CN113987186A

CN113987186A - Method and device for generating marketing scheme based on knowledge graph

Info

Publication number: CN113987186A
Application number: CN202111311376.0A
Authority: CN
Inventors: 李响; 杜正平; 徐勇; 高峰
Original assignee: Beijing Borui Tongyun Technology Co ltd
Current assignee: Beijing Borui Tongyun Technology Co ltd
Priority date: 2021-11-08
Filing date: 2021-11-08
Publication date: 2022-01-28
Anticipated expiration: 2041-11-08
Also published as: CN113987186B

Abstract

The embodiment of the invention relates to a method and a device for generating a marketing scheme based on a knowledge graph, wherein the method comprises the following steps: acquiring first marketing object type data and name data; setting first marketing target type data; marking a first starting node and a first ending node in a first knowledge graph; identifying an optimal node edge connecting line from the first starting node to each first ending node to generate a first optimal node edge connecting line; selecting the shortest node edge connecting line; marking a second end node and a first middle node on the shortest node edge connecting line; performing key intermediate node identification on each first intermediate node and generating a first marketing scheme associated data set according to an identification result; acquiring the associated data of the second end node to generate initial data of a first marketing scheme; and adjusting the initial data of the first marketing scheme by referring to the associated data set of the first marketing scheme to generate data of the first marketing scheme. The method can greatly improve the accuracy and timeliness of scheme formulation.

Description

Method and device for generating marketing scheme based on knowledge graph

Technical Field

The invention relates to the technical field of knowledge graphs, in particular to a method and a device for generating a marketing scheme based on a knowledge graph.

Background

In the field of insurance business, customers are classified mainly by the manual experience of a customer manager for a long time, and the dangerous marketing scheme of the customers is planned based on the manual classification result; after enough client dangerous marketing schemes are settled, the characteristics of the crowd are analyzed based on the manual experience of the data analyst, and the marketing scheme of the crowd is planned according to the marketing activities of the newly added services based on the manual analysis result. Because the manual participation links in the marketing planning process are too many, the problems that the positioning is not accurate enough due to insufficient information understanding, the analysis is not timely enough due to too large information amount and the like often occur.

The Knowledge Graph (Knowledge Graph) technology is one of important branches of an artificial intelligence technology, can display the association between entities in a visual mode, can discover new information, a new unstructured data mode and new Knowledge more quickly and simply, and achieves the purposes of insights on customers and reduction of business transaction risks. The knowledge graph is essentially a semantic network, and is a data structure based on a graph, and consists of nodes (points) and edges (edges). Nodes of the knowledge graph correspond to entities existing in the real world, and each node has attributes such as corresponding identification, name and type. The edges of the knowledge graph refer to connecting lines connecting two nodes and are also called node edges, and each node edge is provided with head and tail node identifications for identifying the mapping direction and attributes such as relationship weight reflecting the association degree between the nodes, namely between entities. In brief, all kinds of information can be connected together through the knowledge graph to obtain a relationship network, and the possibility of intersection generation between different entities can be analyzed through the connection relationship in the graph.

Disclosure of Invention

The invention aims to provide a method, a device, electronic equipment and a computer readable storage medium for generating a marketing scheme based on a knowledge graph, the knowledge graph suitable for the insurance field is created, different types of nodes are created for entities in the field, such as customers, crowds, insurance seeds, marketing activities, field knowledge points, commodity knowledge points, disease knowledge points and the like in the graph, node edges reflecting the incidence relation are established among the nodes, and the formulation of the marketing scheme of the danger seeds and the marketing scheme of the crowds is guided based on the knowledge graph. The method can completely get rid of the influence of artificial factors and greatly improve the precision and timeliness of the scheme.

In order to achieve the above object, a first aspect of the embodiments of the present invention provides a method for generating a marketing solution based on a knowledge graph, where the method includes:

acquiring first marketing object type data and first marketing object name data;

setting first marketing target type data according to the first marketing object type data; when the first marketing object type data is a customer type, setting the first marketing target type data as an insurance type; when the first marketing object type data is a marketing activity class, setting the first marketing target type data as a crowd class;

in a preset first knowledge graph, marking a first node with first node type data matched with the first marketing object type data and first node name data matched with the first marketing object name data as a first starting node, and marking all the first nodes with the first node type data matched with the first marketing target type data as first ending nodes; the first knowledge-graph comprises a plurality of the first nodes; each first node corresponds to a first node data group; the first node data group at least comprises first node identification data, first node name data and first node type data; the first node type data at least comprises a customer class, a crowd class, a risk class and a marketing campaign class;

in the first knowledge graph, identifying an optimal node edge connecting line from the first starting node to each first ending node to generate a corresponding first optimal node edge connecting line; the first knowledge-graph further comprises a plurality of first node edges; each first node edge corresponds to a first node edge data group; the first node edge data group at least comprises first head node identification data and first tail node identification data; the first optimal node edge connecting line is formed by sequentially connecting a plurality of first node edges; on the first optimal node edge connecting line corresponding to each first end node, the first end node identification data of a first one of the first node edges matches with the first node identification data of the first start node, the first head node identification data of a previous one of the two adjacent first node edges matches with the first end node identification data of a next one of the first node edges, and the first head node identification data of a last one of the first node edges matches with the first node identification data of a current first end node;

selecting the first optimal node edge connecting line with the least number of first node edges from the obtained plurality of first optimal node edge connecting lines as a shortest node edge connecting line;

on the shortest node edge connecting line, marking the last first node as a second end node, and marking other first nodes between the first starting node and the second end node as first intermediate nodes;

performing key intermediate node identification on each first intermediate node according to the first marketing object type data, and performing marketing scheme associated data collection according to the identified key intermediate nodes to generate a first marketing scheme associated data set;

acquiring the associated data of the second end node according to the first marketing target type data to generate first marketing scheme initial data;

and adjusting the initial data of the first marketing scheme by referring to the associated data set of the first marketing scheme to generate data of the first marketing scheme.

Preferably, the identifying, in the first knowledge graph, an optimal node edge connecting line from the first start node to each first end node to generate a corresponding first optimal node edge connecting line specifically includes:

taking each first end point as a first current end point; counting all node edge connecting lines from the first starting node to the first current ending point in the first knowledge graph to obtain a plurality of first node edge connecting lines; the first node edge connecting line is formed by sequentially connecting a plurality of first node edges;

selecting the first node edge connection line containing the minimum number of the first node edges from the obtained plurality of first node edge connection lines as the first optimal node edge connection line corresponding to the first current end point.

Preferably, the performing key intermediate node identification on each first intermediate node according to the first marketing object type data, and performing marketing scheme associated data collection according to the identified key intermediate node to generate a first marketing scheme associated data set specifically includes:

performing corresponding association weight value calculation on each first intermediate node according to the first marketing object type data to generate a corresponding first association weight value;

recording the first intermediate node corresponding to the first association weight value exceeding a preset association weight threshold as a first key intermediate node;

and collecting the associated data of all the first key intermediate nodes to generate the first marketing scheme associated data set.

Preferably, the first node edge data group further includes first relationship weight data, and the calculating of the corresponding association weight value of each first intermediate node according to the first marketing object type data to generate a corresponding first association weight value specifically includes:

when the first marketing object type data is a customer type, taking a product of continuous multiplication of the first relation weight data of all the first node edges between the first starting node and the current first intermediate node as the first associated weight value corresponding to the current first intermediate node;

and when the first marketing object type data is a marketing activity class, taking a product of continuous multiplication of the first relation weight data of all the first node edges between the current first intermediate node and the second end node as the first associated weight value corresponding to the current first intermediate node.

Further, the first node type data further includes topic knowledge, commodity knowledge, and disease knowledge, and the collecting the associated data of all the first key intermediate nodes to generate the first marketing scheme associated data set specifically includes:

identifying the first node type data for each of the first critical intermediate nodes; if the first node type data of the current first key intermediate node is a topic knowledge class, extracting a first topic content field of a first topic knowledge record corresponding to the current first key intermediate node from a preset first topic knowledge base to serve as corresponding first associated text data; if the first node type data of the current first key intermediate node is a commodity knowledge class, extracting a first commodity content field of a first commodity knowledge record corresponding to the current first key intermediate node in a preset first commodity knowledge base to serve as corresponding first associated text data; if the first node type data of the current first key intermediate node is a disease knowledge class, extracting a first disease content field of a first disease knowledge record corresponding to the current first key intermediate node from a preset first disease knowledge base to serve as corresponding first associated text data;

forming corresponding first marketing scheme associated data by the first node name data, the first node type data and the first associated text data of each first key intermediate node;

and forming the first marketing scheme associated data set by the obtained plurality of the first marketing scheme associated data.

Preferably, the obtaining the associated data of the second end node according to the first marketing target type data to generate first marketing scheme initial data specifically includes:

when the first marketing target type data is of a dangerous type, extracting a first dangerous type term content field of a first dangerous type term knowledge record corresponding to the second end node from a preset first dangerous type term knowledge base to serve as corresponding second associated text data;

when the first marketing target type data is of a crowd type, extracting a first crowd characteristic content field of a first crowd characteristic knowledge record corresponding to the second end node from a preset first crowd characteristic knowledge base to serve as corresponding second associated text data;

composing the first marketing plan initial data from the first node name data and the second associated text data of the second end node.

A second aspect of the embodiments of the present invention provides an apparatus for generating a marketing plan based on a knowledge graph, which is used to implement the method of the first aspect, and includes: the system comprises an acquisition module, a first knowledge graph processing module, a second knowledge graph processing module, a third knowledge graph processing module, a marketing scheme data processing module and a marketing scheme integration module;

the acquisition module is used for acquiring first marketing object type data and first marketing object name data;

the knowledge graph first processing module is used for setting first marketing target type data according to the first marketing object type data; when the first marketing object type data is a customer type, setting the first marketing target type data as an insurance type; when the first marketing object type data is a marketing activity class, setting the first marketing target type data as a crowd class; in a preset first knowledge graph, marking a first node with first node type data matched with the first marketing object type data and first node name data matched with the first marketing object name data as a first starting node, and marking all the first nodes with the first node type data matched with the first marketing target type data as first ending nodes; wherein the first knowledge-graph comprises a plurality of the first nodes; each first node corresponds to a first node data group; the first node data group at least comprises first node identification data, first node name data and first node type data; the first node type data at least comprises a customer class, a crowd class, a risk class and a marketing campaign class;

the knowledge graph second processing module is used for identifying an optimal node edge connecting line from the first starting node to each first ending node in the first knowledge graph to generate a corresponding first optimal node edge connecting line; wherein the first knowledge-graph further comprises a plurality of first node edges; each first node edge corresponds to a first node edge data group; the first node edge data group at least comprises first head node identification data and first tail node identification data; the first optimal node edge connecting line is formed by sequentially connecting a plurality of first node edges; on the first optimal node edge connecting line corresponding to each first end node, the first end node identification data of a first one of the first node edges matches with the first node identification data of the first start node, the first head node identification data of a previous one of the two adjacent first node edges matches with the first end node identification data of a next one of the first node edges, and the first head node identification data of a last one of the first node edges matches with the first node identification data of a current first end node;

the third processing module of the knowledge graph is used for selecting the first optimal node edge connecting line with the least number of the first node edges from the obtained plurality of first optimal node edge connecting lines as the shortest node edge connecting line; on the shortest node edge connecting line, marking the last first node as a second end node, and marking other first nodes between the first starting node and the second end node as first intermediate nodes;

the marketing scheme data processing module is used for performing key intermediate node identification on each first intermediate node according to the first marketing object type data and performing marketing scheme associated data collection according to the identified key intermediate nodes to generate a first marketing scheme associated data set; acquiring the associated data of the second end node according to the first marketing target type data to generate first marketing scheme initial data;

the marketing scheme integration module is used for referring to the first marketing scheme associated data set, and adjusting the first marketing scheme initial data to generate first marketing scheme data.

A third aspect of an embodiment of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;

the processor is configured to be coupled to the memory, read and execute instructions in the memory, so as to implement the method steps of the first aspect;

the transceiver is coupled to the processor, and the processor controls the transceiver to transmit and receive messages.

A fourth aspect of embodiments of the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a computer, cause the computer to perform the method of the first aspect.

The embodiment of the invention provides a method, a device, electronic equipment and a computer readable storage medium for generating a marketing scheme based on a knowledge graph, wherein the knowledge graph suitable for the insurance field is created, different types of nodes are created for entities in the field, such as clients, crowds, insurance risks, marketing activities, domain knowledge points, commodity knowledge points, disease knowledge points and the like in the graph, node edges reflecting association relations are established among the nodes, and the formulation of the marketing scheme of the risks and the marketing scheme of the crowds is guided based on the knowledge graph. The method of the invention completely gets rid of the influence of artificial factors and greatly improves the precision and timeliness of the scheme.

Drawings

Fig. 1 is a schematic diagram of a method for generating a marketing plan based on a knowledge graph according to an embodiment of the present invention;

fig. 2 is a block diagram of an apparatus for generating a marketing plan based on a knowledge graph according to a second embodiment of the present invention;

fig. 3 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Before describing a method for generating a marketing plan based on a knowledge graph in detail, the graph structure and the data structure of a first knowledge graph, which are created by the method according to the embodiment of the present invention and are suitable for the insurance field, are described in advance as follows.

The first knowledge-graph comprises a plurality of first nodes and a plurality of first node edges; each first node corresponds to a first node data group, and each first node edge corresponds to a first node edge data group; the first node data group at least comprises first node identification data, first node name data and first node type data; the first node type data at least comprises a client class, a crowd class, an insurance class, a marketing activity class, a topic knowledge class, a commodity knowledge class and a disease knowledge class; the first node edge data group includes at least first head node identification data, first tail node identification data, and first relationship weight data.

Here, the first knowledge graph is a customized graph created for the insurance field according to the embodiment of the present invention, and is known by common general knowledge of knowledge graphs, and the graph structure of the first knowledge graph is composed of a plurality of first nodes and a plurality of first node edges; when the data structure of the knowledge-graph is implemented, the data object of the first knowledge-graph is actually composed of two data sets: a data set of a first node and a data set of a first node edge; each first node corresponds to a data group object, namely a first node data group, and the data set of the first node is the total set of the first node data group; each node edge corresponds to a data group object, namely a first node edge data group, and the data set of the first node edge is the total set of the first node edge data group;

each first node at least has three attribute values of node identification, node name and node type, so that the first node data group at least comprises first node identification data reflecting the node identification attribute, first node name data reflecting the node name attribute and first node type data reflecting the node type attribute; because the first knowledge graph is a graph customized for the insurance field, entities corresponding to the nodes at least comprise types of customers, crowds, insurance risk types, marketing activities, field knowledge points, commodity knowledge points, disease knowledge points and the like, and the entities corresponding to the entity types in the first node data group are first node type data, so that the values of the entities at least comprise customer types, crowd types, danger types, marketing activities, topic knowledge types, commodity knowledge types and disease knowledge types;

specifically, the first node name data of the node whose first node type data is the client class is the name of the client, the first node name data of the node whose first node type data is the crowd class is the name of the crowd classification (such as high-consumption crowd, low-consumption crowd, high-tech crowd, logistics distribution crowd, high-blood lipid patient crowd, sub-health crowd, etc. according to the consumption level, and various crowd names according to other classification ways), the first node name data of the node whose first node type data is the danger class is the name of the danger class and/or the specific danger number, and the first node name data of the node whose first node type data is the marketing campaign class is the name of the specific marketing campaign class (such as the online promotion campaign, online promotion, offline classification, online promotion, offline promotion, online promotion, and marketing campaign, online promotion, and marketing campaign, etc.) The names of offline promotional activities, etc., road promotional activities, cell promotional activities, etc., differentiated by the nature of the offline points, and various marketing activities, etc., differentiated by other classification means, the first node name data for a node whose first node type data is topic knowledge, is the name of a specific hot topic (such as specific time topic name, etc., differentiated by hot time, specific area topic name, etc., differentiated by hot geographic location, specific event topic name, etc., differentiated by hot event, specific character topic name, etc., differentiated by hot character, and the names of various topics, etc., differentiated by other hot entity categories), the first node name data for a node whose first node type data is commodity knowledge, is the name of a specific hot commodity (such as specific commodity category name, etc., name of specific commodity model distinguished by hot commodity model, and names of various commodities distinguished by other commodity characteristic categories), the first node name data of the node of which the first node type data is the disease knowledge category is the name of specific disease (such as the name of disease counted by customer health report, the name of disease counted by medical and insurance field report, and the name of disease counted by other report);

each first node edge at least comprises three attribute values of a starting node representing the connecting line direction of the current node edge, a pointing node identifier and a weight value representing the degree of association of two nodes connected by the current node edge, so that the first node edge data group at least comprises first tail node identifier data reflecting the starting node identifier of the current node edge, first head node identifier data reflecting the pointing node identifier of the current node edge and first relation weight data reflecting the weight value of the current node edge; for example, if there is a first node A that characterizes customer A, a first node B that characterizes customer B, and a first node C that characterizes disease C in the first knowledge-graph; knowing that the client has the disease C through the latest health report of the client A, a first node edge A-C is formed when the first node A points to the first node C, the tail node identification of the first node edge A-C is the node identification of the first node A, the head node identification is the node identification of the first node C, and the first relation weight is higher; knowing that the customer has had disease C in history but has been substantially cured through the latest health report of the customer B, there will be a first node edge B-C pointing from the first node B to the first node C, the tail node identifier of the first node edge B-C being the node identifier of the first node B, the head node identifier being the node identifier of the first node C, and the first relationship weight being lower.

After the first knowledge graph is created, each time a worker formulates a dangerous marketing scheme for a specific customer or a crowd marketing scheme for a specific marketing campaign, the worker does not need to consume additional time to organize and manually collect and analyze related data, and only needs to input a customer name of the dangerous marketing scheme or a marketing campaign name of the crowd marketing scheme, and the corresponding dangerous or crowd node is automatically obtained from the optimal node path of the first knowledge graph to form a marketing primary scheme by the method provided by the embodiment of the invention, in addition, the method provided by the embodiment of the invention can obtain more reference data by obtaining other key intermediate nodes closely related to the customer or the crowd to adjust the dangerous or the crowd marketing primary scheme so as to obtain a final optimal scheme, and fig. 1 is a schematic diagram of a method for generating a marketing scheme based on the knowledge graph provided by the embodiment of the invention, as shown in fig. 1, the method mainly comprises the following steps:

step 1, obtaining first marketing object type data and first marketing object name data.

Here, if the staff is setting up a dangerous marketing plan for a specific customer, the marketing object is the specific customer, the first marketing object type data is the customer class, and the first marketing object name data corresponds to the name of the specific customer; if the staff sets up the crowd marketing plan for a specific marketing activity, the marketing object is the specific marketing activity, the first marketing object type data is the marketing activity class, and the first marketing object name data corresponds to the name of the specific marketing activity.

Step 2, setting first marketing target type data according to the first marketing object type data; when the first marketing object type data is a client type, setting the first marketing target type data as an insurance type; and when the first marketing object type data is the marketing activity class, setting the first marketing target type data as the crowd class.

Here, if the first marketing object type data is a customer type, which indicates that the staff is making an insurance marketing plan for a specific customer, the marketing target should be the best insurance type for the user, and naturally the first marketing target type data should be the insurance type; if the first marketing object type data is a marketing campaign type, which indicates that the staff is making a crowd marketing plan for a specific marketing campaign, the marketing target should be the crowd most suitable for the marketing campaign, and naturally the first marketing target type data should be a crowd type.

And 3, in a preset first knowledge graph, marking a first node with the first node type data matched with the first marketing object type data and the first node name data matched with the first marketing object name data as a first starting node, and marking all first nodes with the first node type data matched with the first marketing object type data as first ending nodes.

Here, if the first marketing object type data is a customer class, the corresponding first marketing object name data is a specific customer name, the corresponding first marketing target type data is a risk class, the first start node is a first node whose first node type data is a "customer class" and whose first node name data is a specific customer name, and the first end node is a first node whose first node type data is a "risk class", and since the number of nodes of the risk class is not unique, a plurality of first end nodes can be obtained in the current step;

if the first marketing object type data is a marketing campaign type, the corresponding first marketing object name data is a specific marketing campaign name, the corresponding first marketing target type data is a crowd type, the first start node is a first node whose first node type data is a "marketing campaign type" and whose first node name data is a specific marketing campaign name, and the first end node is a first node whose first node type data is a "crowd type", so that a plurality of first end nodes can be obtained in the current step because the number of nodes of the crowd type is not unique.

For example, the name of the customer is "zhang san", in the first knowledge graph, the first node type data of the first node 1 is "customer class", the first node name data is "zhang san", the first node type data of the first node 2 is "risk class", the first node name data is "life insurance 002", the first node type data of the first node 3 is "risk class", the first node name data is "life insurance 003", the first node type data of the first node 4 is "risk class", the first node name data is "disease insurance 004", the obtained first starting node is the first node 1, and the obtained first ending nodes are 3

first nodes

2, 3, and 4, respectively. The purpose of this is to screen the most appropriate insurance scheme for the user from the largest data range;

for another example, the customer name is "online promotional activity", in the first knowledge graph, the first node type data of the first node 1A is "marketing activity class", the first node name data is "online promotional activity", the first node type data of the first node 2A is "crowd class", the first node name data is "high income crowd 1", the first node type data of the first node 3A is "crowd class", the first node name data is "high income crowd 2", the obtained first starting node is the first node 1A, and 2 obtained first ending nodes are the first nodes 2A and 3A, respectively. The purpose of this is to screen the most suitable population for online promotional activities from the largest data range.

Step 4, in the first knowledge graph, identifying the optimal node edge connecting line from the first starting node to each first ending node to generate a corresponding first optimal node edge connecting line;

the first optimal node edge connecting line is formed by sequentially connecting a plurality of first node edges; on a first optimal node edge connecting line corresponding to each first end node, matching first tail node identification data of a first node edge with first node identification data of a first start node, matching first head node identification data of a previous first node edge of two adjacent first node edges with first tail node identification data of a next first node edge, and matching first head node identification data of a last first node edge with first node identification data of a current first end node;

here, in the knowledge graph, all the connecting lines between nodes are formed by connecting one or more first node edges, which are also called as node edge connecting lines, and the data structure of the node edge connecting lines is actually a sequence formed by arranging one or more first node edges according to the connecting sequence;

here, as the network complexity of the first knowledge graph increases, for each first end node, multiple paths may exist from the first start node to the current first end node, and it is known from the inference theory of the knowledge graph that the longer the path between the nodes, that is, the more the number of node edges connected in sequence from head to tail, the weaker the association relationship between the nodes and the worse the matching degree, so for each first end node, the optimal path from the first start node to the current first end node needs to be identified, because the so-called path is a connection line formed by connecting multiple first node edges from head to tail, and the identification process of the optimal path is the identification process of the connection line of the optimal node edges;

the method specifically comprises the following steps: step 41, taking each first end point as a first current end point; counting all node edge connecting lines from a first starting node to a first current ending point in the first knowledge graph to obtain a plurality of first node edge connecting lines;

the first node edge connecting line is formed by sequentially connecting a plurality of first node edges;

here, it is the first step of the identification process of the edge connecting line of the optimal node, that is, all paths from the first start node to the first current end point are obtained, and one first node edge connecting line corresponds to one path;

for example, in the first knowledge-graph, the first starting node is a

first node

1, and 2 first ending nodes are

first nodes

2 and 3, respectively;

the path from the first node 1 to the first node 2 has 3: first node 1-first node 11-first node 12-first node 2; first node 1-first node 21-first node 2; first node 1-first node 31-first node 32-first node 33-first node 2;

the path from the first node 1 to the first node 3 has 2: first node 1-first node 101-first node 102-first node 3; first node 1-first node 111-first node 112-first node 113-first node 3;

then, for the first node 2, there are 3 first node edge connecting lines, the 1 st first node edge connecting line (first node edges 1-11, first node edges 11-12, first node edges 12-2), the 2 nd first node edge connecting line (first node edges 1-21, first node edges 21-2), the 3 rd first node edge connecting line (first node edges 1-31, first node edges 31-32, first node edges 32-33, first node edges 33-2);

for the first node 3, there are 2 first node edge connecting lines, the 1 st first node edge connecting line (first node edge 1-101, first node edge 101-102, first node edge 102-3), the 2 nd first node edge connecting line (first node edge 1-111, first node edge 111-112, first node edge 112-113, first node edge 113-3);

and 42, selecting the first node edge connecting line containing the minimum number of the first node edges from the obtained plurality of first node edge connecting lines as a first optimal node edge connecting line corresponding to the first current end point.

Here, the second step of the process of identifying the optimal node edge connection line is to select a shortest path from all the obtained paths from the first starting node to the first current ending point as the first optimal node edge connection line, where the shortest path is the connection line with the smallest number of first node edges included in the first node edge connection line;

for example, for the first node 2, there are 3 first node edge connecting lines, the 1 st first node edge connecting line (first node edges 1-11, first node edges 11-12, first node edges 12-2), the 2 nd first node edge connecting line (first node edges 1-21, first node edges 21-2), the 3 rd first node edge connecting line (first node edges 1-31, first node edges 31-32, first node edges 32-33, first node edges 33-2); in the 3 first node edge connecting lines, the 2 nd first node edge connecting line contains the least number of first node edges, and then the first optimal node edge connecting line corresponding to the first node 2 should be (first node edges 1-21, first node edges 21-2);

for the first node 3, there are 2 first node edge connecting lines, the 1 st first node edge connecting line (first node edge 1-101, first node edge 101-102, first node edge 102-3), the 2 nd first node edge connecting line (first node edge 1-111, first node edge 111-112, first node edge 112-113, first node edge 113-3); in the 2 first node edge connecting lines, the 1 st first node edge connecting line with the least number of first node edges is included, and then the first optimal node edge connecting line corresponding to the first node 3 should be (first node edges 1-101, first node edges 101-102, first node edges 102-3).

In summary, if the first marketing object type data is the customer type, the optimal paths from the customer to a plurality of specific insurance products can be obtained in step 4; if the first marketing object type data is a marketing campaign type, the optimal path from the marketing campaign to a plurality of specific crowd categories can be obtained in step 4.

And 5, selecting the first optimal node edge connecting line with the least number of first node edges from the obtained plurality of first optimal node edge connecting lines as the shortest node edge connecting line.

Here, on the principle that the shortest path is the most matched path, the shortest path, which is the shortest node edge connection, is continuously selected from the plurality of paths obtained in step 4 as the most matched path.

For example, in the first knowledge graph, a first starting node is a

first node

1, 2 first ending nodes are respectively

first nodes

2 and 3, a first optimal node edge connecting line corresponding to the first node 2 obtained through the step 4 is (first node edge 1-21, first node edge 21-2), and a first optimal node edge connecting line corresponding to the first node 3 is (first node edge 1-101, first node edge 101-102, first node edge 102-3); then the shortest node edge connection should be the first optimal node edge connection corresponding to first node 2 (first node edges 1-21, first node edges 21-2).

And 6, marking the last first node as a second end node and marking other first nodes between the first starting node and the second end node as first middle nodes on the shortest node edge connecting line.

Here, if the first marketing object type data is a client type, the second end node is actually an insurance product node with the shortest path found for the current user in the embodiment of the present invention, in other words, an insurance product node with the strongest association relationship and the highest matching degree with the current user, and each first intermediate node is actually an association node having a direct or indirect relationship with both the current user and the current insurance product node; if the first marketing object type data is a marketing activity type, the second end node is actually the crowd category node with the shortest path found for the current marketing activity in the embodiment of the present invention, in other words, the crowd category node with the strongest association relationship and the highest matching degree with the current marketing activity, and each first intermediate node is actually an association node having a direct or indirect relationship with both the current marketing activity and the current crowd category node.

For example, in the first knowledge-graph, the first start node is the

first node

1, 2 first end nodes are the

first nodes

2 and 3, respectively, and the shortest node edge connecting line obtained in step 5 is (the first node edges 1 to 21, the first node edges 21 to 2), so the second end node is the first node 2, and the first intermediate node is the first node 21.

Step 7, performing key intermediate node identification on each first intermediate node according to the first marketing object type data, and performing marketing scheme associated data collection according to the identified key intermediate nodes to generate a first marketing scheme associated data set;

here, if all the intermediate nodes between the first start node and the second end node are regarded as reference nodes of the marketing scheme, the data volume is too large and not accurate enough, so that in the step, whether each first intermediate node is a key intermediate node needs to be identified, and only information related to the key intermediate node is extracted to serve as associated data of the marketing scheme;

the method specifically comprises the following steps: step 71, performing corresponding correlation weight value calculation on each first intermediate node according to the first marketing object type data to generate a corresponding first correlation weight value;

here, when the first marketing object type data is a customer type, the embodiment of the present invention takes the association relationship between each first intermediate node and the first start node as a consideration credential of the key intermediate node, and the calculated first association weight value is also a weight accumulation in a direction from the first start node to the current first intermediate node; when the first marketing object type data is a marketing activity class, the embodiment of the invention takes the incidence relation between each first intermediate node and the second end node as a consideration evidence of the key intermediate node, and the calculated first incidence weight value is also subjected to weight multiplication from the current first intermediate node to the second end node;

the method specifically comprises the following steps: step 711, when the first marketing object type data is a customer type, taking a product of continuous multiplication of first relationship weight data of all first node edges between the first starting node and the current first intermediate node as a first associated weight value corresponding to the current first intermediate node;

for example, in the first knowledge graph, the first starting node is the first node 1001, the second ending node is the first node 1002, 3 first intermediate nodes from the first starting node to the second ending node are the first node 1200, the first node 1201 and the first node 1202 in sequence, the corresponding first relationship weight data of the first node edge 1001-;

then, when the first marketing object type data is a customer class,

the first relationship weight data of the 1 st first intermediate node, i.e., the first node 1200 to the first starting node, i.e., the first node 1001, is 0.9;

the 2 nd first intermediate node, that is, the first node 1201 sets the first relationship weight data of the first node edge 1200 ═ 0.9 ═ 0.8 ═ 0.72 to the first relationship weight data of the first node edge 1200-;

the 3 rd first intermediate node, that is, the first node 1202, has the first correlation weight value of the first initial node, that is, the first node 1001 ═ the first relationship weight data of the first node edge 1001-;

step 712, when the first marketing object type data is the marketing activity type, taking the product of the continuous multiplication of the first relationship weight data of all the first node edges between the current first intermediate node and the second end node as a first associated weight value corresponding to the current first intermediate node;

for example, in the first knowledge graph, the first start node is the first node 1101, the second end node is the first node 1102, 3 first intermediate nodes from the first start node to the second end node are the first node 1300, the first node 1301 and the first node 1302 in sequence, the corresponding first relationship weight data of the first node edge 1101-;

then, when the first marketing object type data is of the marketing campaign type,

the first relation weight data of the first node edge 1302 ═ first relation weight data of the first node edge 1300-;

the 2 nd first intermediate node, that is, the first node 1301, has the first correlation weight value of the second end node, that is, the first node 1102, which is the first relationship weight data of the first node edge 1301-;

the first relationship weight data of the 3 rd first intermediate node, i.e., the first node 1302 to the second end node, i.e., the first node 1102, is 0.7;

step 72, recording a first intermediate node corresponding to a first association weight value exceeding a preset association weight threshold as a first key intermediate node;

here, the association weight threshold is a preset weight threshold for identifying the key intermediate node, and whether the corresponding first intermediate node is the key intermediate node is determined according to a comparison result between the threshold and the plurality of first association weight values obtained in step 71;

for example, if the association weight threshold is preset to 0.6, then 3 first intermediate nodes in step 711, only 1 st and 2 nd nodes are qualified; in step 712, only the 3 rd first intermediate node is qualified;

step 73, collecting the associated data of all the first key intermediate nodes to generate a first marketing scheme associated data set;

here, after obtaining the key intermediate node, the embodiment of the present invention obtains detailed content data from the relevant preset knowledge base to form a reference data set of the marketing plan, that is, a first marketing plan associated data set;

the method specifically comprises the following steps: step 731, identifying the first node type data of each first key intermediate node; if the first node type data of the current first key intermediate node is a topic knowledge class, extracting a first topic content field of a first topic knowledge record corresponding to the current first key intermediate node from a preset first topic knowledge base to serve as corresponding first associated text data; if the first node type data of the current first key intermediate node is a commodity knowledge class, extracting a first commodity content field of a first commodity knowledge record corresponding to the current first key intermediate node from a preset first commodity knowledge base to serve as corresponding first associated text data; if the first node type data of the current first key intermediate node is a disease knowledge class, extracting a first disease content field of a first disease knowledge record corresponding to the current first key intermediate node from a preset first disease knowledge base to serve as corresponding first associated text data;

here, each knowledge base (the first topic knowledge base, the first commodity knowledge base, and the first disease knowledge base) in the above text content database may be an independent text content database or a unified text content database; the knowledge base is composed of a plurality of knowledge records (a first topic knowledge record, a first commodity knowledge record and a first disease knowledge record), each knowledge record corresponds to a first node in the first knowledge graph, and each knowledge record is provided with a content field (a first topic content field, a first commodity content field and a first disease content field) for storing associated text information corresponding to the first node; for example, the first node type data of the first key intermediate node is a topic knowledge class, the first node name data is a title of a specific topic, and then the first topic content field of the corresponding first topic knowledge record in the first topic knowledge base stores the detailed content information of the specific topic; the first node type data of the first key intermediate node is a commodity knowledge class, the first node name data is the name of a specific commodity, and the detailed introduction information of the specific commodity is stored in the first topic content field of the corresponding first commodity knowledge record in the first commodity knowledge base; the first node type data of the first key intermediate node is a disease knowledge class, the first node name data is a name of a specific disease, and the first topic content field of the corresponding first disease knowledge record in the first disease knowledge base stores detailed introduction information of the specific disease;

step 732, composing corresponding first marketing scheme associated data by the first node name data, the first node type data and the first associated text data of each first key intermediate node;

in step 733, a first marketing plan associated data set is formed by the obtained plurality of first marketing plan associated data.

Here, the first marketing plan associated data set is associated data of a marketing plan, that is, reference data for specifying the marketing plan; if the first marketing object type data is of a client type, hot topic information concerned by the current client, commodity information related to the current client and disease information related to the current client can be seen through the first marketing scheme related data set; if the first marketing object type data is a marketing activity type, hot topic information concerned by the positioning crowd, commodity information related to the positioning crowd and disease information related to the positioning crowd can be seen through the first marketing scheme associated data set.

Step 8, acquiring the associated data of the second end node according to the first marketing target type data to generate first marketing scheme initial data;

the method specifically comprises the following steps: step 81, when the first marketing target type data is of a dangerous type, extracting a first dangerous type clause content field of a first dangerous type clause knowledge record corresponding to the second end node in a preset first dangerous type clause knowledge base to serve as corresponding second associated text data;

here, the first risk type clause knowledge base in the above may be an independent text content database, or may be combined with the first topic knowledge base, the first commodity knowledge base, and the first disease knowledge base in the above to form a unified text content database; the first risk type clause knowledge base consists of a plurality of first risk type clause knowledge records, each first risk type clause knowledge record corresponds to a first node in the first knowledge graph, and each first risk type clause knowledge record has a first risk type clause content field for storing associated text information corresponding to the first node;

when the first marketing target type data is a dangerous type, it is stated that the second end node is actually an insurance product node with the strongest association relationship and the highest matching degree for the current user in the embodiment of the present invention, and the content of the first dangerous type term content field of the first dangerous type term knowledge record corresponding to the first dangerous type term knowledge base by the second end node is also the standard term text of the insurance product, that is, the second associated text data is actually the standard term text corresponding to the second end node;

step 82, when the first marketing target type data is of a crowd type, extracting a first crowd characteristic content field of a first crowd characteristic knowledge record corresponding to the second end node from a preset first crowd characteristic knowledge base to serve as corresponding second associated text data;

here, the first crowd characteristic knowledge base may be an independent text content database, or may be combined with the first risk type term knowledge base, the first topic knowledge base, the first commodity knowledge base, and the first disease knowledge base to form a unified text content database; the first personal group feature knowledge base is composed of a plurality of first personal group feature knowledge records, each first personal group feature knowledge record corresponds to a first node in the first knowledge graph, and each first personal group feature knowledge record has a first personal group feature content field for storing associated text information corresponding to the first node;

when the first marketing target type data is a crowd class, it is indicated that the second end node is actually a crowd class node with the strongest association relationship and the highest matching degree found for the current marketing activity in the embodiment of the present invention, and the content of the first crowd characteristic content field of the first crowd characteristic knowledge record corresponding to the first crowd characteristic knowledge base of the second end node is also the crowd characteristic information of the crowd class, that is, the second associated text data is actually the crowd class characteristic information corresponding to the second end node;

and 83, forming first marketing scheme initial data by the first node name data and the second associated text data of the second end node.

Here, since the second end node is the final marketing target node, the present embodiment takes the data associated with the second end node as the main data of the marketing scheme formulated this time, that is, the initial data of the first marketing scheme;

when the first marketing target type data is a dangerous type, the first node name data of the second end node is actually the name data of the best matched insurance product, and the first marketing scheme initial data is actually the insurance product name + insurance product standard clause text;

when the first marketing target type data is the crowd class, the first node name data of the second end node is the name data which is most matched with the crowd class, and the first marketing scheme initial data is the crowd class name plus the crowd class characteristic information.

And 9, referring to the first marketing scheme associated data set, and adjusting the first marketing scheme initial data to generate first marketing scheme data.

Here, as can be seen from the foregoing, if the first marketing object type data is a client type, the first marketing target type data is a dangerous type, in this case, the content of the first marketing plan initial data in this step is insurance product name + insurance product standard clause text selected for the current user, on this basis, the insurance product standard clause text in the first marketing plan initial data can be adjusted in a targeted manner through the hot topic information of current client interest, the commodity information of current client association, and the disease information of current client association, which are provided by the first marketing plan association data set, so that the generated first marketing plan data is closer to the actual demand of the current user;

in addition, if the first marketing object type data is a marketing activity type, the first marketing target type data is a crowd type, and in this case, the content of the first marketing scheme initial data in the step is the name of the crowd type located for the current marketing activity plus the crowd type characteristic information, on the basis, the crowd type characteristic information in the first marketing scheme initial data can be specifically refined through the hot topic information of the located crowd concern, the commodity information of the located crowd relevance and the disease information of the located crowd relevance, which are provided by the first marketing scheme relevance data set, so that the generated first marketing scheme data can be closer to the actual demand of the current crowd.

Fig. 2 is a block diagram of a device for generating a marketing plan based on a knowledge graph according to a second embodiment of the present invention, where the device may be a terminal device or a server for implementing the method according to the second embodiment of the present invention, or may be a device connected to the terminal device or the server for implementing the method according to the second embodiment of the present invention, and for example, the device may be a device or a chip system of the terminal device or the server. As shown in fig. 2, the apparatus includes: the marketing scheme data processing system comprises an acquisition module 201, a knowledge graph first processing module 202, a knowledge graph second processing module 203, a knowledge graph third processing module 204, a marketing scheme data processing module 205 and a marketing scheme integration module 206.

The obtaining module 201 is configured to obtain first marketing object type data and first marketing object name data.

The knowledge graph first processing module 202 is used for setting first marketing target type data according to the first marketing object type data; when the first marketing object type data is a client type, setting the first marketing target type data as an insurance type; when the first marketing object type data is a marketing activity class, setting the first marketing target type data as a crowd class; in a preset first knowledge graph, marking a first node with the first node type data matched with the first marketing object type data and the first node name data matched with the first marketing object name data as a first starting node, and marking all first nodes with the first node type data matched with the first marketing object type data as first ending nodes; wherein the first knowledge-graph comprises a plurality of first nodes; each first node corresponds to a first node data group; the first node data group at least comprises first node identification data, first node name data and first node type data; the first node type data includes at least a customer class, a crowd class, a risk class, and a marketing campaign class.

The knowledge graph second processing module 203 is configured to identify an optimal node edge connection line from the first start node to each first end node in the first knowledge graph, and generate a corresponding first optimal node edge connection line; wherein the first knowledge-graph further comprises a plurality of first node edges; each first node edge corresponds to a first node edge data group; the first node edge data group at least comprises first head node identification data and first tail node identification data; the first optimal node edge connecting line is formed by sequentially connecting a plurality of first node edges; on a first optimal node edge connecting line corresponding to each first end node, first tail node identification data of a first node edge is matched with first node identification data of a first starting node, first head node identification data of a previous first node edge of two adjacent first node edges is matched with first tail node identification data of a next first node edge, and first head node identification data of a last first node edge is matched with first node identification data of a current first end node.

The knowledge graph third processing module 204 is configured to select, from the obtained plurality of first optimal node edge connection lines, a first optimal node edge connection line that includes the smallest number of first node edges as a shortest node edge connection line; and on the shortest node edge connecting line, marking the last first node as a second end node, and marking other first nodes between the first starting node and the second end node as first intermediate nodes.

The marketing scheme data processing module 205 is configured to perform key intermediate node identification on each first intermediate node according to the first marketing object type data, and perform marketing scheme associated data collection according to the identified key intermediate node to generate a first marketing scheme associated data set; and acquiring the associated data of the second end node according to the first marketing target type data to generate first marketing scheme initial data.

The marketing plan integration module 206 is configured to adjust the first marketing plan initial data to generate first marketing plan data by referring to the first marketing plan associated data set.

The device for generating the marketing scheme based on the knowledge graph provided by the embodiment of the invention can execute the method steps in the method embodiment, and the implementation principle and the technical effect are similar, so that the detailed description is omitted.

It should be noted that the division of the modules of the above apparatus is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity, or may be physically separated. And these modules can be realized in the form of software called by processing element; or may be implemented entirely in hardware; and part of the modules can be realized in the form of calling software by the processing element, and part of the modules can be realized in the form of hardware. For example, the obtaining module may be a processing element separately set up, or may be implemented by being integrated in a chip of the apparatus, or may be stored in a memory of the apparatus in the form of program code, and a processing element of the apparatus calls and executes the functions of the determining module. Other modules are implemented similarly. In addition, all or part of the modules can be integrated together or can be independently realized. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.

For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more Digital Signal Processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), etc. For another example, when some of the above modules are implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor that can invoke the program code. As another example, these modules may be integrated together and implemented in the form of a System-on-a-chip (SOC).

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, bluetooth, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), etc.

Fig. 3 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention. The electronic device may be the terminal device or the server, or may be a terminal device or a server connected to the terminal device or the server and implementing the method according to the embodiment of the present invention. As shown in fig. 3, the electronic device may include: a processor 301 (e.g., a CPU), a memory 302, a transceiver 303; the transceiver 303 is coupled to the processor 301, and the processor 301 controls the transceiving operation of the transceiver 303. Various instructions may be stored in memory 302 for performing various processing functions and implementing the methods and processes provided in the above-described embodiments of the present invention. Preferably, the electronic device according to an embodiment of the present invention further includes: a power supply 304, a system bus 305, and a communication port 306. The system bus 305 is used to implement communication connections between the elements. The communication port 306 is used for connection communication between the electronic device and other peripherals.

The system bus mentioned in fig. 3 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus or the like. The system bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown, but this does not mean that there is only one bus or one type of bus. The communication interface is used for realizing communication between the database access device and other equipment (such as a client, a read-write library and a read-only library). The Memory may include a Random Access Memory (RAM) and may also include a Non-Volatile Memory (Non-Volatile Memory), such as at least one disk Memory.

The Processor may be a general-purpose Processor, including a central processing unit CPU, a Network Processor (NP), and the like; but also a digital signal processor DSP, an application specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic device, discrete gate or transistor logic, discrete hardware components.

It should be noted that the embodiment of the present invention also provides a computer-readable storage medium, which stores instructions that, when executed on a computer, cause the computer to execute the method and the processing procedure provided in the above-mentioned embodiment.

The embodiment of the invention also provides a chip for running the instructions, and the chip is used for executing the method and the processing process provided by the embodiment.

Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A method for generating a marketing plan based on a knowledge graph, the method comprising:

2. The method for generating a marketing solution based on a knowledge-graph of claim 1, wherein the identifying, in the first knowledge-graph, an optimal node-edge connecting line from the first start node to each of the first end nodes to generate a corresponding first optimal node-edge connecting line specifically comprises:

3. The method for generating a marketing plan based on a knowledge graph of claim 1, wherein the identifying key intermediate nodes of each first intermediate node according to the first marketing object type data, and collecting marketing plan association data according to the identified key intermediate nodes to generate a first marketing plan association data set specifically comprises:

4. The method for generating a marketing plan based on a knowledge graph of claim 3, wherein the first node edge data set further includes first relationship weight data, and the calculating of the corresponding association weight value for each of the first intermediate nodes according to the first marketing object type data to generate the corresponding first association weight value specifically includes:

5. The method for generating a marketing plan based on a knowledge graph of claim 3, wherein the first node type data further comprises a topic knowledge class, a commodity knowledge class and a disease knowledge class, and the collecting the associated data of all the first key intermediate nodes generates the first marketing plan associated data set, which specifically comprises:

6. The method for generating a marketing plan based on a knowledge graph of claim 1, wherein the obtaining the associated data of the second end node according to the first marketing target type data to generate first marketing plan initial data specifically comprises:

7. An apparatus for implementing the method of generating a marketing plan based on a knowledge-graph of any one of claims 1-6, the apparatus comprising: the system comprises an acquisition module, a first knowledge graph processing module, a second knowledge graph processing module, a third knowledge graph processing module, a marketing scheme data processing module and a marketing scheme integration module;

8. An electronic device, comprising: a memory, a processor, and a transceiver;

the processor is used for being coupled with the memory, reading and executing the instructions in the memory to realize the method steps of any one of claims 1-6;

9. A computer-readable storage medium having stored thereon computer instructions which, when executed by a computer, cause the computer to perform the method of any of claims 1-6.