CN119831663A - Medical intelligent marketing system based on SCRM - Google Patents
Medical intelligent marketing system based on SCRM Download PDFInfo
- Publication number
- CN119831663A CN119831663A CN202411926318.2A CN202411926318A CN119831663A CN 119831663 A CN119831663 A CN 119831663A CN 202411926318 A CN202411926318 A CN 202411926318A CN 119831663 A CN119831663 A CN 119831663A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- formula
- marketing
- module
- 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
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention discloses a medical intelligent marketing system based on SCRM, which belongs to the technical field of marketing in the medical industry and comprises a data acquisition module, a data processing and analyzing module, an AI recommendation engine module, a client interaction platform module and an effect evaluation and feedback mechanism module, wherein the invention guides a user to access public information of the user through an OAuth2.0 authorization protocol by using an official API interface provided by each large social media platform, authorizes application to call the API interface by using an HTTP request library to acquire the public information of the user, encrypts sensitive data by using a symmetric encryption technology, collects real-time data such as search records and page browsing records of the user through a front-end application and a website, triggers a recommendation system to generate new recommended content when detecting relevant information of a user searching a certain disease, then matches related science popularization articles and expert lecture videos from a content library according to search keywords of the user, pushes the matched content to the user through a popup window and a notification column, supports multilingual switching in the front-end application, and the user can select different language versions to meet the requirements of international patients.
Description
Technical Field
The invention belongs to the technical field of marketing in the medical industry, and particularly relates to a medical intelligent marketing system based on SCRM.
Background
With the acceleration of digital transformation, the service mode of the medical industry is undergoing a deep change from a traditional mode to digital and intelligent, on one hand, the needs of patients on medical services are more and more diversified and personalized, on the other hand, the medical institutions face challenges of effectively touching target groups, improving the viscosity of the patients, optimizing resource allocation and the like, and on the other hand, under the background, the development of a set of medical intelligent marketing system capable of integrating online and offline resources, deeply analyzing user behaviors and providing personalized services is particularly important;
However, the prior art is insufficient in data integration capability, the existing medical information system is often scattered in different platforms and departments, the phenomenon of data islanding is serious, data cannot be effectively integrated and shared, data formats and standards of different sources are not uniform, the difficulty of data integration is increased, the quality and usability of the data are affected, the prior art mainly depends on static information and a small amount of behavior data of users, multi-dimensional dynamic data support is lacking, the user portrait is not comprehensive and accurate enough, meanwhile, in the process of constructing the user portrait, protection measures for user privacy are insufficient, and concern of users on data safety is easily caused.
Disclosure of Invention
The invention aims to provide an SCRM-based medical intelligent marketing system so as to solve the problems in the background technology.
In order to achieve the aim, the medical intelligent marketing system based on SCRM comprises a data acquisition module, a data processing and analyzing module, an AI recommendation engine module, a client interaction platform module and an effect evaluation and feedback mechanism module;
the data acquisition module can collect the multidimensional data of personal information, health condition and consumption behavior of a patient from a hospital information system, a health application program and a plurality of social media channels;
The data processing and analyzing module is used for preprocessing the collected data, performing data cleaning, format conversion and outlier detection operation, and performing deep mining on the data by adopting an advanced data analysis technology and a machine learning algorithm to construct a detailed user portrait;
The AI recommendation engine provides personalized medical service or product recommendation based on user portraits and historical interaction data by using a deep learning technology, and the recommendation content covers multiple aspects of disease prevention, health management and disease treatment;
The client interaction platform supports various communication modes, including instant messaging, telephone and mail, and promotes effective communication between medical staff and patients;
the effect evaluation and feedback mechanism quantifies the effect of the marketing campaign by setting a KPI index system, and continuously optimizes a recommendation model and a service flow according to user feedback.
The data acquisition module guides a user to access public information of the user through an authority API interface provided by each large social media platform by using an OAuth 2.0 authorization protocol, an application is authorized to access the public information of the user and then calls the API interface by using an HTTP request library, the data acquisition module is compatible with a third-party health application interface, is compatible with related data information transmitted and acquired by interfaces of wearable equipment and health management APP, acquires more comprehensive patient health data in the same access authorization mode, calculates abstracts of the data by using a hash function when checking the data, compares whether the abstracts of two data sources are consistent, encrypts sensitive data by using a symmetric encryption technology, and an encryption formula is divided into four steps, namely an initial round key is added, the realization formula is as follows:
In the formula, P is a plaintext block (128 bits), K 0 is an initial key (128 bits), C 0 is the output of an initial round, and the following steps are carried out by multi-round conversion:
Si,j＝Sbox[Si,j]
In the formula, S i,j is the byte of the ith row and the jth column in the state matrix, S box is an S-box table, and the step three is final round transformation, and the realization formula is as follows:
In the formula, W [ i ] is the i-th word, N k is the word number T (W [ i-1 ]) of the key, and is a transformation function comprising byte substitution, cyclic shift and round constant exclusive OR, and the efficient encryption and decryption of data are realized by using a symmetric encryption algorithm, and the transformation comprises byte substitution, row shift, column mixing and key addition steps, so that the high security and the attack resistance of the algorithm are ensured.
The data processing and analyzing module analyzes unstructured data by using natural language processing and machine learning algorithms, extracts text features by using a TF-IDF method, and then selects a machine learning algorithm of a Support Vector Machine (SVM) for training, wherein the implementation formula is as follows:
TF-IDF(t,d0=TF(t,d)×IDF(t)
In the formula, TF (t, d) is the frequency of occurrence of the word t in the document d, IDF (t) is the inverse document frequency, and the calculation formula is:
In the formula, N is the total number of documents, N t is the number of documents containing word t, the potential requirements of patients and the possible health risks which are valuable for marketing decisions are identified through a trained model, and meanwhile, emotion analysis technology is introduced to select a naive Bayesian model through a machine learning method to carry out model training, so that the formula is realized:
In the formula, P (C k |x) is the posterior probability of a category C k under a given feature vector x, P (x|C k) is the conditional probability of the feature vector x under a category C k, P (C k) is the prior probability of the category C k, P (x) is the edge probability of the feature vector x, a trained model is used for carrying out emotion classification on a new text, judging positive, negative and neutral emotion, evaluating the attitude of a patient to specific services and products, assisting in the formulation of marketing strategies, a relationship network among users is constructed by utilizing a graph neural network technology, and the embedded representation of nodes is learned by using a GCN model, wherein the realization formula is as follows:
In the formula, H (l+1) is the node feature matrix of the first layer, Is an adjacency matrix added with a self-loop,Is thatW (l) is a weight matrix of a first layer, sigma is an activation function, potential community effects are found, support is provided for group marketing, and the technical means and the implementation steps together form an efficient and intelligent data processing and analyzing system, so that strong support is provided for other modules of the medical intelligent marketing system.
The AI recommendation engine module can dynamically adjust recommended content according to real-time requirements of users, firstly collects real-time data such as search records and page browsing records of the users through front-end application and websites, when detecting that the users search for relevant information of certain diseases, triggers a recommendation system to generate new recommended content, then matches relevant science popularization articles and expert lecture videos from a content library according to search keywords of the users, and pushes the matched content to the users through popup windows and notification columns, wherein the matching calculation of the similarity between the keywords and the content library comprises the following formulas:
in the formula, A and B are two vectors, respectively representing TF-IDF vectors of two documents, & represents direction
Volume dot product, lock A lock
Obtaining geographic position information of a user through an IP address of the user and a GPS positioning technology, updating a user portrait according to the geographic position and time factors of the user, increasing characteristics of geographic and time dimensions, generating recommendation according to historical behaviors of the user and behaviors of similar users through a collaborative filtering formula, wherein the implementation formula is as follows:
In the formula (i) the formula (ii), Is the predictive score of user u for item i,Is the average score for user u, sin (u, v) is the similarity between user u and user v, r vi is the actual score for user v for item i,The recommendation engine detects the user browser setting and the language preference set by the user through NLP technology, and translates the recommended content into the language selected by the user by using a machine translation model Transformer, and the realization formula is as follows:
in the formula, Q is a query vector, K is a key vector, V is a value vector, d k is a dimension of the key vector, softmax is a softmax function for normalizing probability distribution, multi-language switching is supported in front-end application, and a user can select different language versions to meet the requirements of international patients.
The client interaction platform module has an intelligent customer service function, collects and sorts common questions and answers thereof, builds a knowledge base, uses a Support Vector Machine (SVM) to identify intention of a user, and realizes the following formulas:
In the formula, Y is the intention of the user, x is the input text of the user, Y is all possible intention sets, P (y|x) is the probability of intention Y under the given input text x, meanwhile, the module supports the recognition and understanding of voice input by using automatic voice recognition, the front end converts the voice input of the user into text, the operation of cleaning and word segmentation preprocessing is carried out on the converted text, the intention of the user is recognized by using the voice recognition formula, key entities are extracted from the text, finally, proper answers are selected and corresponding operations are executed according to the intention and the entities of the user, and the realization formula is as follows:
In the formula (i) the formula (ii), Is the recognized text, X is the input voice signal, P (w|X) is the probability of the text w given the voice signal X, and the realization formula of entity extraction is:
In the formula, Y i is the entity tag of the ith word, x i is the ith word, x i-1 and x i+1 are adjacent words, Y is all possible entity tag sets, P (Y i|xi,xi-1,xi+1) is the probability of the entity tag under the given word and the context condition, and the intention of a user can be accurately captured and responded appropriately through intelligent customer service supporting voice recognition and natural language understanding technologies.
The effect evaluation and feedback mechanism module defines a series of KPI indexes according to marketing targets, including click rate, conversion rate and review rate, firstly collects click times, purchase times and review times data of users through front-end application, cleans, deduplicates and format conversion preprocessing operation is carried out on the collected data to ensure data quality, then calculates values of various indexes according to defined KPI formulas, finally displays the calculated results in a form of a chart and an instrument panel so as to facilitate viewing and analysis of management layers and marketing teams, meanwhile, the A/B test is carried out to compare effects of different marketing strategies to find out an optimal scheme, the module randomly divides the users into a plurality of groups, each group uses different marketing strategies, simultaneously runs marketing activities of a plurality of groups, records user behavior data of each group, analyzes data of each group by using a statistical method, evaluates effects of different strategies, and selects a strategy with optimal performance as a final scheme according to data analysis results, wherein the method is used for comparing whether classification variables of the groups have obvious differences or not by using a chi test, and the method is realized as follows:
In the formula, O i is an observation value, E i is an expected value, n is the number of classification variables, and finally the module periodically collects and gathers data of marketing activities, carries out deep analysis on the gathered data, identifies reasons of success and failure, compiles detailed marketing reports including data charts, analysis conclusions and recommended measures, distributes the marketing reports to relevant management layers, marketing teams and collects feedback comments of relevant personnel in a mail and conference mode, continuously perfects report contents and analysis methods, summarizes experience teaching training and guides future marketing work.
Wherein, the system comprises the following steps:
s1, acquiring relevant information of a patient through a data acquisition module, and ensuring legal compliance of data sources;
S2, generating a user portrait by utilizing a data processing and analyzing module, wherein the user portrait comprises basic information, health conditions and consumption habits;
S3, providing personalized service or product recommendation based on the user portrait by the AI recommendation engine, wherein the recommendation content is required to be subjected to strict auditing, so that the scientificity and the legality of the recommendation content are ensured;
s4, realizing doctor-patient communication through a client interaction platform and providing convenient consultation service;
S5, monitoring marketing campaign effectiveness by utilizing an effect evaluation and feedback mechanism, and timely adjusting strategies to improve marketing efficiency.
In the step S1, the data acquisition process strictly complies with laws and regulations related to network security laws and personal information protection laws, ensures information security and personal privacy protection, and clearly informs users of the purpose, scope and use rules of data collection, thereby respecting the user' S right of knowledge and right of choice.
In step S3, personalized recommendation considers not only preference of users, but also professionals and safety of medical services, avoids damage to benefits of patients due to excessive marketing, and for special groups (such as minors and elderly people), recommended content is cautious to prevent misguidance or adverse effects, and medical professionals are invited to examine the recommended content regularly, so that the recommended content meets the latest medical standard and ethical standard.
In step S5, besides the conventional marketing effect evaluation, the method further includes investigation of patient satisfaction as an important reference for optimizing services, conducting regular user investigation, collecting feedback about system functions, interface design, operation convenience and the like, continuously improving user experience, establishing a sound complaint processing mechanism, giving positive response to reasonable complaints of users, and maintaining good brand images.
Compared with the prior art, the invention has the beneficial effects that:
1. The invention can integrate multi-source data from different platforms and departments through the data acquisition module, break data islands, realize comprehensive sharing and high-efficiency utilization of data, ensure consistency and high quality of the data by establishing unified data standards and formats, improve the usability and reliability of the data, and simultaneously support real-time data stream processing by the system, timely capture the latest behaviors and demands of users and ensure timeliness and effectiveness of marketing strategies;
2. the invention not only considers static information (such as age and sex) of the user, but also synthesizes dynamic behavior data (such as search history and treatment record) of the user, and constructs a comprehensive and accurate user image, and adopts machine learning and natural language processing data mining technology to deeply analyze the real demands and potential preferences of the user, thereby improving the accuracy and satisfaction degree of personalized service;
3. The invention combines multiple recommendation algorithms of deep learning and graph neural network, improves the diversity and accuracy of recommendation, effectively solves the problem of cold start of new users or new articles by introducing content-based recommendation and new user behavior analysis, improves the accuracy of initial recommendation, and ensures that the recommended content always meets the current demands and interests of users by dynamically adjusting the recommended content according to the real-time behaviors of the users;
4. The invention establishes a KPI index system covering multiple aspects of click rate, conversion rate, review rate and the like, comprehensively and accurately measures the performance of marketing activities, supports scientific A/B test, finds an optimal scheme by comparing the effects of different marketing strategies, improves the scientificity and effectiveness of the marketing strategies, simultaneously can rapidly collect and process user feedback, timely adjusts the marketing strategies, and improves the response speed and market adaptability of the system;
5. the invention pays attention to the friendliness of the user interface and the interactive design, provides a concise and visual operation interface, improves the use experience and satisfaction of users, supports multiple language versions, can meet the requirements of international patients, and expands the application range and user groups of the system.
Drawings
FIG. 1 is a flowchart illustrating the operation of the SCRM-based medical intelligent marketing system of the present invention;
FIG. 2 is a second flowchart of operation of the SCRM-based medical intelligent marketing system of the present invention;
FIG. 3 is a flow chart of operation of the SCRM-based medical intelligent marketing system of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
1-3, The present invention provides a technical solution, including a data acquisition module, a data processing and analyzing module, an AI recommendation engine module, a client interaction platform module, and an effect evaluation and feedback mechanism module;
the data acquisition module can collect the multidimensional data of personal information, health condition and consumption behavior of a patient from a hospital information system, a health application program and a plurality of social media channels;
The data processing and analyzing module is used for preprocessing the collected data, performing data cleaning, format conversion and outlier detection operation, and performing deep mining on the data by adopting an advanced data analysis technology and a machine learning algorithm to construct a detailed user portrait;
The AI recommendation engine provides personalized medical service or product recommendation based on user portraits and historical interaction data by using a deep learning technology, and the recommendation content covers multiple aspects of disease prevention, health management and disease treatment;
The client interaction platform supports various communication modes, including instant messaging, telephone and mail, and promotes effective communication between medical staff and patients;
the effect evaluation and feedback mechanism quantifies the effect of the marketing campaign by setting a KPI index system, and continuously optimizes a recommendation model and a service flow according to user feedback.
The data acquisition module guides a user to access public information of the user through an authority API interface provided by each large social media platform by using an OAuth 2.0 authorization protocol, an application is authorized to access the public information of the user and then calls the API interface by using an HTTP request library, the data acquisition module is compatible with a third-party health application interface, is compatible with related data information transmitted and acquired by interfaces of wearable equipment and health management APP, acquires more comprehensive patient health data in the same access authorization mode, calculates abstracts of the data by using a hash function when checking the data, compares whether the abstracts of two data sources are consistent, encrypts sensitive data by using a symmetric encryption technology, and an encryption formula is divided into four steps, namely an initial round key is added, the realization formula is as follows:
In the formula, P is a plaintext block (128 bits), K 0 is an initial key (128 bits), C 0 is the output of an initial round, and the following steps are carried out by multi-round conversion:
Si,j＝Sbox[Si,j]
In the formula, S i,j is the byte of the ith row and the jth column in the state matrix, S box is an S-box table, and the step three is final round transformation, and the realization formula is as follows:
In the formula, W [ i ] is the i-th word, N k is the word number T (W [ i-1 ]) of the key, and is a transformation function comprising byte substitution, cyclic shift and round constant exclusive OR, and the efficient encryption and decryption of data are realized by using a symmetric encryption algorithm, and the transformation comprises byte substitution, row shift, column mixing and key addition steps, so that the high security and the attack resistance of the algorithm are ensured.
The data processing and analyzing module analyzes unstructured data by using natural language processing and machine learning algorithms, extracts text features by using a TF-IDF method, and then selects a machine learning algorithm of a Support Vector Machine (SVM) for training, wherein the implementation formula is as follows:
TF-IDF(t,d)=TF(t,d)×IDF(t)
In the formula, TF (t, d) is the frequency of occurrence of the word t in the document d, IDF (t) is the inverse document frequency, and the calculation formula is:
In the formula, N is the total number of documents, N t is the number of documents containing word t, the potential requirements of patients and the possible health risks which are valuable for marketing decisions are identified through a trained model, and meanwhile, emotion analysis technology is introduced to select a naive Bayesian model through a machine learning method to carry out model training, so that the formula is realized:
In the formula, P (C k |x) is the posterior probability of a category C k under a given feature vector x, P (x|C k) is the conditional probability of the feature vector x under a category C k, P (C k) is the prior probability of the category C k, P (x) is the edge probability of the feature vector x, a trained model is used for carrying out emotion classification on a new text, judging positive, negative and neutral emotion, evaluating the attitude of a patient to specific services and products, assisting in the formulation of marketing strategies, a relationship network among users is constructed by utilizing a graph neural network technology, and the embedded representation of nodes is learned by using a GCN model, wherein the realization formula is as follows:
In the formula, H (l+1) is the node feature matrix of the first layer, Is an adjacency matrix added with a self-loop,Is thatW (l) is a weight matrix of a first layer, sigma is an activation function, potential community effects are found, support is provided for group marketing, and the technical means and the implementation steps together form an efficient and intelligent data processing and analyzing system, so that strong support is provided for other modules of the medical intelligent marketing system.
The AI recommendation engine module can dynamically adjust recommended content according to real-time requirements of users, firstly collects real-time data such as search records and page browsing records of the users through front-end application and websites, when detecting that the users search for relevant information of certain diseases, triggers a recommendation system to generate new recommended content, then matches relevant science popularization articles and expert lecture videos from a content library according to search keywords of the users, and pushes the matched content to the users through popup windows and notification columns, wherein the matching calculation of the similarity between the keywords and the content library comprises the following formulas:
in the formula, A and B are two vectors, respectively representing TF-IDF vectors of two documents, & represents direction
Volume dot product, lock A lock
Obtaining geographic position information of a user through an IP address of the user and a GPS positioning technology, updating a user portrait according to the geographic position and time factors of the user, increasing characteristics of geographic and time dimensions, generating recommendation according to historical behaviors of the user and behaviors of similar users through a collaborative filtering formula, wherein the implementation formula is as follows:
In the formula (i) the formula (ii), Is the predictive score of user u for item i,Is the average score for user u, sin (u, v is the similarity between user u and user v, r vi is the actual score for user v for item i,The recommendation engine detects the user browser setting and the language preference set by the user through NLP technology, and translates the recommended content into the language selected by the user by using a machine translation model Transformer, and the realization formula is as follows:
in the formula, Q is a query vector, K is a key vector, V is a value vector, d k is a dimension of the key vector, softmax is a softmax function for normalizing probability distribution, multi-language switching is supported in front-end application, and a user can select different language versions to meet the requirements of international patients.
The client interaction platform module has an intelligent customer service function, collects and sorts common questions and answers thereof, builds a knowledge base, uses a Support Vector Machine (SVM) to identify intention of a user, and realizes the following formulas:
In the formula, Y is the intention of the user, x is the input text of the user, Y is all possible intention sets, P (y|x) is the probability of intention Y under the given input text x, meanwhile, the module supports the recognition and understanding of voice input by using automatic voice recognition, the front end converts the voice input of the user into text, the operation of cleaning and word segmentation preprocessing is carried out on the converted text, the intention of the user is recognized by using the voice recognition formula, key entities are extracted from the text, finally, proper answers are selected and corresponding operations are executed according to the intention and the entities of the user, and the realization formula is as follows:
In the formula (i) the formula (ii), Is the recognized text, X is the input voice signal, P (w|X) is the probability of the text w given the voice signal X, and the realization formula of entity extraction is:
In the formula, Y i is the entity tag of the ith word, x i is the ith word, x i-1 and x i+1 are adjacent words, Y is all possible entity tag sets, P (Y i|xi,xi-1,xi+1) is the probability of the entity tag under the given word and the context condition, and the intention of a user can be accurately captured and responded appropriately through intelligent customer service supporting voice recognition and natural language understanding technologies.
The effect evaluation and feedback mechanism module defines a series of KPI indexes according to marketing targets, including click rate, conversion rate and review rate, firstly collects click times, purchase times and review times data of users through front-end application, cleans, deduplicates and format conversion preprocessing operation is carried out on the collected data to ensure data quality, then calculates values of various indexes according to defined KPI formulas, finally displays the calculated results in a form of a chart and an instrument panel so as to facilitate viewing and analysis of management layers and marketing teams, meanwhile, the A/B test is carried out to compare effects of different marketing strategies to find out an optimal scheme, the module randomly divides the users into a plurality of groups, each group uses different marketing strategies, simultaneously runs marketing activities of a plurality of groups, records user behavior data of each group, analyzes data of each group by using a statistical method, evaluates effects of different strategies, and selects a strategy with optimal performance as a final scheme according to data analysis results, wherein the method is used for comparing whether classification variables of the groups have obvious differences or not by using a chi test, and the method is realized as follows:
In the formula, O i is an observation value, E i is an expected value, n is the number of classification variables, and finally the module periodically collects and gathers data of marketing activities, carries out deep analysis on the gathered data, identifies reasons of success and failure, compiles detailed marketing reports including data charts, analysis conclusions and recommended measures, distributes the marketing reports to relevant management layers, marketing teams and collects feedback comments of relevant personnel in a mail and conference mode, continuously perfects report contents and analysis methods, summarizes experience teaching training and guides future marketing work.
Wherein, the system comprises the following steps:
s1, acquiring relevant information of a patient through a data acquisition module, and ensuring legal compliance of data sources;
S2, generating a user portrait by utilizing a data processing and analyzing module, wherein the user portrait comprises basic information, health conditions and consumption habits;
S3, providing personalized service or product recommendation based on the user portrait by the AI recommendation engine, wherein the recommendation content is required to be subjected to strict auditing, so that the scientificity and the legality of the recommendation content are ensured;
s4, realizing doctor-patient communication through a client interaction platform and providing convenient consultation service;
S5, monitoring marketing campaign effectiveness by utilizing an effect evaluation and feedback mechanism, and timely adjusting strategies to improve marketing efficiency.
In the step S1, the data acquisition process strictly complies with laws and regulations related to network security laws and personal information protection laws, ensures information security and personal privacy protection, and clearly informs users of the purpose, scope and use rules of data collection, thereby respecting the user' S right of knowledge and right of choice.
In step S3, personalized recommendation considers not only preference of users, but also professionals and safety of medical services, avoids damage to benefits of patients due to excessive marketing, and for special groups (such as minors and elderly people), recommended content is cautious to prevent misguidance or adverse effects, and medical professionals are invited to examine the recommended content regularly, so that the recommended content meets the latest medical standard and ethical standard.
In step S5, besides the conventional marketing effect evaluation, the method further includes investigation of patient satisfaction as an important reference for optimizing services, conducting regular user investigation, collecting feedback about system functions, interface design, operation convenience and the like, continuously improving user experience, establishing a sound complaint processing mechanism, giving positive response to reasonable complaints of users, and maintaining good brand images.
The system utilizes the official API interface provided by each large social media platform to acquire the public information of the user in an authorized access mode, guides the user to access the public information of the user through an OAuth 2.0 authorization protocol, uses an HTTP request library to call the API interface to acquire the public information of the user, analyzes the returned JSON or XML format data into usable structured data, and finally stores the analyzed data into a MongoDB database;
The system acquires geographic position information of the user through the IP address of the user and a GPS positioning technology, records the behavior time of the user, such as access time and search time, updates user portraits according to the geographic position and time factors of the user, increases the characteristics of geographic and time dimensions, combines the geographic position and time factors of the user and generates personalized recommended content;
The client interaction platform module has natural language processing and dialogue management system technology, realizes the function of automatically answering common questions, collects and sorts the common questions and answers thereof, constructs a knowledge base, uses natural language understanding technology to identify the intention of a user, selects proper answers or executes corresponding operations according to the intention and the context of the user, generates natural language replies, and sends the natural language replies to the user in a text or voice form;
A series of KPI indexes of click rate, conversion rate and review rate are defined according to marketing targets, the front-end application is used for collecting action data of click times, purchase times and review times of users, preprocessing operations such as cleaning, duplication removal and format conversion are carried out on the collected data, data quality is ensured, values of various indexes are calculated according to a defined KPI formula, and calculation results are displayed in the forms of charts, dashboards and the like, so that a management layer and marketing team can check and analyze conveniently.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations may be made therein without departing from the spirit and scope of the invention as defined by the appended claims and their equivalents.
The invention and its embodiments have been described above with no limitation, and the actual construction is not limited to the embodiments of the invention as shown in the drawings. In summary, if one of ordinary skill in the art is informed by this disclosure, a structural manner and an embodiment similar to the technical solution should not be creatively devised without departing from the gist of the present invention.
Claims (10)
1. The SCRM-based medical intelligent marketing system comprises a data acquisition module, a data processing and analyzing module, an AI recommendation engine module, a client interaction platform module and an effect evaluation and feedback mechanism module;
the data acquisition module can collect the multidimensional data of personal information, health condition and consumption behavior of a patient from a hospital information system, a health application program and a plurality of social media channels;
The data processing and analyzing module is used for preprocessing the collected data, performing data cleaning, format conversion and outlier detection operation, and performing deep mining on the data by adopting an advanced data analysis technology and a machine learning algorithm to construct a detailed user portrait;
The AI recommendation engine provides personalized medical service or product recommendation based on user portraits and historical interaction data by using a deep learning technology, and the recommendation content covers multiple aspects of disease prevention, health management and disease treatment;
The client interaction platform supports various communication modes, including instant messaging, telephone and mail, and promotes effective communication between medical staff and patients;
the effect evaluation and feedback mechanism quantifies the effect of the marketing campaign by setting a KPI index system, and continuously optimizes a recommendation model and a service flow according to user feedback.
2. The SCRM-based medical intelligent marketing system of claim 1, wherein the data acquisition module utilizes an official API interface provided by each large social media platform to guide a user to access public information of the user through an OAuth2.0 authorization protocol, an authorized application calls the API interface by using an HTTP request library to acquire the public information of the user, meanwhile, the data acquisition module is compatible with a third-party health application interface, and is compatible with related data information transmitted by interfaces of wearable equipment and health management APP, and more comprehensive patient health data is acquired in a mode of authorizing access, when checking the data, a hash function is used for calculating the abstract of the data, whether the abstract of two data sources are consistent or not is compared, a symmetric encryption technology is used for encrypting the sensitive data, and an encryption formula is divided into four steps, namely, an initial round key is added, the realization formula is as follows:
In the formula, P is a plaintext block (128 bits), K 0 is an initial key (128 bits), C 0 is the output of an initial round, and the following steps are carried out by multi-round conversion:
Si,j＝Sbox[Si,j]
In the formula, S i,j is the byte of the ith row and the jth column in the state matrix, S box is an S-box table, and the step three is final round transformation, and the realization formula is as follows:
In the formula, W [ i ] is the i-th word, N k is the word number T (W [ i-1 ]) of the key, and is a transformation function comprising byte substitution, cyclic shift and round constant exclusive OR, and the efficient encryption and decryption of data are realized by using a symmetric encryption algorithm, and the transformation comprises byte substitution, row shift, column mixing and key addition steps, so that the high security and the attack resistance of the algorithm are ensured.
3. The SCRM-based medical intelligent marketing system of claim 1, wherein the data processing and analysis module parses unstructured data using natural language processing and machine learning algorithms, extracts text features using a TF-IDF method and then selects a machine learning algorithm supporting a vector machine SVM for training, the implementation formula is:
TF-IDF(t,d)=TF(t,d)×IDF(t)
In the formula, TF (t, d) is the frequency of occurrence of the word t in the document d, IDF (t) is the inverse document frequency, and the calculation formula is:
In the formula, N is the total number of documents, N t is the number of documents containing word t, the potential requirements of patients and the possible health risks which are valuable for marketing decisions are identified through a trained model, and meanwhile, emotion analysis technology is introduced to select a naive Bayesian model through a machine learning method to carry out model training, so that the formula is realized:
In the formula, P (C k |x) is the posterior probability of a category C k under a given feature vector x, P (x|C k) is the conditional probability of the feature vector x under a category C k, P (C k) is the prior probability of the category C k, P (x) is the edge probability of the feature vector x, a trained model is used for carrying out emotion classification on a new text, judging positive, negative and neutral emotion, evaluating the attitude of a patient to specific services and products, assisting in the formulation of marketing strategies, a relationship network among users is constructed by utilizing a graph neural network technology, and the embedded representation of nodes is learned by using a GCN model, wherein the realization formula is as follows:
In the formula, H (l+1) is the node feature matrix of the first layer, Is an adjacency matrix added with a self-loop,Is thatW (l) is a weight matrix of a first layer, sigma is an activation function, potential community effects are found, support is provided for group marketing, and the technical means and the implementation steps together form an efficient and intelligent data processing and analyzing system, so that strong support is provided for other modules of the medical intelligent marketing system.
4. The SCRM-based medical intelligent marketing system of claim 1, wherein the AI recommendation engine module is capable of dynamically adjusting recommended content according to real-time requirements of users, collecting real-time data of search records and page browsing records of the users through front-end application and websites, triggering the recommendation system to generate new recommended content when detecting relevant information of searching a certain disease of the users, matching relevant science popularization articles and expert lecture videos from a content library according to search keywords of the users, and pushing the matched content to the users through popup windows and notification bars, wherein the similarity matching calculation formula of the keywords and the content library is as follows:
in the formula, A and B are two vectors, representing the TF-IDF vectors of the two documents, respectively, representing the vector dot product, A-
And (B) per represents the modulus of the vectors A and B respectively, acquiring the geographic position information of the user through the IP address of the user and the GPS positioning technology, updating the user portrait according to the geographic position and the time factor of the user, increasing the characteristics of geographic dimension and time dimension,
Generating recommendation according to the historical behaviors of the user and the behaviors of similar users through a collaborative filtering formula, wherein the realization formula is as follows:
In the formula (i) the formula (ii), Is the predictive score of user u for item i,Is the average score for user u, sin (u, v is the similarity between user u and user v, r vi is the actual score for user v for item i,The recommendation engine detects the user browser setting and the language preference set by the user through NLP technology, and translates the recommended content into the language selected by the user by using a machine translation model Transformer, and the realization formula is as follows:
in the formula, Q is a query vector, K is a key vector, V is a value vector, d k is a dimension of the key vector, softmax is a softmax function for normalizing probability distribution, multi-language switching is supported in front-end application, and a user can select different language versions to meet the requirements of international patients.
5. The SCRM-based medical intelligent marketing system of claim 1, wherein the customer interaction platform module has intelligent customer service function, collects and sorts common questions and answers thereof, constructs a knowledge base, uses a support vector machine SVM to identify intent, identifies intent of a user, and realizes the following formula:
In the formula, Y is the intention of the user, x is the input text of the user, Y is all possible intention sets, P (y|x) is the probability of intention Y under the given input text x, meanwhile, the module supports the recognition and understanding of voice input by using automatic voice recognition, the front end converts the voice input of the user into text, the operation of cleaning and word segmentation preprocessing is carried out on the converted text, the intention of the user is recognized by using the voice recognition formula, key entities are extracted from the text, finally, proper answers are selected and corresponding operations are executed according to the intention and the entities of the user, and the realization formula is as follows:
In the formula (i) the formula (ii), Is the recognized text, X is the input voice signal, P (w|X) is the probability of the text w given the voice signal X, and the realization formula of entity extraction is:
In the formula, Y i is the entity tag of the ith word, x i is the ith word, x i-1 and x i+1 are adjacent words, Y is all possible entity tag sets, P (Y i|xi,xi-1,xi+1) is the probability of the entity tag under the given word and the context condition, and the intention of a user can be accurately captured and responded appropriately through intelligent customer service supporting voice recognition and natural language understanding technologies.
6. The SCRM-based medical intelligent marketing system of claim 1, wherein the effect evaluation and feedback mechanism module defines a series of KPI indexes according to marketing targets, including click rate, conversion rate and review rate, firstly collects click times, purchase times and review times data of users through front-end application, performs cleaning, deduplication and format conversion preprocessing operations on the collected data to ensure data quality, calculates values of each index according to a defined KPI formula, finally displays calculation results in a form of a chart and an instrument panel so as to facilitate viewing and analysis of management layers and marketing teams, meanwhile, compares effects of different marketing strategies through A/B tests to find out an optimal scheme, the module randomly divides users into a plurality of groups, each group uses different marketing strategies, simultaneously operates marketing activities of a plurality of groups under the same environment, records user behavior data of each group, analyzes data of each group by using a statistical method, evaluates effects of different strategies, and selects a strategy with optimal performance as a final scheme according to data analysis results, wherein whether the groups have significant differences are realized by using a plurality of comparison formulas or not:
In the formula, O i is an observation value, E i is an expected value, n is the number of classification variables, and finally the module periodically collects and gathers data of marketing activities, carries out deep analysis on the gathered data, identifies reasons of success and failure, compiles detailed marketing reports including data charts, analysis conclusions and recommended measures, distributes the marketing reports to relevant management layers, marketing teams and collects feedback comments of relevant personnel in a mail and conference mode, continuously perfects report contents and analysis methods, summarizes experience teaching training and guides future marketing work.
7. The SCRM-based medical intelligent marketing system of claim 1, wherein the system comprises the steps of:
s1, acquiring relevant information of a patient through a data acquisition module, and ensuring legal compliance of data sources;
S2, generating a user portrait by utilizing a data processing and analyzing module, wherein the user portrait comprises basic information, health conditions and consumption habits;
S3, providing personalized service or product recommendation based on the user portrait by the AI recommendation engine, wherein the recommendation content is required to be subjected to strict auditing, so that the scientificity and the legality of the recommendation content are ensured;
s4, realizing doctor-patient communication through a client interaction platform and providing convenient consultation service;
S5, monitoring marketing campaign effectiveness by utilizing an effect evaluation and feedback mechanism, and timely adjusting strategies to improve marketing efficiency.
8. The SCRM-based medical intelligent marketing system of claim 7, wherein in step S1, the data collection process strictly complies with laws and regulations related to the laws and regulations of network Security and personal information protection, ensures information security and personal privacy protection, and explicitly informs the user of the purpose, scope and usage rules of data collection, respecting the user' S right of awareness and option.
9. The SCRM-based medical intelligent marketing system of claim 7, wherein in step S3, personalized recommendations take into account not only the preferences of the user, but also the professionals and security of the medical service, avoiding damaging patient benefits due to overstocking, recommended content is more cautious for special people (e.g., minors, elderly), preventing misguidance or adverse effects, and medical professionals are invited to review the recommended content regularly, ensuring that it meets the latest medical standards and ethics specifications.
10. The SCRM-based medical intelligent marketing system of claim 7, wherein in step S5, in addition to the conventional marketing effect evaluation, the system further comprises investigation of patient satisfaction as an important reference for optimizing services, and further comprises the steps of conducting regular user investigation, collecting feedback on aspects of system functions, interface design, operation convenience and the like, continuously improving user experience, establishing a sound complaint handling mechanism, giving positive response to reasonable complaints of users, and maintaining good brand images.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411926318.2A CN119831663A (en) | 2024-12-25 | 2024-12-25 | Medical intelligent marketing system based on SCRM |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411926318.2A CN119831663A (en) | 2024-12-25 | 2024-12-25 | Medical intelligent marketing system based on SCRM |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN119831663A true CN119831663A (en) | 2025-04-15 |
Family
ID=95303766
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411926318.2A Pending CN119831663A (en) | 2024-12-25 | 2024-12-25 | Medical intelligent marketing system based on SCRM |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119831663A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120416297A (en) * | 2025-07-03 | 2025-08-01 | 深圳圣马歌科技有限公司 | An intelligent interactive system based on APP five-network integration technology and its implementation method |
-
2024
- 2024-12-25 CN CN202411926318.2A patent/CN119831663A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120416297A (en) * | 2025-07-03 | 2025-08-01 | 深圳圣马歌科技有限公司 | An intelligent interactive system based on APP five-network integration technology and its implementation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10672012B2 (en) | Brand personality comparison engine | |
| US11315149B2 (en) | Brand personality inference and recommendation system | |
| CN110598016A (en) | Method, device, equipment and medium for recommending multimedia information | |
| WO2016035072A2 (en) | Sentiment rating system and method | |
| Perdana et al. | Combining likes-retweet analysis and naive bayes classifier within twitter for sentiment analysis | |
| US20140032657A1 (en) | System and Method for Measuring the Positive or Negative Impact of Digital and Social Media Content on Intent and Behavior | |
| Zhang et al. | The impact of big data on research methods in information science | |
| CN119202140B (en) | Data report generation method, electronic device, storage medium and computer program product | |
| CN111104590A (en) | Information recommendation method, device, medium and electronic equipment | |
| Wang et al. | Research on influencing factors and governance of disinformation dissemination on science and technology topics: an empirical study on the topic of “metaverse” | |
| CN118503373A (en) | Intelligent interactive question-answering system in multichannel safety emergency knowledge field | |
| CN119831663A (en) | Medical intelligent marketing system based on SCRM | |
| CN117807323A (en) | Online interactive smart Prime big data platform | |
| Qiu et al. | Why do knowledge seekers turn to knowledge contributors in social Q&A platforms? Evidence from China | |
| CN117079772A (en) | Intelligent correction system and terminal based on mental evaluation analysis of community correction object | |
| CN120508701B (en) | Personalized Visual Communication Element Recommendation Method and System Based on Internet Big Data | |
| Liu et al. | From News to Knowledge: Predicting Hate Crime Trends through Event Extraction from Media Content. | |
| CN112968873A (en) | Encryption method and device for private data transmission | |
| Alshanqiti et al. | Intelligent parallel mixed method approach for characterising viral Youtube videos in Saudi Arabia | |
| CN117726231A (en) | A video customer service quality analysis method | |
| Jiang | The role of social media in news avoidance: A cognitive load and technology affordance perspective | |
| Uğur et al. | Investigatingsocialmediaactivities: Astudyoncelebrity posts | |
| O'Neill et al. | Rating by ranking: An improved scale for judgement-based labels | |
| Zhang et al. | Analysis and research on factors affecting information dissemination of emergencies in social media environment | |
| CN117370448B (en) | Brand digital asset insight analysis method |
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 |