CN109241198B

CN109241198B - Competitor mining method and device, electronic equipment and storage medium

Info

Publication number: CN109241198B
Application number: CN201810797322.1A
Authority: CN
Inventors: 黄剑飞; 罗恒亮; 梁世朴
Original assignee: Beijing Sankuai Online Technology Co Ltd
Current assignee: Beijing Sankuai Online Technology Co Ltd
Priority date: 2018-07-19
Filing date: 2018-07-19
Publication date: 2021-04-23
Anticipated expiration: 2038-07-19
Also published as: CN109241198A

Abstract

The application discloses a competitor mining method, belongs to the technical field of computers, and solves the problem that a competitor mined in the prior art is inaccurate. The competitor mining method disclosed by the application comprises the following steps: according to a pre-constructed competition relationship network, candidate competition merchants mined by a target merchant based on the competition relationship network and corresponding first competition strength indexes are determined; determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index; and determining the competitive merchant of the target merchant by comparing the candidate competitive merchant to be compared with the competitive seed merchant of the target merchant. The competitor mining method disclosed by the application improves the accuracy of mining to the competitor.

Description

Competitor mining method and device, electronic equipment and storage medium

Technical Field

The present application relates to the field of computer technologies, and in particular, to a competitor mining method, a competitor mining device, an electronic device, and a storage medium.

Background

And competitors are found in time, so that the merchant can adjust the marketing strategy timely and comprehensively, and customers can be better served. In the prior art, a method for mining a competitor of a merchant generally includes first collecting data such as a geographical location of the merchant, a location of the merchant (such as star level, type, price, etc.), category information of a product hung under the merchant, and the like, then automatically classifying and automatically rearranging the collected data through a classification technique (such as a classification template based on the category automatic classification technique), and finally calculating similarity between text contents according to text content features in the collected data, thereby determining a ranking of associated merchants, and evaluating a ranking effect by means of expert experience to determine a competitor. The applicant finds that the competitors obtained by mining by the method in the prior art are evaluated based on expert experience, and the evaluation method is not objective enough and has the defect of inaccurate mining result; meanwhile, no objective accuracy measurement method exists, and the excavation accuracy is not favorably and continuously improved.

Disclosure of Invention

The application provides a competitor mining method, which at least solves the problem that a mined competitor is inaccurate in the prior art.

In a first aspect, an embodiment of the present application provides a competitor mining method, including:

determining candidate competitive merchants mined by a target merchant based on a competitive relationship network according to a pre-constructed competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on a corresponding competitive relationship network;

determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index;

and determining the competitive merchant of the target merchant by comparing the candidate competitive merchant to be compared with the competitive seed merchant of the target merchant.

In a second aspect, an embodiment of the present application provides a competitor digging device, including:

the candidate competitive merchant information determining module is used for determining candidate competitive merchants mined by a target merchant based on a competitive relationship network according to a pre-established competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on the corresponding competitive relationship network;

the candidate competitive merchant determining module is used for determining candidate competitive merchants to be compared from the candidate competitive merchants mined on the basis of the competitive relationship network according to the first competitive strength index determined by the candidate competitive merchant information determining module;

and the competitive merchant list mining module is used for comparing the candidate competitive merchants to be compared with the competitive seed merchants of the target merchants to determine the competitive merchants of the target merchants.

In a third aspect, an embodiment of the present application further discloses an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the computer program, the competitor mining method according to the embodiment of the present application is implemented.

In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, performs the steps of the competitor mining method disclosed in the embodiments of the present application.

The method for mining the competitors, disclosed by the embodiment of the application, comprises the steps of determining candidate competitive merchants, mined based on a competitive relationship network, of a target merchant according to the pre-established competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on the corresponding competitive relationship network; determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index; the candidate competitive merchants to be compared are compared with the competitive seed merchants of the target merchants to determine the competitive merchants of the target merchants, so that the problem that competitors mined in the prior art are inaccurate is solved. According to the competitor mining method disclosed by the embodiment of the application, candidate competitive merchants are mined through a competitive relationship network constructed based on data, and the mined candidate competitive merchants are further sorted by combining with seed competitive merchants, so that the accuracy of mining to competitors is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.

FIG. 1 is a flowchart of a competitor mining method according to a first embodiment of the present application;

FIG. 2 is a flowchart of a competitor mining method according to the second embodiment of the present application;

fig. 3 is a schematic structural diagram of a competitor digging device according to a third embodiment of the present application;

fig. 4 is a second schematic structural view of a competitor digging device according to a third embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. 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 application.

In the embodiment of the present application, the competition strength indexes such as the first competition strength index, the second competition strength index, and the third competition strength index are used for strength or magnitude of the competition relationship between the constant merchants. For example, the larger the value of the first competition strength index is, the stronger or larger the competition relationship between the merchants is; the smaller the value of the first competition strength index is, the weaker or smaller the competition relationship between the merchants is.

Example one

As shown in fig. 1, the method for mining a competitor includes: step 110 to step 130.

Step 110, according to a pre-constructed competition relationship network, determining candidate competition merchants mined by a target merchant based on the competition relationship network, and a first competition strength index between the candidate competition merchants and the target merchant based on a corresponding competition relationship network.

In specific implementation, firstly, determining candidate competitive merchant information of a target merchant according to a pre-constructed competitive relationship network, wherein the candidate competitive merchant information comprises: and the candidate competitive merchants are mined based on the competitive relationship network, and the candidate competitive merchants are based on the first competitive strength index of the corresponding competitive relationship network and the target merchant.

The competition relationship network is used for recording competition relationships among the merchants and first competition strength indexes, and in some specific embodiments of the application, the competition relationship network can be constructed by constructing directed graphs among the merchants. In the specific implementation of the application, the merchants with the competitive relationship can be mined based on any one or combination of multiple data of static data of the merchants, on-line dynamic data of the merchants, off-line dynamic data of the merchants, user on-line dynamic data of the merchants and user off-line dynamic data of the merchants, and a first competitive strength index between the merchants with the competitive relationship is determined. The merchant having the competitive relationship may be determined according to specific service requirements, for example, the method may include: a plurality of merchants operated by the user in sequence in one action, a plurality of merchants with similar geographic positions, a plurality of merchants with similar business categories and the like. The first competition strength index is used for measuring the competition strength between the merchants and can be determined according to the behavior data of the merchants or the user behavior data of the merchants.

And then, constructing a directed edge by taking the merchants as nodes according to the determined competitive relationship among the merchants, and establishing a directed graph. The directed edge points to the target node from the source node, and the user indicates that a merchant corresponding to the target node has a competitive relationship with a merchant corresponding to the source node. Further, a first competition strength index between merchants corresponding to the target node and the source node is used as the attribute information of the directed edge. Thus, the construction of a competition relation network is completed.

When the method is implemented specifically, the merchants with competition relationships and the first competition strength indexes among the merchants can be determined from multiple dimensions, and corresponding competition relationship networks can be respectively constructed according to the competition relationships determined by each dimension and the first competition strength indexes, or only one competition relationship network can be constructed.

In a specific application process, nodes corresponding to target merchants can be respectively searched in each pre-constructed competition relationship network, merchants corresponding to the target nodes pointed by directional edges of the target merchants as source nodes are determined as candidate competition merchants of the target merchants in the current competition relationship network, and attribute information of the directional edges is used as a first competition strength index of the corresponding candidate competition merchants and the target merchants.

And 120, determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index.

In specific implementation, a group of candidate competitive merchants of the target merchant can be determined through one network, a plurality of groups of candidate competitive merchants of the target merchant can be determined through different networks, and the first competition strength index of each candidate competitive merchant and the target merchant in each group of candidate competitive merchants. The first competition strength index is a first competition strength index of the target merchant and a candidate competition merchant determined according to a competition relationship network. In specific implementation, if a competition relationship network is pre-constructed, a group of candidate competitive merchants determined by the target merchant based on the competition relationship network is ranked according to the first competition strength index, and then a preset number of candidate competitive merchants with the highest first competition strength index in the candidate competitive merchants are selected as candidate competitive merchants to be compared. If a plurality of competition networks are pre-constructed, a group of candidate competition merchants determined by the target merchant based on each competition network can be respectively sorted according to the first competition strength index, and then a preset number of candidate competition merchants with the highest first competition strength index in each group of candidate competition merchants are selected as candidate competition merchants to be compared corresponding to the corresponding competition networks. The preset number is determined according to business needs, and may be, for example, the number of competitive seed merchants of the target merchant.

Step 130, comparing the candidate competitive merchants to be compared with the competitive seed merchants of the target merchant, and determining the competitive merchants of the target merchant.

If a competition relationship network is pre-constructed, whether the candidate competition merchants to be compared can be used as the competition merchants of the target merchant or not can be determined by calculating the coverage rate of the candidate competition merchants to be compared to the competition seed merchants of the target merchant.

If a plurality of competition relationship networks are constructed in advance, in the concrete implementation of the application, the competition seed merchants of the target merchants need to be specified in advance, and a third competition strength index of each competition seed merchant and the target merchant is determined. And then, performing weighted fusion on the first competition strength indexes of the candidate competitive merchants to be compared, which are determined according to each competition relationship network, based on different competition relationship networks, and determining a second competition strength index of each candidate competitive merchant to be compared. And then, calculating the difference index of the candidate competitive merchants to be compared and the competitive seed merchants of the target merchant based on the second competition strength index, namely the competition strength relation obtained by synthesizing the first competition strength indexes of the multiple competition relation networks, so as to determine whether the candidate competitive merchants to be compared are accurately ranked and mined based on the current second competition strength index. For example, a root mean square error between the second competition strength index of the candidate competition merchant to be compared and the third competition strength index of the competition seed merchant is calculated as the difference index. Or determining whether the candidate competitive merchants to be compared are accurately ranked and mined based on the current second competition strength index based on the difference index of the list of the candidate competitive merchants to be compared and the list of the competitive seed merchants. For example, the accuracy or recall of the competitive seed merchants in the candidate competitive merchant list to be compared is calculated as the difference index.

And when the difference index meets a preset difference index condition, sequencing and mining the candidate competitive merchants to be compared based on the current second competition strength index. When the difference index does not meet the preset difference index condition, iteratively executing a step of determining a second competition strength index of each candidate competition merchant to be compared by performing weighted fusion on the first competition strength indexes of the candidate competition merchants to be compared, which are determined according to each competition relationship network, based on different competition relationship networks and by adjusting the weight of the first competition strength indexes corresponding to each competition relationship network until the difference index meets the preset difference index condition. And determining the preset difference index condition according to specific service requirements.

The method for mining the competitors, disclosed by the embodiment of the application, comprises the steps of determining candidate competitive merchants mined by a target merchant based on a competitive relationship network according to the pre-established competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on the corresponding competitive relationship network; determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index; the candidate competitive merchants to be compared are compared with the competitive seed merchants of the target merchants to determine the competitive merchants of the target merchants, so that the problem that competitors mined in the prior art are inaccurate is solved. According to the competitor mining method disclosed by the embodiment of the application, candidate competitive merchants are mined through a competitive relationship network constructed based on data, and the determined candidate competitive merchants are further sorted by combining with seed competitive merchants, so that the accuracy of mining to competitors is improved.

Example two

As shown in fig. 2, the method for mining a competitor includes: step 210 to step 250.

And step 210, constructing a competition relationship network according to the obtained merchant data.

In some embodiments of the present application, the competitive relationship network is constructed from the obtained merchant data. For example, a competition network may be constructed based on a certain dimension according to merchant data, or a competition network corresponding to at least two dimensions in a one-to-one manner may be constructed from at least two dimensions according to acquired merchant data.

In some embodiments of the present application, before determining, according to a pre-constructed competition relationship network, a candidate competitive merchant mined by a target merchant based on the competition relationship network, and before the candidate competitive merchant is based on a first competition strength index between the corresponding competition relationship network and the target merchant, the method further includes: constructing a competition relationship network according to the obtained merchant data; wherein the merchant data comprises dynamic data comprising at least one of: the online dynamic data of the merchant, the offline dynamic data of the merchant, the online dynamic data of the user of the merchant and the offline dynamic data of the user of the merchant. In some embodiments of the present application, the merchant data further comprises static data, the static data further comprising: intrinsic attribute information of the merchant. The static data refers to relatively stable data, such as inherent attribute information of the merchant. Taking a merchant as an example of a hotel, the inherent attribute information includes: the merchant's geographic location, star rating, time of opening, date of fitment, and the like. The dynamic data refers to data which changes frequently, and a competition relationship network is constructed by combining the dynamic data, so that the timeliness of the data is effectively guaranteed, and the accuracy of the constructed competition relationship network can be further improved. In specific implementation, the dynamic data includes online dynamic data of the merchant, which may include recent sales, recent price, and the like; the dynamic data may include offline dynamic data for the merchant that may include: merchant offers, etc.; the online dynamic data of the user of the merchant may include: browsing data of the user, ordering data of the user, comment data generated by the user and the like; the dynamic data of the user of the merchant under the user's line can comprise the actual data of the user to the store and the like. By combining the merchant data and the user data to construct a competition relationship network, the data is more comprehensive, and the accuracy of a mined competitor is improved. A competition relationship network is established by combining the online and offline dynamic behavior data of the user, so that real user data driving is realized, and the practicability and accuracy of competitor mining are improved.

After the merchant data is obtained, a competition relationship network is further constructed according to the obtained merchant data.

The competition relationship network is used for recording competition relationships among the merchants and first competition strength indexes, and in some specific embodiments of the application, the competition relationship network can be constructed by constructing directed graphs among the merchants. The contention relationship network comprises: the node comprises a source node and a target node, the directed edge is connected with the nodes corresponding to the merchants with competition, and the attribute information of the directed edge comprises first competition strength indexes of the target node of the directed edge and the merchants corresponding to the source node. Wherein the merchant having the competitive relationship is determined by any one or more of the following methods: determining the merchants related to the primary behavior of the user as merchants with competitive relationship; determining merchants related to the same behavior of the user as merchants with competitive relationship; determining merchants operated by user equipment of a specified user as merchants with competitive relationship; determining the commercial tenant of which the attribute information meets the preset correlation condition as the commercial tenant with a competitive relationship; and determining the related merchants of the preset merchant behaviors as merchants with competitive relationship.

For example, when a user browses at the page of the merchant A, B, C and finally performs a ordering action at the merchant a, it is considered that a competitive relationship exists between the merchants A, B and C; for another example, if the user generates the order-leaving behavior at the merchant a and the merchant B within a week, the merchant a and the merchant B are considered to have a competitive relationship; for another example, after the user performs the store-to-store consumption operation through the device 1, it may be considered that a competitive relationship exists between merchants related to the online operation of the device 1; also for example, merchant C and merchant D are considered to be in a competitive relationship if the geographic locations of merchant C and merchant D are relatively close or the underhung product categories are the same. In some embodiments of the present application, merchant a browses merchant E, and merchant a and merchant E may be considered to have a competitive relationship.

In some embodiments of the present application, the merchant having a competitive relationship is further determined by: and taking the merchants having competition relation with the same merchant as the merchants having competition relation. For example, if merchant a and merchant B have a competitive relationship and merchant B and merchant F have a competitive relationship, merchant a and merchant F may also be considered to have a competitive relationship.

In specific implementation, a screening strategy can be set according to specific business requirements to determine merchants with competitive relationships. Further, the competitive relationship may be determined from different dimensions. In some embodiments of the present application, the at least two dimensions are selected from individual dimensions or combinations of dimensions among a user rating dimension, a merchant inflow and outflow dimension, a to store overlap dimension, a business circle dimension, a price dimension, a star rating dimension. When the competition relationship network is constructed based on different dimensions, the merchant data is analyzed by adopting a corresponding data analysis method so as to determine the merchant competition relationship required by constructing the competition relationship network and a first competition strength index between merchants corresponding to the currently constructed competition relationship network. The first competition strength index is used for measuring the competition strength between merchants and can be obtained by adopting a general ranking algorithm such as a pagerank method for the constructed competition network. The merchant data may also be determined by analyzing the merchant data in the dimension corresponding to the competitive relationship network, for example, according to any one or more of attribute information of the merchant, the number of times that the merchant is associated with the preset behavior, and the number of times that the merchant is rejected by the preset behavior.

The following describes some specific construction processes of the contention resolution network by way of example.

When a competitive relationship network is constructed based on user evaluation dimensions, user evaluation data in merchant data are extracted first, and then merchants with competitive relationships are determined according to the similarity of the user evaluation data. For example, if a certain user or a certain specified user group includes "breakfast is very delicious" in the postings of different business a and business B pages, it is determined that the business a and the business B have a competitive relationship, and the business a is used as the source node a and the business B is used as the target node NodeB to construct the directed edge AB. Furthermore, a full directed graph can be constructed through the comment data of the users in the whole network, and then the full directed graph is searched through a ranking method, so that the competition strength index of the commercial tenant A and the commercial tenant B can be determined. And determining a first competition strength index of the merchant A and the merchant B based on the user evaluation dimension according to the times of the similar comments of the same user appearing in the merchant A and the merchant B. For example, the occurrence probability of the similar comment of the same user in the comment of the merchant a is compared with the occurrence probability of the similar comment of the same user in the comment of the merchant B, then the difference is added by 1 to obtain the reciprocal, and the obtained reciprocal is used as the first competition strength index based on the user evaluation dimension of the merchant a and the merchant B. And finally, constructing a competition relation network based on the user evaluation dimension according to the source node NodeA, the target node NodeB and the directed edge AB by taking the first competition strength index based on the user evaluation dimension of the user A and the merchant B as the attribute information of the directed edge AB. In specific implementation, the establishment of the competition relationship network based on the user evaluation dimension may further include nodes corresponding to other merchants having competition relationships with the merchant a or the merchant B based on the user evaluation dimension, and other nodes having competition relationships.

When a competitive relationship network is constructed based on merchant inflow and outflow dimensions, data of browsing, ordering, check-in and the like of a user in merchant data are extracted first, and then merchants with competitive relationships are determined according to the extracted data. In some embodiments of the present application, the browsing, selection, and presence of the user is not complete on-line, and the offline and online data need to be integrated for competitive network construction. For example, when a user logs in a hotel under a certain subscriber line, the platform acquires the equipment-related information of the user by accessing the check-in system of the hotel, and backtracks and restores the hotel selection process of the user according to the equipment-related information, so that the hotel related in the selection process is determined as a merchant with competitive relationship. And adding nodes and directed edges corresponding to the determined merchants with the competitive relationship in a competitive relationship network constructed based on the merchant inflow and outflow dimensions. Furthermore, a total digraph can be constructed through the browsing data of the users of the whole network, and then the total digraph is searched through a ranking method, so that the competition strength index of the commercial tenant A and the commercial tenant B can be determined. And determining a first competition strength index of the merchants with the competitive relationship based on the merchant inflow and outflow dimensions according to a probability curve of the merchants with the competitive relationship appearing in the ordering process of the same user. For example, the reciprocal of the average difference of the probability distributions of the merchant a and the merchant B occurring in the ordering process of the multiple users is taken as a first competition strength index of the merchant a and the merchant B based on the merchant inflow and outflow dimension. And finally, taking the first competition strength index as the attribute information of the directed edge of the node corresponding to the connection commercial tenant A and the commercial tenant B in the current competition relationship network.

When a competitive relationship network is constructed based on business community dimensionality, static data such as geographic positions, administrative region attributes and the like in merchant data are extracted, and then merchants with competitive relationships are determined according to information such as geographic position distance and administrative region attributes of the merchants. For example, merchants in the same administrative area are considered to be competitive merchants. Further, a first competition strength index of the merchant can be determined according to the distance between the geographical positions of the merchant, and the closer the distance is, the stronger the competition is. And then, adding nodes and directed edges in the competition relationship network based on the business circle dimensionality according to the similar method to construct the competition relationship network.

In some embodiments of the present application, a corresponding network of competitive relationships may also be constructed based on price dimensions, store-to-store overlap dimensions, star-level dimensions, and the like. A competitive relationship network can also be constructed based on price and business circle dimensions, business circle and star-level dimensions and the like. With the continuous enrichment of merchant data types and the improvement of business requirements, a competitive relationship network can be constructed based on more dimensions.

Step 220, according to a pre-constructed competition relationship network, determining candidate competition merchants mined by a target merchant based on the competition relationship network, and a first competition strength index between the candidate competition merchants and the target merchant based on a corresponding competition relationship network.

In some embodiments of the present application, if a competitive relationship network is constructed in advance according to merchant data based on a certain dimension, the candidate competitive merchant information of the target merchant is determined according to the competitive relationship network constructed in advance, where the candidate competitive merchant information includes: and the candidate competitive merchants are mined based on the competitive relationship network, and the candidate competitive merchants are based on the first competitive strength index of the corresponding competitive relationship network and the target merchant.

In some embodiments of the present application, the determining, according to at least two pre-constructed competition relationship networks, candidate competitive merchants mined by a target merchant based on each competition relationship network, and a first competition strength index between each candidate competitive merchant and the target merchant based on a corresponding competition relationship network includes: respectively executing the following operations for each pre-constructed competition relationship network: determining a directed edge of a source node corresponding to a connection target merchant in the competition relationship network as a target directed edge; using the commercial tenant corresponding to the target node pointed by the target directed edge as a candidate competitive commercial tenant of the target commercial tenant based on the competitive relationship network; and taking the attribute information of the target directed edge as a first competition strength index of a candidate competition merchant corresponding to the target node pointed by the target directed edge based on the competition relationship network and the target merchant.

By taking an example that the competition network Net1 includes nodes corresponding to the merchant A, B, C, assuming that the node corresponding to the merchant a is a source node, and the nodes corresponding to the merchant B and the merchant C are connected to the source node a through the directed edges AB and AC, respectively, and point to the nodes corresponding to the merchant B and the merchant C, the nodes corresponding to the merchant B and the merchant C are target nodes, and are denoted as NodeB and node C. Firstly, searching a node NodeA corresponding to a target merchant A in a competition relation network Net1, and according to network information; further traversing directed edges connected with the NodeA, and determining directed edges taking the NodeA as a source node as a target directed edge; after the target directed edges AC and AB are determined, target nodes NodeB and NodeB pointed to by the target directed edges AC and AB may be determined, and then, the merchant B and the merchant C corresponding to NodeB and NodeB are candidate competitive merchants of the target merchant a based on the competitive relationship network Net 1. Then, the attribute information of the target directed edges AC and AB is used as the first competition strength index of the competition-based network Net1 of the merchant B and the merchant C with the target merchant a, that is, the first competition strength index of the competition-based network Net 1.

Traversing other competition networks according to the same method, a group of candidate competitive merchants of the target merchant a corresponding to each competition network can be determined, and a first competition strength index between each candidate competitive merchant and the target merchant a in each group of candidate competitive merchants can be determined.

In some embodiments of the present application, after taking the attribute information of the target directed edge as a first competition strength indicator between the candidate competitive merchant corresponding to the target node pointed by the target directed edge and the target merchant based on the competition relationship network, the method further includes: respectively executing the following operations for each pre-constructed competition relationship network: taking the candidate competitive merchants of the target merchant based on the competitive relationship network as assumed target merchants; determining candidate competitive merchants of the assumed target merchant based on the competitive relationship network, and using the candidate competitive merchants as predicted candidate competitive merchants of the target merchant based on the competitive relationship network; determining a first competition strength index of the predicted candidate competition merchant based on the competition relationship network and the target merchant according to a first competition strength index of the candidate competition merchant of the assumed target merchant based on the competition relationship network and a first competition strength index of the assumed target merchant based on the competition relationship network and the target merchant; and supplementing the predicted candidate competitive merchants to the candidate competitive merchants of the target merchants based on the competitive relationship network.

After the competition relationship network is constructed, potential competition relationships can be further mined through the competition relationship network, so that potential competition merchants are prevented from being omitted. For example, in a certain competition network, the merchants corresponding to the node a and the node B have a competition relationship, and the merchants corresponding to the node C and the node B have a competition relationship, it can be considered that the merchants corresponding to the node a and the node C have a competition relationship. In specific implementation, for example, the candidate competitive merchants mined to the merchant a through the competitive relationship network Net1 include the merchant B, the candidate competitive merchants of the merchant B are mined through the competitive relationship network Net1, for example, the candidate competitive merchants of the merchant B competitive relationship network Net1 include the merchant H, and the merchant H is used as the predicted candidate competitive merchant of the target merchant a based on the competitive relationship network Net 1. Then, determining a single network competition strength index of the merchant H based on the competition relationship network Net1 and the candidate competition merchant B, and determining a single network competition strength index of the merchant H based on the competition relationship network Net1 and the target merchant A according to the single network competition strength index of the merchant H based on the competition relationship network Net1 and the candidate competition merchant B and the single network competition strength index of the target merchant A based on the competition relationship network Net1 and the candidate competition merchant B. And finally, supplementing the predicted candidate competitive merchant H to the candidate competitive merchant of the target merchant A based on the competitive relationship network, and expanding the candidate competitive merchant of the target merchant A based on the competitive relationship network into a merchant B and a merchant H.

And step 230, determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index.

In particular, a set of candidate competing merchants may be determined by each competing relationship network. If a plurality of competition relationship networks are constructed in advance, a plurality of groups of candidate competition merchants of the target merchant and a first competition strength index of each candidate competition merchant and the target merchant in each group of candidate competition merchants can be determined through different competition relationship networks. The first competition strength index is a competition strength index of the target merchant and a candidate competition merchant determined according to a competition relationship network. In specific implementation, a group of candidate competitive merchants determined by the target merchant based on each competitive relationship network may be respectively ranked according to the first competitive strength index, and then a preset number of candidate competitive merchants with the highest first competitive strength index in each group of candidate competitive merchants are selected as candidate competitive merchants to be compared corresponding to the corresponding competitive relationship network. The preset number is determined according to business needs, and may be, for example, the number of competitive seed merchants of the target merchant. According to the first competition strength index, partial candidate competitive merchants are selected from the candidate competitive merchants recorded by each competition relationship network and compared with competitive seed merchants, and the accuracy of competitive merchant mining can be improved.

Assuming that the constructed competitive relationship network comprises a competitive relationship network Net1 constructed based on the business circle dimension and a competitive relationship network Net2 constructed based on the price dimension, the candidate competitive merchants of the target merchant a determined according to the competitive relationship network Net1 comprise merchant B (Net1_ Socre _ B), C (Net1_ Socre _ C), D (Net1_ Socre _ D) and E (Net1_ Socre _ E), and the candidate competitive merchants of the target merchant a determined according to the competitive relationship network Net2 comprise merchant B (Net2_ Socre _ B), C (Net2_ Socre _ C), D (Net2_ Socre _ D) and F (Net2_ Socre _ F). Wherein Net1_ Socre _ B, Net1_ Socre _ C, Net1_ Socre _ D and Net1_ Socre _ E are the first competition strength index of the merchant B, C, D, E with the target merchant A recorded in the competition network Net1, and Net2_ Socre _ B, Net2_ Socre _ C, Net2Socre _ D and Net2_ Socre _ F are the first competition strength index of the merchant B, C, D, F with the target merchant A recorded in the competition network Net 2. Assuming that the seed competitive merchants of the target merchant include merchant B, merchant C, and merchant D, 3 candidate competitive merchants with the highest first competitive strength index may be selected from candidate competitive merchants B, C, D, E of target merchant a determined according to the competitive relationship network Net1 as a group of candidate competitive merchants to be compared, for example, the determined candidate competitive merchants to be compared are merchants B, C and E; then, 3 candidate competitive merchants with the highest first competitive strength index are selected from the candidate competitive merchants B, C, D, F of the target merchant a determined according to the competitive relationship network Net2 as a group of candidate competitive merchants to be compared, for example, the determined candidate competitive merchants to be compared are merchants B, C and D.

Step 240, comparing the candidate competitive merchants to be compared with the competitive seed merchants of the target merchant, and determining the competitive merchants of the target merchant.

When a competition relationship network is pre-constructed, the step of determining the competitive merchant of the target merchant by comparing the candidate competitive merchant to be compared with the competitive seed merchant of the target merchant comprises the following steps: determining the coverage rate of the candidate competitive merchants to be compared to the competitive seed merchants; if the coverage rate meets a preset coverage rate threshold, determining the candidate competitive commercial tenants to be compared as competitive commercial tenants of the target commercial tenant; and if the coverage rate does not meet the preset coverage rate threshold value, outputting that the mining competition merchant fails.

Still taking candidate competitive merchants to be compared of the target merchant a determined according to the competitive relationship network Net1 as follows, where the candidate competitive merchants to be compared of the target merchant a include merchant B (Net1_ sorre _ B), merchant C (Net1_ sorre _ C), and merchant E (Net1_ sorre _ E), the competitive seed merchant of the target merchant a includes merchants B, C and merchant F, the preset coverage threshold is 80%, the coverage rate of the candidate competitive merchants to be compared to the competitive seed merchant is 66.7%, and 66.7% is less than 80%, and when the coverage rate is considered not to satisfy the preset coverage threshold, the output mining competitive merchant fails. If the preset coverage rate threshold is 60%, the coverage rate of the candidate competitive merchants to be compared to the competitive seed merchants is 66.7%, and 66.7% is greater than 60%, the coverage rate is considered to meet the preset coverage rate threshold, all the candidate competitive merchants to be compared may be determined as competitive merchants of the target merchant, and a preset number of candidate competitive merchants to be compared, of which the first competitive strength index is the highest, may also be determined as competitive merchants of the target merchant.

Still in the following, by taking as an example that the candidate competing merchants to be compared of the target merchant a determined according to the competition relationship network Net1 include merchant B (Net1_ sorre _ B), C (Net1_ sorre _ C) and E (Net1_ sorre _ E), and the candidate competing merchants to be compared of the target merchant a determined according to the competition relationship network Net2 include merchant B (Net2_ sorre _ B), C (Net2_ sorre _ C) and D (Net2_ sorre _ D), the technical solution for determining the competing merchants of the target merchant by comparing the candidate competing merchants to be compared with the competing seed merchants of the target merchant is explained in detail.

In some embodiments of the present application, the determining a competing merchant of the target merchant by comparing the candidate competing merchant to be compared with the competing seed merchant of the target merchant includes: determining a second competition strength index of the candidate competition merchant to be compared and the target merchant by performing weighted fusion on the first competition strength index of the candidate competition merchant to be compared with the corresponding weight; determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant; if the difference index meets the preset difference index condition, determining competitive merchants of the target merchants from the candidate competitive merchants to be compared according to the second competitive strength index; if the difference index does not meet the preset difference index condition, taking the difference index meeting the preset difference index condition as a target, adjusting the weight corresponding to the first competition strength index, and skipping to the step of determining a second competition strength index of the candidate competition merchant to be compared and the target merchant by performing weighted fusion on the first competition strength index of the candidate competition merchant to be compared with the corresponding weight. The difference index described in the embodiment of the present application is used for the difference between the constant merchants.

In this embodiment, candidate competing merchants to be compared for target merchant a determined according to the competing relationship networks Net1 and Net2 include merchants B, C, D and E. Firstly, the first competition strength indexes of the candidate competition merchants to be compared are weighted and fused by corresponding weights, and second competition strength indexes of the candidate competition merchants to be compared and the target merchant are determined. Taking merchant B as an example, the first competition strength index of merchant B and target merchant a determined according to the competition network Net1 is Net1_ sorre _ B, the first competition strength index of merchant B and target merchant a determined according to the competition network Net2 is Net2_ sorre _ B, and assuming that the initial weights of the competition networks Net1 and Net2 are both 0.5, the second competition strength index of merchant B and target merchant a is calculated by the following formula:

net _ Socre (i) ═ a × Net1_ Socre (i) + B × Net2_ Socre (i), where i ═ B and Net _ Socre (i) are second competition strength indicators of the merchant B and the target merchant a, and a ═ B ═ 0.5, Net1_ Socre (i) ═ Net1_ Socre _ B, Net2_ Socre (i) ═ Net2_ Socre _ B.

Referring to the above formula, the second competition strength index of the candidate competing merchants B, C, D and E to be compared may be determined.

Then, the difference index of the candidate competitive merchant to be compared and the competitive seed merchant is determined.

In some embodiments of the present application, each competitive seed merchant corresponds to a third competitive strength index that is obtained in advance and is associated with the target merchant, and the determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant includes at least one of the following: determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant according to the candidate competitive merchant to be compared and the competitive seed merchant; and determining the difference index of the candidate competitive merchant to be compared and the competitive seed merchant according to the second competitive strength index of the candidate competitive merchant to be compared and the third competitive strength index of the competitive seed merchant and the target merchant.

Taking the example that the competitive seed merchants of the target merchant a include merchants B, C and F, the candidate competitive merchants to be compared include: B. c, D and E, including competing seed merchants B and C, the recall accuracy can be determined to be 66.7%, i.e., the difference indicator is 66.7%.

In some embodiments of the present application, according to the candidate competitive merchant to be compared and the competitive seed merchant, determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant may be: and taking the recall rate of the candidate competitive merchants to be compared in the candidate competitive merchants to be compared or the accuracy rate of the candidate competitive merchants to be compared relative to the competitive seed merchants as the difference index of the information of the candidate competitive merchants to be compared and the competitive seed merchants.

In some embodiments of the present application, determining, according to the second competition strength index of the candidate competition merchant to be compared and the third competition strength index of the competition seed merchant and the target merchant, the difference index between the candidate competition merchant to be compared and the competition seed merchant may be: and taking the root mean square error of the second competition strength index of the candidate competition merchant to be compared and the third competition strength index of the competition seed merchant and the target merchant as the difference index of the information of the candidate competition merchant to be compared and the competition seed merchant.

Still taking the example that the competitive seed merchants of the target merchant a include merchants B, C and F, assuming that the third competitive strength indexes of the competitive seed merchants B, C and F and the target merchant a are 0.8, 0.1 and 0.1, respectively, the candidate competitive merchants to be compared include: B. c, D and E are 0.55, 0.3, 0.1, respectively, then the root mean square error of the third competition strength index of the competitor seed merchant B, C and F and the second competition strength index of the candidate competitor merchant to be compared is:

after the difference indexes of the competitive seed commercial tenants and the candidate competitive commercial tenants to be compared are determined, judging the difference indexes according to preset difference index conditions to determine whether the competitive commercial tenants of the target commercial tenants are mined from the candidate competitive commercial tenants to be compared according to the current second competition strength indexes. If the difference index meets the preset difference index condition, for example, the recall accuracy is greater than a preset accuracy threshold or the root mean square error is less than a preset root mean square error threshold, then all candidate competitive merchants to be compared or the part of the candidate competitive merchants to be compared with the highest second competition strength index are determined as the competitive merchants of the target merchant. If the difference index does not meet the preset difference index condition, for example, the recall accuracy is less than or equal to a preset accuracy threshold or the root mean square error is greater than or equal to a preset root mean square error threshold, iteratively adjusting the weight corresponding to the first competition strength index by taking the difference index meeting the preset difference index condition as a target, and determining a second competition strength index between the candidate competition merchant to be compared and the target merchant by performing weighted fusion on the first competition strength index of the candidate competition merchant to be compared with the corresponding weight until it is determined that the competition merchant mining the target merchant among the candidate competition merchants to be compared by using the current second competition strength index is appropriate.

Thus far, competing merchants for the target merchant have been identified.

In specific implementation, the iterative adjustment is essentially a data fitting process, each competition relationship network corresponds to a first competition strength index for any two merchants, the first competition strength index is input for data fitting, the specific data fitting method can adopt a least square method for linear fitting and optimization, and can also adopt other optimization methods such as a logistic regression method for optimizing the ranking results of candidate competitive merchants to be compared.

In some embodiments of the present application, the competitor seed merchant is determined by any one of the following methods: mining preset behavior data of the target merchant to determine competitive seed merchants of the target merchant; and determining competitive seed merchants of the target merchants according to the labeling information of the target merchants to the merchants. Taking a hotel merchant as an example, a target hotel can log in hotel account or software at a hotel end, then select other hotels by self, and mark the selected competitive hotel as a competitive seed merchant by adding the competitive hotel to the 'my attention' list; or, the logs of other hotels browsed by the target hotel in a certain time are subjected to statistical analysis, the browsed hotels are sorted according to browsing frequency, and the high-frequency browsed hotels of the same type are used as competitive seed merchants.

In a specific implementation, the third competition strength index of the competitive seed merchant and the target merchant may refer to a determination method for determining the first competition strength index of the candidate competitive merchant and the target merchant, and details are not repeated here. In some embodiments of the present application, the third competition strength index between the competitive seed merchant and the target merchant may also be determined according to a time for adding the competitive seed to the "my attention" list, or a browsing frequency. In other embodiments of the present application, the third competition strength index between the competitive seed merchant and the target merchant may be further set by the merchant.

In some embodiments of the present application, the method further comprises: and updating the competitive seed commercial tenant according to the operation of the target commercial tenant on the competitive seed commercial tenant. In other embodiments of the present application, after the comparing the candidate competing merchants to be compared with the competing seed merchant of the target merchant and mining the competing merchant list of the target merchant, the method further includes: and updating the competitive seed commercial tenant according to the operation of the mined competitive commercial tenant. In still other embodiments of the application, the competitive seed merchant may be updated by using the above two methods simultaneously. And updating the operation of the competitive seed merchant is not limited to after determining the competitive merchant of the target merchant according to the operation of the target merchant on the competitive seed merchant.

Step 250, updating the competitive seed merchants according to the operation of the target merchants on the competitive seed merchants and/or the mined competitive merchants.

After the competing merchants of the target merchant are determined, the competing merchant list may be fed back to the target merchant. In particular implementations, the target merchant may designate merchants in the competitive merchant list as seed competitive merchants, for example, by adding one or some merchants in the competitive merchant list to "my attention" through interface interaction. The target merchant may also remove a merchant from "my attention" and update the competing seed merchant.

And when the competitive merchant of the target merchant is mined next time, comparing the determined candidate competitive merchant to be compared with the updated competitive seed merchant to determine the competitive merchant of the target merchant.

The competitor mining method disclosed in the embodiment of the application constructs a competition relationship network corresponding to at least two dimensions one by one from the at least two dimensions according to the obtained merchant data, determines candidate competitive merchants mined by a target merchant based on each competition relationship network according to at least two pre-constructed competition relationship networks, determines candidate competitive merchants to be compared from the candidate competitive merchants mined based on each competition relationship network based on the corresponding competition relationship network and a first competition strength index of the target merchant, then determines candidate competitive merchants to be compared from the candidate competitive merchants mined based on each competition relationship network according to the first competition strength index, further determines competitive merchants of the target merchant by comparing the candidate competitive merchants to be compared with competitive seed merchants of the target merchant, and finally determines the competitive merchants of the target merchant according to the operation of the target merchant on the competitive seed merchants and/or the mined competitive merchants, and the competitive seed merchants are updated, so that the problem that competitors mined in the prior art are inaccurate is solved. According to the competitor mining method disclosed by the embodiment of the application, the competitor seed commercial tenants are updated according to the operation of the commercial tenants, the personalized dynamic sequencing of the competitive commercial tenants can be realized, the self-learning competitive commercial tenant list is output, the commercial tenant requirements can be really understood, and the competitors can be intelligently and accurately mined.

EXAMPLE III

The competitor digging device disclosed by the embodiment is shown in fig. 3, and comprises:

a candidate competitive merchant information determining module 310, configured to determine, according to a pre-established competitive relationship network, a candidate competitive merchant mined by a target merchant based on the competitive relationship network, and a first competitive strength index between the candidate competitive merchant and the target merchant based on a corresponding competitive relationship network;

a candidate competitive merchant determining module 320 for determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index determined by the candidate competitive merchant information determining module 310;

a competitive merchant list mining module 330, configured to determine a competitive merchant of the target merchant by comparing the candidate competitive merchant to be compared with the competitive seed merchant of the target merchant.

Optionally, as shown in fig. 4, the apparatus further includes a competitive seed merchant updating module 340, where the competitive seed merchant updating module 340 is further configured to:

updating the competitive seed commercial tenant according to the operation of the target commercial tenant on the competitive seed commercial tenant; and/or the presence of a gas in the gas,

and updating the competitive seed commercial tenant according to the operation of the mined competitive commercial tenant.

Optionally, the competitive relationship network is constructed according to the obtained merchant data, and in some embodiments of the present application, as shown in fig. 4, the apparatus further includes:

a competition relationship network construction module 350, configured to construct a competition relationship network according to the obtained merchant data;

wherein the merchant data comprises dynamic data comprising at least one of: the online dynamic data of the merchant, the offline dynamic data of the merchant, the online dynamic data of the user of the merchant and the offline dynamic data of the user of the merchant. A competition relationship network is established by combining the online and offline dynamic behavior data of the user, so that real user data driving is realized, and the practicability and accuracy of competitor mining are improved. In some embodiments of the present application, the merchant data further comprises static data, the static data further comprising: intrinsic attribute information of the merchant.

Optionally, the contention relation network includes: the node includes a source node and a target node, the directed edge connects the nodes corresponding to merchants having a competitive relationship, the attribute information of the directed edge includes a first competitive strength index of the target node of the directed edge and the first competitive strength index of the merchant corresponding to the source node, and the candidate competitive merchant information determining module 310 is further configured to:

determining a directed edge of a source node corresponding to a connection target merchant in the competition relationship network as a target directed edge;

using the commercial tenant corresponding to the target node pointed by the target directed edge as a candidate competitive commercial tenant of the target commercial tenant based on the competitive relationship network;

and taking the attribute information of the target directed edge as a first competition strength index of a candidate competition merchant corresponding to the target node pointed by the target directed edge based on the competition relationship network and the target merchant.

Optionally, the candidate competing merchant information determining module 310 is further configured to:

taking the candidate competitive merchants of the target merchant based on the competitive relationship network as assumed target merchants;

determining candidate competitive merchants of the assumed target merchant based on the competitive relationship network, and using the candidate competitive merchants as predicted candidate competitive merchants of the target merchant based on the competitive relationship network;

determining a first competition strength index of the predicted candidate competition merchant based on the competition relationship network and the target merchant according to a first competition strength index of the candidate competition merchant of the assumed target merchant based on the competition relationship network and a first competition strength index of the assumed target merchant based on the competition relationship network and the target merchant;

and supplementing the predicted candidate competitive merchants to the candidate competitive merchants of the target merchants based on the competitive relationship network.

Optionally, the competitive merchant list mining module 330 is further configured to:

determining the coverage rate of the candidate competitive merchants to be compared to the competitive seed merchants;

if the coverage rate meets a preset coverage rate threshold, determining the candidate competitive commercial tenants to be compared as competitive commercial tenants of the target commercial tenant;

and if the coverage rate does not meet the preset coverage rate threshold value, outputting that the mining competition merchant fails.

determining a second competition strength index of the candidate competition merchant to be compared and the target merchant by performing weighted fusion on the first competition strength index of the candidate competition merchant to be compared with the corresponding weight;

determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant;

if the difference index meets the preset difference index condition, determining competitive merchants of the target merchants from the candidate competitive merchants to be compared according to the second competitive strength index;

if the difference index does not meet the preset difference index condition, taking the difference index meeting the preset difference index condition as a target, adjusting the weight corresponding to the first competition strength index, and skipping to the step of determining a second competition strength index of the candidate competition merchant to be compared and the target merchant by performing weighted fusion on the first competition strength index of the candidate competition merchant to be compared with the corresponding weight.

Optionally, each competitive seed merchant corresponds to a third competitive strength index, which is obtained in advance, of the target merchant, and the determining of the difference index between the candidate competitive merchant to be compared and the competitive seed merchant includes at least one of the following:

determining a difference index between the candidate competitive merchant to be compared and the competitive seed merchant according to the candidate competitive merchant to be compared and the competitive seed merchant;

and determining the difference index of the candidate competitive merchant to be compared and the competitive seed merchant according to the second competitive strength index of the candidate competitive merchant to be compared and the third competitive strength index of the competitive seed merchant and the target merchant.

Optionally, the competitive seed merchant is determined by any one of the following methods:

mining preset behavior data of the target merchant to determine competitive seed merchants of the target merchant;

and determining competitive seed merchants of the target merchants according to the labeling information of the target merchants to the merchants.

The competitor excavation device disclosed in the embodiment of the application is used for implementing each step of the competitor excavation method described in the first embodiment and the second embodiment of the application, and specific implementation modes of each module of the device refer to the corresponding step, and are not described herein again.

The competitor mining device disclosed by the embodiment of the application determines a candidate competitive merchant of a target merchant, which is mined based on a competitive relationship network, according to the pre-established competitive relationship network, and determines a first competitive strength index of the candidate competitive merchant and the target merchant based on the corresponding competitive relationship network; determining candidate competitive merchants to be compared from the candidate competitive merchants mined based on the competitive relationship network according to the first competitive strength index; the candidate competitive merchants to be compared are compared with the competitive seed merchants of the target merchants to determine the competitive merchants of the target merchants, so that the problem that competitors mined in the prior art are inaccurate is solved. The competitor mining device disclosed by the embodiment of the application excavates the candidate competitive merchants through the competitive relationship network constructed based on the data, and further sorts the excavated candidate competitive merchants by combining the seed competitive merchants, so that the accuracy of excavating to competitors is improved.

Furthermore, the competitor mining device disclosed in the embodiment of the application updates the competitor seed commercial tenants according to the operation of the commercial tenants, can realize personalized dynamic sequencing of the competitive commercial tenants, outputs a self-learning competitive commercial tenant list, can really understand the requirements of the commercial tenants, and can intelligently and more accurately mine competitors.

Accordingly, the present application also discloses an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the computer program to implement the competitor mining method according to the first embodiment and the second embodiment of the present application. The electronic device can be a PC, a mobile terminal, a personal digital assistant, a tablet computer and the like.

The present application also discloses a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps of the competitor mining method as described in the first and second embodiments of the present application.

The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.

The competitor mining method and device provided by the application are introduced in detail, a specific example is applied in the description to explain the principle and the implementation mode of the application, and the description of the embodiment is only used for helping to understand the method and the core idea of the application; meanwhile, for a person skilled in the art, according to the idea of the present application, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present application.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.

Claims

1. A competitor mining method is applied to electronic equipment and comprises the following steps:

determining candidate competitive merchants mined by a target merchant based on a competitive relationship network according to a pre-constructed competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on a corresponding competitive relationship network; the method specifically comprises the following steps: respectively searching nodes corresponding to target merchants in each pre-constructed competition relationship network, determining merchants corresponding to the target nodes pointed by the directional edges of the target merchants as source nodes as candidate competitive merchants of the target merchants in the current competition relationship network, and using the attribute information of the directional edges as a first competition strength index of the corresponding candidate competitive merchants and the target merchants;

the first competition strength index is used for measuring the competition strength among the merchants, and is obtained by searching the constructed competition relationship network by adopting a universal ranking algorithm, or is determined by analyzing the merchant data in the dimension corresponding to the competition relationship network;

determining competitive merchants of the target merchants by comparing the candidate competitive merchants to be compared with competitive seed merchants of the target merchants;

the competition relationship network is constructed according to the obtained merchant data, and the merchant data comprises dynamic data.

2. The method of claim 1, further comprising:

3. The method of claim 1 or 2, wherein the dynamic data comprises at least one of: the online dynamic data of the merchant, the offline dynamic data of the merchant, the online dynamic data of the user of the merchant and the offline dynamic data of the user of the merchant.

4. The method of claim 3, wherein the contention relationship network comprises: the method comprises the following steps that each node corresponds to a merchant, each node comprises a source node and a target node, each directed edge is connected with the corresponding node of the merchant with a competition relationship, the attribute information of each directed edge comprises a first competition strength index of the merchant corresponding to the target node and the source node of the directed edge, the candidate competition merchants mined by the target merchants based on the competition relationship network are determined according to a pre-established competition relationship network, and the candidate competition merchants and the first competition strength index of the target merchants based on the corresponding competition relationship network comprise:

5. The method according to claim 4, wherein after the step of using the attribute information of the target directed edge as a first competition strength indicator between the candidate competitive merchant corresponding to the target node pointed by the target directed edge and the target merchant based on the competition relationship network, the method further comprises:

6. The method of claim 1, wherein the step of determining the competing merchant of the target merchant by comparing the candidate competing merchant to be compared with the competing seed merchant of the target merchant comprises:

7. The method of claim 1, wherein the step of determining the competing merchant of the target merchant by comparing the candidate competing merchant to be compared with the competing seed merchant of the target merchant comprises:

if the difference index meets a preset difference index condition, determining competitive merchants of the target merchants from the candidate competitive merchants to be compared according to the second competitive strength index;

8. The method according to claim 7, wherein each of the competitive seed merchants corresponds to a third competitive strength index with the target merchant, and the step of determining the difference index between the candidate competitive merchant to be compared and the competitive seed merchant includes at least one of:

9. The method of claim 1, wherein the competing seed merchant is determined by any one of the following methods:

10. A competitor digging device is applied to electronic equipment and is characterized by comprising:

the candidate competitive merchant information determining module is used for determining candidate competitive merchants mined by a target merchant based on a competitive relationship network according to a pre-established competitive relationship network, and determining a first competitive strength index of the candidate competitive merchants and the target merchant based on the corresponding competitive relationship network; the method is specifically used for: respectively searching nodes corresponding to target merchants in each pre-constructed competition relationship network, determining merchants corresponding to the target nodes pointed by the directional edges of the target merchants as source nodes as candidate competitive merchants of the target merchants in the current competition relationship network, and using the attribute information of the directional edges as a first competition strength index of the corresponding candidate competitive merchants and the target merchants;

the first competition strength index is used for measuring the competition strength among the commercial tenants and is obtained by searching the constructed competition network by adopting a universal ranking algorithm, or the first competition strength index is determined by analyzing the commercial tenant data in the dimension corresponding to the competition relationship network;

a competitive merchant list mining module, configured to determine competitive merchants of the target merchant by comparing the candidate competitive merchants to be compared with competitive seed merchants of the target merchant;

11. The apparatus of claim 10, further comprising: a competitive seed merchant update module, the competitive seed merchant update module further to:

12. The apparatus of claim 10 or 11, wherein the dynamic data comprises at least one of: the online dynamic data of the merchant, the offline dynamic data of the merchant, the online dynamic data of the user of the merchant and the offline dynamic data of the user of the merchant.

13. The apparatus of claim 12, wherein the contention relationship network comprises: the node comprises a source node and a target node, the directed edge is connected with the nodes corresponding to the merchants with competition relationships, the attribute information of the directed edge comprises first competition strength indexes of the target node of the directed edge and the merchants corresponding to the source node, and the candidate competition merchant information determining module is further used for:

14. The apparatus of claim 13, wherein the candidate competing merchant information determination module is further configured to:

15. The apparatus of claim 10, wherein the competitive merchant list mining module is further configured to:

16. The apparatus of claim 10, wherein the competitive merchant list mining module is further configured to:

17. The apparatus according to claim 16, wherein each of the competitive seed merchants corresponds to a third competitive strength index with the target merchant, and the determining of the difference index between the candidate competitive merchant to be compared and the competitive seed merchant includes at least one of:

18. The apparatus of claim 10, wherein the competing seed merchant is determined by any one of the following methods:

19. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the competitor mining method of any of claims 1-9 when executing the computer program.

20. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the competitor mining method of any one of claims 1 to 9.