interaction.
Background
With the increasing update of network technology, the management mode of the consultation service has also changed. The traditional manual answer mode for the consultation questions has long answer time, and can not finish higher-quality answer due to uneven personnel capacity level. In the prior art, the user consultation service has low processing efficiency and answers the technical problem of poor quality.
Disclosure of Invention
The application provides a consultation service decision management method and system based on digital intelligence interaction, which are used for solving the technical problems of low processing efficiency and poor response quality of user consultation service in the prior art.
In view of the above problems, the present application provides a method and a system for decision management of consultation services based on digital intelligence interaction.
The application provides a consultation service decision management method based on intelligent interaction, wherein the service decision management platform is in communication connection with a data interaction device, and the method comprises the following steps:
p pieces of consultation window data of P target users in a preset time window are obtained, wherein the consultation window data comprise business application categories and consultation content data, and P is a positive integer greater than 1;
the method comprises the steps of calling P historical service consultation data of P target users based on the data interaction device, inputting the P historical service consultation data into a semantic understanding analysis model, and obtaining P semantic understanding categories, wherein the P semantic understanding categories are in one-to-one correspondence with the P target users;
generating Q interested service feature sets according to P service application categories and P historical service consultation data in the P consultation window data;
generating M service depth characteristics according to P consultation content data and P historical service consultation data in the P consultation window data;
inputting the Q interesting service feature sets, M service depth features, P target users and P semantic understanding categories into a service decision management model, and outputting P service decision schemes;
and carrying out consultation business decision management on the P target users according to the P business decision schemes.
In a second aspect of the present application, there is provided a consultation service decision management system based on digital intelligence interaction, the system comprising:
the system comprises a window data acquisition module, a window data processing module and a data processing module, wherein the window data acquisition module is used for acquiring P pieces of consultation window data of P target users in a preset time window, the consultation window data comprise business application categories and consultation content data, and P is a positive integer greater than 1;
the system comprises an understanding category obtaining module, a semantic understanding module and a semantic analysis module, wherein the understanding category obtaining module is used for calling P historical business consultation data of P target users based on a data interaction device, inputting the P historical business consultation data into the semantic analysis model and obtaining P semantic understanding categories, and the P semantic understanding categories are in one-to-one correspondence with the P target users;
the business feature generation module is used for generating Q interested business feature sets according to P business application categories and P historical business consultation data in the P consultation window data;
the depth feature generation module is used for generating M service depth features according to P consultation content data and P historical service consultation data in the P consultation window data;
the business decision scheme output module is used for inputting the Q interesting business feature sets, M business depth features, P target users and P semantic understanding categories into a business decision management model to output P business decision schemes;
and the decision management module is used for carrying out consultation business decision management on the P target users according to the P business decision schemes.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
the application obtains P consulting window data of P target users in a preset time window, wherein the consulting window data comprises business application categories and consulting content data, P is a positive integer greater than 1, then P historical business consulting data of the P target users are called based on a data interaction device, the P historical business consulting data are input into a semantic understanding analysis model to obtain P semantic understanding categories, the P semantic understanding categories are in one-to-one correspondence with the P target users, Q interested business feature sets are generated according to the P business application categories and the P historical business consulting data in the P consulting window data, M business depth features are generated according to the P consulting content data and the P historical business consulting data in the P consulting window data, and P business decision schemes are output by inputting the Q interested business feature sets, the M business depth features, the P target users and the P semantic understanding categories into a business decision management model, and the P target users are subjected to consulting business decision management according to the P business decision schemes. The technical effect of improving the decision management efficiency of the consultation service is achieved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for decision management of a consultation service based on digital intelligence interaction according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of outputting P semantic understanding categories in a consultation service decision management method based on digital intelligence interaction according to the embodiment of the present application;
FIG. 3 is a schematic flow chart of generating Q interesting service feature sets in a consultation service decision management method based on intelligent interaction according to the embodiment of the present application;
fig. 4 is a schematic structural diagram of a consultation service decision management system based on digital intelligence interaction according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a window data acquisition module 11, an understanding category acquisition module 12, a service characteristic generation module 13, a depth characteristic generation module 14, a service decision scheme output module 15 and a decision management module 16.
Detailed Description
The application provides a consultation service decision management method and system based on digital intelligence interaction, which are used for solving the technical problems of low processing efficiency and poor response quality of user consultation service in the prior art.
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
It should be noted that the terms "comprises" and "comprising," along with any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus, but may include other steps or modules not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
As shown in fig. 1, the present application provides a consultation service decision management method based on intelligent interaction, wherein the method is applied to a service decision management platform, and the service decision management platform is in communication connection with a data interaction device, and the method comprises:
step S100: p pieces of consultation window data of P target users in a preset time window are obtained, wherein the consultation window data comprise business application categories and consultation content data, and P is a positive integer greater than 1;
in a possible embodiment, the data interaction device is a device for extracting relevant data of the target user, including an application port and the like. The preset time window is a preset time period for making a consultation service processing decision, and is set by the staff at will, without limitation. And the P target users are users for carrying out business consultation in a preset time period. The P consultation window data are all data generated in the period of time when P target users conduct consultation. Wherein the consultation window data includes business application category and consultation content data. The service application category is data for describing the service types required to be consulted by the target user, and comprises the service types such as identity registration, social security subsidy, certificate inquiry and the like. The counseling content data describes the concrete content of the counseling of the target user. Illustratively, the user consulting content is "whether the undergraduate has been the candidate for one year just after graduation, how to reserve the candidate identity", and the like. Basic analysis data is provided for subsequent business decision management, namely corresponding business consultation reply, by acquiring consultation window data of the target user.
Step S200: the method comprises the steps of calling P historical service consultation data of P target users based on the data interaction device, inputting the P historical service consultation data into a semantic understanding analysis model, and obtaining P semantic understanding categories, wherein the P semantic understanding categories are in one-to-one correspondence with the P target users;
further, as shown in fig. 2, the step S200 of the embodiment of the present application further includes:
step S210: extracting P consultation frequencies of P target users based on the P historical service consultation data, and generating P first semantic analysis items;
step S220: taking K business application categories as clustering targets, carrying out cluster analysis on the P consultation frequencies to obtain K consultation frequencies of the K business application categories of the P target users, and generating P second semantic analysis items;
step S230: based on the P second semantic analysis items, traversing the P historical business consultation data to calculate K business application category overlapping degrees of P target users, and generating P third semantic analysis items;
step S240: inputting the P first semantic analysis items, the P second semantic analysis items and the P third semantic analysis items into the semantic understanding analysis model, and outputting P semantic understanding categories.
Further, step S200 of the embodiment of the present application further includes:
step S250: acquiring a plurality of sample first semantic analysis items, a plurality of sample second semantic analysis items, a plurality of sample third semantic analysis items and a plurality of semantic understanding domains as construction data, and marking the plurality of semantic understanding domains;
step S260: training a framework constructed by a feedforward neural network foundation by utilizing the construction data, and supervising by using a plurality of semantic understanding categories of the marks in the training process;
step S270: and until the training reaches convergence, obtaining the semantic understanding analysis model after the training is completed.
In one embodiment, the data interaction device is used to acquire P historical business consultation data of P target users, that is, data generated by consulting on the consultation platform by the P target users in a historical time period is acquired, and the interpretation degree, that is, the semantic understanding category, which can be understood by the target users when the business consultation reply is carried out on the target users according to analysis. Illustratively, when a user repeatedly asks a business question 20 times, indicating that the user has poor acceptance of the information, the user needs to explain by white speech and describe by way of example only, so that the user can understand the business content of the consultation. The semantic understanding analysis model is a functional model for intelligently analyzing semantic understanding capacity of a user based on historical business consultation data, wherein the input data are P pieces of historical business consultation data, and the output data are P pieces of semantic understanding categories.
In one embodiment, P consultation frequencies of P target users are extracted based on the P historical service consultation data, so that overall consultation frequencies of the target users are obtained, and the average consultation frequency spent by the users in solving the problem can be judged and used as P first semantic analysis items. And taking K business application categories as clustering targets, carrying out clustering analysis on P consultation frequencies, namely analyzing the consultation times corresponding to each business application category in the K business application categories, and reflecting the degree that a user does not understand the business content in one business application category as P second semantic analysis items. Furthermore, based on the P second semantic analysis items, dividing the consultation frequency corresponding to each business application category of each target user by the total consultation frequency of each target user in the history time, carrying out mean processing on the consultation frequency corresponding to the K business application categories, taking the result of the mean processing as the overlapping degree of the K business application categories, namely quantifying the repetition degree required by understanding the business content when carrying out business consultation on the target user, and taking the result as a third semantic analysis item. And then, inputting the P first semantic analysis items, the P second semantic analysis items and the P third semantic analysis items into the semantic understanding analysis model, and obtaining P semantic understanding categories through intelligent operation of the model.
Specifically, a plurality of sample first semantic analysis items, a plurality of sample second semantic analysis items, a plurality of sample third semantic analysis items and a plurality of semantic understanding categories are obtained from big data by taking semantic understanding as indexes and used as construction data, the plurality of semantic understanding categories are marked, and in the process of training a framework constructed on the basis of a feedforward neural network by using the construction data, the marked plurality of semantic understanding categories are used for supervision until an output result is converged, and the semantic understanding analysis model is obtained.
Step S300: generating Q interested service feature sets according to P service application categories and P historical service consultation data in the P consultation window data;
further, as shown in fig. 3, step S300 of the embodiment of the present application further includes:
step S310: extracting P historical business application category sets from the P historical business consultation data;
step S320: respectively judging whether the P business application categories are successfully matched with the P historical business application category sets, if so, comparing the P successful matching numbers with the total number of the categories in the P historical business application category sets, and taking the P calculation results as P first attention features;
step S330: generating P second attention features according to P interval time of successful matching;
step S340: and generating Q interested service feature sets according to the P first attention features and the P second attention features.
Further, step S340 of the embodiment of the present application further includes:
step S341: randomly not replacing the first attention degree features of a sample from the first attention degree features of a plurality of samples to select the first attention degree features of the sample as first internal nodes, and performing two-classification on the first attention degree features of the plurality of samples by the first attention degree features of the sample to obtain a first division result;
step S342: randomly not replacing the first attention degree features of one sample from the first attention degree features of the plurality of samples to select the first attention degree features of the one sample as second internal nodes, and performing two classification on the first division result by the first attention degree features of the sample to obtain a second division result;
step S343: randomly not replacing the first attention degree characteristic of one sample from the first attention degree characteristics of the plurality of samples to select the first attention degree characteristic of the sample as an N internal node, and carrying out second classification on the N-1 division result by the first attention degree characteristic of the sample to obtain an N division result;
step S344: marking the first division result, the second division result and the N division result by using the corresponding service feature mean value of interest in each division result, and generating according to the first internal node, the second internal node, the N internal node, the marked first division result, the marked second division result and the marked N division result;
step S345: inputting the P first attention features into the directional branches of the interested service feature analysis decision tree, outputting Q1 directional service feature sets, inputting the P second attention features into the persistent branches of the interested service feature analysis decision tree, and outputting Q 2 A set of persistent business features;
step S346: according to the Q1 directional service characteristic sets and Q 2 The persistent service feature sets generate Q service feature sets of interest.
In one possible embodiment, the service application category is taken as an index, P historical service application category sets are extracted from the P historical service consultation data, further, P service application categories corresponding to P target users are respectively matched with the P historical service application category sets, when the matching is successful, the matching success number corresponding to the P matching success data is compared with the total number of categories in the P historical service application category sets, and P calculation results are obtained, wherein the P calculation results respectively reflect the proportion of the service consulted by the P target users in the historical time period by the service consulted by the P target users in the preset time window, and therefore the P calculation results are used as P first attention features. And carrying out average processing on the P successfully matched data according to a plurality of historical interval times in the historical time period to obtain P successfully matched interval times, and taking the P successfully matched interval times as P second attention features. Q sets of business features of interest are then generated based on the P first features of interest and the P second features of interest.
Specifically, a first sample attention feature is randomly not replaced from a plurality of first sample attention features to be selected as a first internal node, the first sample attention features are subjected to secondary classification to obtain a first classification result, a first sample attention feature is randomly not replaced from the plurality of first sample attention features to be selected as a second internal node, the first classification result is subjected to secondary classification to obtain a second classification result, a first sample attention feature is randomly not replaced from the plurality of first sample attention features to be selected as an N internal node, the first sample attention feature is subjected to secondary classification to the N-1 classification result, and the N-1 classification result is subjected to secondary classification to obtain a N-th classification result. And marking each division result by utilizing the corresponding service feature mean value of interest in each division result, namely the directivity degree in the service feature of interest, and obtaining a marked first division result, second division result and Nth division result. Generating directional branches of the service feature analysis decision tree of interest according to the first internal node, the second internal node, the Nth internal node, the marked first division result, the marked second division result and the marked Nth division result, inputting P first attention features into the directional branches of the service feature analysis decision tree of interest, and outputting Q 1 A set of directional traffic characteristics. Wherein the Q is 1 The individual directional service feature sets reflect the specific attention degree of the target user to the service in the preset time window.
Specifically, a plurality of sample second attention features are acquired, and a persistent branch of the service feature analysis decision tree of interest is constructed based on the same construction method as the directional branch of the service feature analysis decision tree of interest. Output Q by inputting P second-degree-of-interest features into persistent branches of the business feature analysis decision tree of interest 2 A set of persistent business features. Wherein the Q is 2 The persistent business feature set reflects the consulting persistence degree of the target user on the business. The Q is set to 1 Individual directional traffic feature set and Q 2 After the persistent service feature sets are summarized, Q interesting service feature sets are generated.
Step S400: generating M service depth characteristics according to P consultation content data and P historical service consultation data in the P consultation window data;
in one possible embodiment, the data byte amount is compared by comparing the information amount of P counsel content data and P historical business counsel data in the P counsel window data. Preferably, the data byte amounts of the P pieces of advisory content data and the data byte amounts of the P pieces of history service advisory data are compared, respectively, and the obtained result is characterized as the P data byte amounts. And carrying out cluster analysis on the P data byte quantity features by taking the size of the data byte quantity features as a clustering target to obtain M data byte quantity features, and carrying out service depth assignment on the M data byte quantity features according to a preset feature depth standard table to obtain M service depth features.
Step S500: inputting the Q interesting service feature sets, M service depth features, P target users and P semantic understanding categories into a service decision management model, and outputting P service decision schemes;
further, step S500 of the embodiment of the present application further includes:
step S510: taking the P semantic understanding categories as a first constraint condition;
step S520: and analyzing the Q interesting service feature sets, the M service depth features and the P target user input service decision management models, and restricting the analysis process by using a first restriction condition to obtain P service decision schemes.
In one possible embodiment, the P semantic understanding categories are used as a first constraint condition set by a service decision scheme, Q interested service feature sets, M service depth features and P target users are then input into a service decision management model, preferably, one target user is randomly selected from the P target users, the target user is used as an index, the Q interested service feature sets and the M service depth features are matched to obtain corresponding directional service features and directional service features, the corresponding directional service features and the directional service features are input into the service decision management model, and the decision process is constrained by using the first constraint condition corresponding to the target user to obtain a corresponding service decision scheme. Furthermore, based on the same method, P business decision schemes corresponding to P target users are obtained.
Step S600: and carrying out consultation business decision management on the P target users according to the P business decision schemes.
Further, after the consulting service decision management is performed on the P target users according to the P service decision schemes, step S600 in the embodiment of the present application further includes:
step S610: p pieces of feedback evaluation information of P target users in a preset feedback time period are obtained;
step S620: information entropy calculation is carried out based on the P pieces of feedback evaluation information, and the P pieces of feedback evaluation information are screened according to information entropy calculation results, so that an effective information set is obtained;
step S630: and updating and correcting the network parameters of the business decision management model according to the effective information set.
In one possible embodiment, the service consultation process of the P target users is managed according to the P service decision schemes, and feedback evaluation collection is performed on the P target users in a preset feedback time period, so as to obtain P feedback evaluation information. And respectively calculating the data information quantity in the P pieces of feedback evaluation information to obtain an information entropy calculation result, and reserving the feedback evaluation information meeting the preset information entropy to obtain an effective information set. Therefore, the data volume of analysis is reduced, and the analysis efficiency is improved. And correcting network parameters in the service decision management model according to the information reflected in the effective information set, and if the information in the service decision scheme reflected by the user cannot meet the requirement, adjusting the first constraint condition so that the first constraint condition can meet the requirement of the user.
In summary, the embodiment of the application has at least the following technical effects:
according to the application, the target user consultation data in the preset time window is analyzed, the characteristics of the user consultation service are analyzed by utilizing the historical consultation data of the user, and the basis is provided for service decision from three dimensions of semantic understanding, interesting service and service depth, so that the technical effects of improving the service decision management quality and intelligently improving the analysis efficiency are achieved.
Example two
Based on the same inventive concept as the counseling service decision management method based on intelligent interaction in the foregoing embodiments, as shown in fig. 4, the present application provides a counseling service decision management system based on intelligent interaction, and the system and method embodiments in the embodiments of the present application are based on the same inventive concept. Wherein the system comprises:
the window data acquisition module 11 is configured to acquire P pieces of counseling window data of P target users in a preset time window, where the counseling window data includes a business application category and counseling content data, and P is a positive integer greater than 1;
the understanding category obtaining module 12 is configured to invoke P historical service consultation data of P target users based on the data interaction device, and input the P historical service consultation data into the semantic understanding analysis model to obtain P semantic understanding categories, where the P semantic understanding categories are in one-to-one correspondence with the P target users;
the business feature generation module 13 is used for generating Q interested business feature sets according to P business application categories and P historical business consultation data in the P consultation window data;
the depth feature generating module 14 is configured to generate M service depth features according to P pieces of advisory content data and P pieces of historical service advisory data in the P pieces of advisory window data;
the service decision scheme output module 15, where the service decision scheme output module 15 is configured to input the Q service feature sets of interest, M service depth features, P target users, and P semantic understanding categories into a service decision management model, and output P service decision schemes;
and the decision management module 16 is used for carrying out consultation business decision management on the P target users according to the P business decision schemes.
Further, the understanding category obtaining module 12 is configured to perform the following method:
extracting P consultation frequencies of P target users based on the P historical service consultation data, and generating P first semantic analysis items;
taking K business application categories as clustering targets, carrying out cluster analysis on the P consultation frequencies to obtain K consultation frequencies of the K business application categories of the P target users, and generating P second semantic analysis items;
based on the P second semantic analysis items, traversing the P historical business consultation data to calculate K business application category overlapping degrees of P target users, and generating P third semantic analysis items;
inputting the P first semantic analysis items, the P second semantic analysis items and the P third semantic analysis items into the semantic understanding analysis model, and outputting P semantic understanding categories.
Further, the understanding category obtaining module 12 is configured to perform the following method:
acquiring a plurality of sample first semantic analysis items, a plurality of sample second semantic analysis items, a plurality of sample third semantic analysis items and a plurality of semantic understanding domains as construction data, and marking the plurality of semantic understanding domains;
training a framework constructed by a feedforward neural network foundation by utilizing the construction data, and supervising by using a plurality of semantic understanding categories of the marks in the training process;
and until the training reaches convergence, obtaining the semantic understanding analysis model after the training is completed.
Further, the service feature generating module 13 is configured to perform the following method:
extracting P historical business application category sets from the P historical business consultation data;
respectively judging whether the P business application categories are successfully matched with the P historical business application category sets, if so, comparing the P successful matching numbers with the total number of the categories in the P historical business application category sets, and taking the P calculation results as P first attention features;
generating P second attention features according to P interval time of successful matching;
and generating Q interested service feature sets according to the P first attention features and the P second attention features.
Further, the service feature generating module 13 is configured to perform the following method:
randomly not replacing the first attention degree features of a sample from the first attention degree features of a plurality of samples to select the first attention degree features of the sample as first internal nodes, and performing two-classification on the first attention degree features of the plurality of samples by the first attention degree features of the sample to obtain a first division result;
randomly not replacing the first attention degree features of one sample from the first attention degree features of the plurality of samples to select the first attention degree features of the one sample as second internal nodes, and performing two classification on the first division result by the first attention degree features of the sample to obtain a second division result;
randomly not replacing the first attention degree characteristic of one sample from the first attention degree characteristics of the plurality of samples to select the first attention degree characteristic of the sample as an N internal node, and carrying out second classification on the N-1 division result by the first attention degree characteristic of the sample to obtain an N division result;
marking the first division result, the second division result and the N division result by using the corresponding service feature mean value of interest in each division result, and generating directional branches of the service feature analysis decision tree of interest according to the first internal node, the second internal node, the N internal node, the marked first division result, the second division result and the N division result;
inputting the P first attention features into directional branches of the interesting business feature analysis decision tree, and outputting Q 1 A directional service feature set, inputting the P second attention features into the persistence branch of the service feature analysis decision tree of interest, and outputting Q 2 A set of persistent business features;
according to said Q 1 Individual directional traffic feature set and Q 2 The persistent service feature sets generate Q service feature sets of interest.
Further, the service decision scheme output module 15 is configured to execute the following method:
taking the P semantic understanding categories as a first constraint condition;
and analyzing the Q interesting service feature sets, the M service depth features and the P target user input service decision management models, and restricting the analysis process by using a first restriction condition to obtain P service decision schemes.
Further, the decision management module 16 is configured to perform the following method:
p pieces of feedback evaluation information of P target users in a preset feedback time period are obtained;
information entropy calculation is carried out based on the P pieces of feedback evaluation information, and the P pieces of feedback evaluation information are screened according to information entropy calculation results, so that an effective information set is obtained;
and updating and correcting the network parameters of the business decision management model according to the effective information set.
It should be noted that the sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The foregoing description of the preferred embodiments of the application is not intended to limit the application to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the application are intended to be included within the scope of the application.
The specification and figures are merely exemplary illustrations of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the scope of the application. Thus, the present application is intended to include such modifications and alterations insofar as they come within the scope of the application or the equivalents thereof.
Claims (4)
1. The consultation business decision management method based on digital intelligence interaction is characterized by being applied to a business decision management platform, wherein the business decision management platform is in communication connection with a data interaction device, and the method comprises the following steps:
p pieces of consultation window data of P target users in a preset time window are obtained, wherein the consultation window data comprise business application categories and consultation content data, and P is a positive integer greater than 1;
the method comprises the steps of calling P historical service consultation data of P target users based on the data interaction device, inputting the P historical service consultation data into a semantic understanding analysis model, and obtaining P semantic understanding categories, wherein the P semantic understanding categories are in one-to-one correspondence with the P target users;
generating Q interested service feature sets according to P service application categories and P historical service consultation data in the P consultation window data;
generating M service depth characteristics according to P consultation content data and P historical service consultation data in the P consultation window data;
inputting the Q interesting service feature sets, M service depth features, P target users and P semantic understanding categories into a service decision management model, and outputting P service decision schemes;
carrying out consultation business decision management on P target users according to the P business decision schemes;
the P historical business consultation data are input into a semantic understanding analysis model, and the method comprises the following steps:
extracting P consultation frequencies of P target users based on the P historical service consultation data, and generating P first semantic analysis items;
taking K business application categories as clustering targets, carrying out cluster analysis on the P consultation frequencies to obtain K consultation frequencies of the K business application categories of the P target users, and generating P second semantic analysis items;
based on the P second semantic analysis items, traversing the P historical business consultation data to calculate K business application category overlapping degrees of P target users, and generating P third semantic analysis items;
inputting the P first semantic analysis items, the P second semantic analysis items and the P third semantic analysis items into the semantic understanding analysis model, and outputting P semantic understanding categories;
acquiring a plurality of sample first semantic analysis items, a plurality of sample second semantic analysis items, a plurality of sample third semantic analysis items and a plurality of semantic understanding domains as construction data, and marking the plurality of semantic understanding domains;
training a framework constructed by a feedforward neural network foundation by utilizing the construction data, and supervising by using a plurality of semantic understanding categories of the marks in the training process;
until the training reaches convergence, obtaining the semantic understanding analysis model after the training is completed;
the generating Q service feature sets of interest according to the P service application categories and the P historical service advisory data in the P advisory window data includes:
extracting P historical business application category sets from the P historical business consultation data;
respectively judging whether the P business application categories are successfully matched with the P historical business application category sets, if so, comparing the P successful matching numbers with the total number of the categories in the P historical business application category sets, and taking the P calculation results as P first attention features;
generating P second attention features according to P interval time of successful matching;
generating Q interesting service feature sets according to the P first attention features and the P second attention features;
randomly not replacing the first attention degree features of a sample from the first attention degree features of a plurality of samples to select the first attention degree features of the sample as first internal nodes, and performing two-classification on the first attention degree features of the plurality of samples by the first attention degree features of the sample to obtain a first division result;
randomly not replacing the first attention degree features of one sample from the first attention degree features of the plurality of samples to select the first attention degree features of the one sample as second internal nodes, and performing two classification on the first division result by the first attention degree features of the sample to obtain a second division result;
randomly not replacing the first attention degree characteristic of one sample from the first attention degree characteristics of the plurality of samples to select the first attention degree characteristic of the sample as an N internal node, and carrying out second classification on the N-1 division result by the first attention degree characteristic of the sample to obtain an N division result;
marking the first division result, the second division result and the N division result by using the corresponding service feature mean value of interest in each division result, and generating directional branches of the service feature analysis decision tree of interest according to the first internal node, the second internal node, the N internal node, the marked first division result, the second division result and the N division result;
inputting the P first attention features into directional branches of the interesting business feature analysis decision tree, and outputting Q 1 Personal directional traffic characteristicsThe P second attention features are input into the persistence branches of the interested business feature analysis decision tree to output Q 2 A set of persistent business features;
according to said Q 1 Individual directional traffic feature set and Q 2 Generating Q interested service feature sets by the persistent service feature sets;
the generating M service depth features according to the P pieces of advisory content data and the P pieces of historical service advisory data in the P pieces of advisory window data includes:
respectively obtaining data byte amounts of P consultation content data and P historical service consultation data in the P consultation window data; comparing the data byte quantity of the P consulting content data with the data byte quantity of the P historical service consulting data to obtain P data byte quantity characteristics; and carrying out cluster analysis on the P data byte quantity features by taking the size of the data byte quantity features as a clustering target to obtain M data byte quantity features, and carrying out service depth assignment on the M data byte quantity features according to a preset feature depth standard table to obtain the M service depth features.
2. The method of claim 1, wherein the method comprises:
taking the P semantic understanding categories as a first constraint condition;
and analyzing the Q interesting service feature sets, the M service depth features and the P target user input service decision management models, and restricting the analysis process by using a first restriction condition to obtain P service decision schemes.
3. The method of claim 1, wherein after consulting business decision management for P target users according to the P business decision schemes, the method comprises:
p pieces of feedback evaluation information of P target users in a preset feedback time period are obtained;
information entropy calculation is carried out based on the P pieces of feedback evaluation information, and the P pieces of feedback evaluation information are screened according to information entropy calculation results, so that an effective information set is obtained;
and updating and correcting the network parameters of the business decision management model according to the effective information set.
4. A consultation service decision management system based on digital intelligence interaction, the system comprising:
the system comprises a window data acquisition module, a window data processing module and a data processing module, wherein the window data acquisition module is used for acquiring P pieces of consultation window data of P target users in a preset time window, the consultation window data comprise business application categories and consultation content data, and P is a positive integer greater than 1;
the system comprises an understanding category obtaining module, a semantic understanding module and a semantic analysis module, wherein the understanding category obtaining module is used for calling P historical business consultation data of P target users based on a data interaction device, inputting the P historical business consultation data into the semantic analysis model and obtaining P semantic understanding categories, and the P semantic understanding categories are in one-to-one correspondence with the P target users;
the business feature generation module is used for generating Q interested business feature sets according to P business application categories and P historical business consultation data in the P consultation window data;
the depth feature generation module is used for generating M service depth features according to P consultation content data and P historical service consultation data in the P consultation window data;
the business decision scheme output module is used for inputting the Q interesting business feature sets, M business depth features, P target users and P semantic understanding categories into a business decision management model to output P business decision schemes;
the decision management module is used for carrying out consultation business decision management on P target users according to the P business decision schemes;
the understanding category obtaining module includes:
extracting P consultation frequencies of P target users based on the P historical service consultation data, and generating P first semantic analysis items;
taking K business application categories as clustering targets, carrying out cluster analysis on the P consultation frequencies to obtain K consultation frequencies of the K business application categories of the P target users, and generating P second semantic analysis items;
based on the P second semantic analysis items, traversing the P historical business consultation data to calculate K business application category overlapping degrees of P target users, and generating P third semantic analysis items;
inputting the P first semantic analysis items, the P second semantic analysis items and the P third semantic analysis items into the semantic understanding analysis model, and outputting P semantic understanding categories;
acquiring a plurality of sample first semantic analysis items, a plurality of sample second semantic analysis items, a plurality of sample third semantic analysis items and a plurality of semantic understanding domains as construction data, and marking the plurality of semantic understanding domains;
training a framework constructed by a feedforward neural network foundation by utilizing the construction data, and supervising by using a plurality of semantic understanding categories of the marks in the training process;
until the training reaches convergence, obtaining the semantic understanding analysis model after the training is completed;
the service characteristic generating module comprises:
extracting P historical business application category sets from the P historical business consultation data;
respectively judging whether the P business application categories are successfully matched with the P historical business application category sets, if so, comparing the P successful matching numbers with the total number of the categories in the P historical business application category sets, and taking the P calculation results as P first attention features;
generating P second attention features according to P interval time of successful matching;
generating Q interesting service feature sets according to the P first attention features and the P second attention features;
randomly not replacing the first attention degree features of a sample from the first attention degree features of a plurality of samples to select the first attention degree features of the sample as first internal nodes, and performing two-classification on the first attention degree features of the plurality of samples by the first attention degree features of the sample to obtain a first division result;
randomly not replacing the first attention degree features of one sample from the first attention degree features of the plurality of samples to select the first attention degree features of the one sample as second internal nodes, and performing two classification on the first division result by the first attention degree features of the sample to obtain a second division result;
randomly not replacing the first attention degree characteristic of one sample from the first attention degree characteristics of the plurality of samples to select the first attention degree characteristic of the sample as an N internal node, and carrying out second classification on the N-1 division result by the first attention degree characteristic of the sample to obtain an N division result;
marking the first division result, the second division result and the N division result by using the corresponding service feature mean value of interest in each division result, and generating directional branches of the service feature analysis decision tree of interest according to the first internal node, the second internal node, the N internal node, the marked first division result, the second division result and the N division result;
inputting the P first attention features into directional branches of the interesting business feature analysis decision tree, and outputting Q 1 A directional service feature set, inputting the P second attention features into the persistence branch of the service feature analysis decision tree of interest, and outputting Q 2 A set of persistent business features;
according to said Q 1 Individual directional traffic feature set and Q 2 Generating Q interested service feature sets by the persistent service feature sets;
the depth feature generation module comprises:
respectively obtaining data byte amounts of P consultation content data and P historical service consultation data in the P consultation window data; comparing the data byte quantity of the P consulting content data with the data byte quantity of the P historical service consulting data to obtain P data byte quantity characteristics; and carrying out cluster analysis on the P data byte quantity features by taking the size of the data byte quantity features as a clustering target to obtain M data byte quantity features, and carrying out service depth assignment on the M data byte quantity features according to a preset feature depth standard table to obtain the M service depth features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310672533.3A CN116401355B (en) | 2023-06-08 | 2023-06-08 | Consultation business decision management method and system based on digital intelligence interaction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310672533.3A CN116401355B (en) | 2023-06-08 | 2023-06-08 | Consultation business decision management method and system based on digital intelligence interaction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116401355A CN116401355A (en) | 2023-07-07 |
| CN116401355B true CN116401355B (en) | 2023-08-15 |
Family
ID=87012717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310672533.3A Active CN116401355B (en) | 2023-06-08 | 2023-06-08 | Consultation business decision management method and system based on digital intelligence interaction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116401355B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117893216B (en) * | 2024-03-14 | 2024-05-14 | 江苏鑫埭信息科技有限公司 | User behavior analysis method and system based on big data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109885695A (en) * | 2019-01-28 | 2019-06-14 | 平安科技（深圳）有限公司 | Assets suggest generation method, device, computer equipment and storage medium |
| CN110688454A (en) * | 2019-09-09 | 2020-01-14 | 深圳壹账通智能科技有限公司 | Method, device, equipment and storage medium for processing consultation conversation |
| CN112597390A (en) * | 2020-12-25 | 2021-04-02 | 李华兵 | Block chain big data processing method based on digital finance and big data server |
| CN113438515A (en) * | 2021-06-17 | 2021-09-24 | 重庆大牛认知科技有限公司 | IPTV terminal government affair consultation method and system based on intelligent interaction |
| KR20210118634A (en) * | 2020-03-23 | 2021-10-01 | 주식회사 베스트브라더스 | Call center service efficiency improvement system based on data and method thereof |
| WO2022095378A1 (en) * | 2020-11-03 | 2022-05-12 | 平安科技（深圳）有限公司 | Artificial-intelligence-based training method and apparatus, and computer device and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230140828A1 (en) * | 2021-10-28 | 2023-05-04 | Mckinsey & Company, Inc. | Machine Learning Methods And Systems For Cataloging And Making Recommendations Based On Domain-Specific Knowledge |
-
2023
- 2023-06-08 CN CN202310672533.3A patent/CN116401355B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109885695A (en) * | 2019-01-28 | 2019-06-14 | 平安科技（深圳）有限公司 | Assets suggest generation method, device, computer equipment and storage medium |
| CN110688454A (en) * | 2019-09-09 | 2020-01-14 | 深圳壹账通智能科技有限公司 | Method, device, equipment and storage medium for processing consultation conversation |
| KR20210118634A (en) * | 2020-03-23 | 2021-10-01 | 주식회사 베스트브라더스 | Call center service efficiency improvement system based on data and method thereof |
| WO2022095378A1 (en) * | 2020-11-03 | 2022-05-12 | 平安科技（深圳）有限公司 | Artificial-intelligence-based training method and apparatus, and computer device and storage medium |
| CN112597390A (en) * | 2020-12-25 | 2021-04-02 | 李华兵 | Block chain big data processing method based on digital finance and big data server |
| CN113438515A (en) * | 2021-06-17 | 2021-09-24 | 重庆大牛认知科技有限公司 | IPTV terminal government affair consultation method and system based on intelligent interaction |
Non-Patent Citations (1)
| Title |
|---|
| 基于文本与Web语义分析的智能咨询服务模式及体系架构研究;唐晓波;魏巍;;情报科学(第11期);113-116 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116401355A (en) | 2023-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108540320B (en) | Evaluation method for mining user satisfaction based on signaling | |
| CN116401355B (en) | Consultation business decision management method and system based on digital intelligence interaction | |
| CN108874959A (en) | A kind of user's dynamic interest model method for building up based on big data technology | |
| Pan et al. | A systematic review of information security risk assessment | |
| CN115906954A (en) | Multivariate time sequence prediction method and device based on graph neural network | |
| CN112054943A (en) | Traffic prediction method for mobile network base station | |
| CN112200375B (en) | Prediction model generation method, prediction model generation device, and computer-readable medium | |
| CN111695084A (en) | Model generation method, credit score generation method, device, equipment and storage medium | |
| CN116186359B (en) | Integrated management method, system and storage medium for multi-source heterogeneous data of universities | |
| CN113961712A (en) | Knowledge graph-based fraud telephone analysis method | |
| CN116775879A (en) | Fine tuning training method of large language model, contract risk review method and system | |
| Omonov | Using the theories of fuzzy sets for researching the processes of diagnostics of data communication networks | |
| CN115964570B (en) | Cloud service recommendation method and device based on QoS multi-period change feature prediction | |
| CN112988948B (en) | Service processing method and device | |
| Ballin et al. | Optimization of stratified sampling with the r package samplingstrata: applications to network data | |
| Elsaid et al. | Proposed framework for planning software releases using fuzzy rule‐based system | |
| CN110175700A (en) | Adult degree English grade examzation prediction technique based on weighting depth forest | |
| da Silva et al. | Automated machine learning for time series prediction | |
| CN117893216B (en) | User behavior analysis method and system based on big data | |
| Kvet et al. | Non-dominated Set of Public Service System Designs using Simulated Annealing Approach | |
| Mozhaiev et al. | Method of modeling of a social profile using big data structure transformation optimization | |
| Sason et al. | ORACLE‑BASED DATA GENERATION FOR HIGHLY EFFICIENT DIGITAL TWIN NETWORK TRAINING | |
| CN114820085B (en) | User screening method, related device and storage medium | |
| Antoniou et al. | Preferred solutions of the ground station scheduling problem using NSGA-III with weighted reference points selection | |
| Madanayake et al. | A strategic approach to exploiting and implementing BIM, Big Data Analytics, and Internet of Things (BBI) for competitive advantage in the construction industry |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |