WO2018205371A1 - 风险评估方法、装置、服务器和存储介质 - Google Patents

风险评估方法、装置、服务器和存储介质 Download PDF

Info

Publication number
WO2018205371A1
WO2018205371A1 PCT/CN2017/090575 CN2017090575W WO2018205371A1 WO 2018205371 A1 WO2018205371 A1 WO 2018205371A1 CN 2017090575 W CN2017090575 W CN 2017090575W WO 2018205371 A1 WO2018205371 A1 WO 2018205371A1
Authority
WO
WIPO (PCT)
Prior art keywords
node
risk
activated
social network
level
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2017/090575
Other languages
English (en)
French (fr)
Inventor
王健宗
黄章成
吴天博
肖京
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Publication of WO2018205371A1 publication Critical patent/WO2018205371A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities

Definitions

  • the present invention relates to the field of computer technologies, and in particular, to a risk assessment method, apparatus, server, and storage medium.
  • a risk assessment is a quantitative assessment of the likely impact and risk of a business or individual to be assessed before or after a risk event (but not yet). That is, risk assessment is to quantify the extent to which an event or thing is affected or lost.
  • the traditional evaluation methods commonly used in risk assessment mainly include: expert forecasting method, which means that a group of experts consists of a group of experts to conduct predictive discussions through symposiums. This method not only evaluates inefficiency, but also does not ensure objective reflection of reality; LEC Risk assessment method (this method uses the product of three factors related to system risk to evaluate the risk of operator casualties. These three factors are: L (likelihood), E (exposure, personnel) The degree of exposure to a hazardous environment) and C (consequence, the consequences of an accident) are a quantitative method of risk calculation. According to the hazard source identification record, quantitatively calculate the sub-segment brought by each hazard source, determine the maximum risk, and list the planned and planned control risks.
  • the LEC risk assessment method is based on empirical judgments on the classification of hazard levels. It is limited in application and cannot be universally applied. That is to say, the traditional risk assessment methods are mostly empirical static decision evaluation methods, the evaluation efficiency is low, and the evaluation process is not retroactive.
  • a risk assessment method and apparatus a server, and a storage medium are provided.
  • a risk assessment method comprising:
  • the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, and an evaluation subject node corresponding to the evaluation subject identifier And an edge set between the nodes, the edges of the edge set are all directed edges with weights;
  • Monitoring a risk factor determining, according to the risk factor, an activated first level activation node in the social network map, the first level activation node including at least one of the associated nodes;
  • a risk assessment device comprising:
  • An evaluation requesting module configured to receive a risk assessment request sent by the terminal, and extract an evaluation subject identifier carried in the risk assessment request;
  • a social network map finding module configured to search for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes an associated node directly or indirectly associated with the evaluation subject identifier, Evaluating the evaluation subject node corresponding to the subject identifier and the edge set between the nodes, and the edges in the edge set are all directed edges with weights;
  • a risk factor monitoring module configured to monitor a risk factor, determine an activated first level active node in the social network map according to the risk factor, and the first level activation node includes at least one of the associated nodes;
  • a risk transfer module configured to calculate, according to the social network graph, an accumulated weight of the first-level active node to its neighboring associated node, and determine whether the accumulated weight exceeds a preset activation threshold of the adjacent associated node to determine a second level Activate the node;
  • An evaluation result feedback module configured to determine, according to the first level activation node and the second level activation node, a next level activation node, until an accumulated weight of the activated association node is transmitted to the evaluation subject node, to obtain the evaluation The cumulative weight of the subject node, thereby obtaining the risk assessment result of the evaluation subject node, and transmitting the risk assessment result to the terminal.
  • a server comprising a memory and a processor, the memory storing computer executable instructions, the instructions being executed by the processor, causing the processor to perform the following steps:
  • the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, and an evaluation subject node corresponding to the evaluation subject identifier And an edge set between the nodes, the edges of the edge set are all directed edges with weights;
  • Monitoring a risk factor determining, according to the risk factor, an activated first level activation node in the social network map, the first level activation node including at least one of the associated nodes;
  • One or more non-volatile readable storage media storing computer-executable instructions, when executed by one or more processors, cause the one or more processors to perform the following steps:
  • the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, and an evaluation subject node corresponding to the evaluation subject identifier And an edge set between the nodes, the edges of the edge set are all directed edges with weights;
  • Monitoring a risk factor determining, according to the risk factor, an activated first level activation node in the social network map, the first level activation node including at least one of the associated nodes;
  • 1 is an application environment diagram of a risk assessment method in an embodiment
  • FIG. 2 is a schematic diagram showing the internal structure of a server in an embodiment
  • FIG. 3 is a flow chart of a risk assessment method in an embodiment
  • FIG. 4 is a diagram of a weighted social network in one embodiment
  • FIG. 5 is a flow chart involved in determining a first level activation node in an embodiment
  • FIG. 6 is a flow chart of a risk assessment method in another embodiment
  • FIG. 7 is a flow chart involved in generating a risk propagation path map in one embodiment
  • Figure 8 is a diagram of a weighted social network in another embodiment
  • FIG. 9 is a structural block diagram of a risk assessment apparatus in an embodiment
  • FIG. 10 is a structural block diagram of a risk factor monitoring module in an embodiment
  • Figure 11 is a block diagram showing the structure of a risk assessment apparatus in another embodiment
  • Figure 12 is a block diagram showing the structure of a risk assessment apparatus in still another embodiment
  • Figure 13 is a block diagram showing the structure of a risk assessment apparatus in still another embodiment.
  • an application environment diagram of a risk assessment method including a terminal 110 and a server 120.
  • Terminal 110 can communicate with server 120 over a network.
  • the terminal 110 may be at least one of a smartphone, a tablet, a notebook, and a desktop computer, but is not limited thereto.
  • the server 120 may be an independent physical server or a server cluster composed of a plurality of physical servers.
  • the terminal 110 transmits a risk assessment request to the server 120, which specifies the evaluation subject identifier to be evaluated.
  • the server receives the risk assessment request sent by the terminal, and searches for a social network map including the identifier of the evaluation subject, the social network map further includes an associated node directly or indirectly associated with the evaluation subject, and each node (including the evaluation subject node and the associated node) The weight of the association between them.
  • the impact weight of the risk factor on each node is determined according to the social network map, and it is determined whether the impact weight reaches the preset activation weight of the affected node, and if so, the first level activation node is determined.
  • the first-level activation node is used as the risk factor to determine the next-level activation node until the cumulative weight of the activated associated node is transmitted to the evaluation subject, and finally, the risk assessment result of the evaluation subject is determined.
  • the server feeds back the obtained risk assessment results to the terminal.
  • a server 120 includes a processor coupled through a system bus, a non-volatile readable storage medium, an internal memory, and a network interface.
  • the non-volatile readable storage medium of the server 120 stores an operating system, a database, and at least one computer executable instruction.
  • the processor can be caused to perform a risk assessment method as shown in FIG.
  • the database is used to store data, such as data such as social network maps involved in the execution of the risk assessment method.
  • the processor is used to provide computing and control capabilities to support the operation of the entire server 120.
  • the internal memory in the server provides a cached operating environment for operating systems, databases, and computer executable instructions in a non-volatile readable storage medium.
  • the network interface is used for communication connection with the terminal 110.
  • a risk assessment method is provided, which is applied to the server shown in FIG. 2, and specifically includes the following steps:
  • Step S202 Receive a risk assessment request sent by the terminal, and extract an evaluation subject identifier carried in the risk assessment request.
  • the terminal may send a risk assessment request to the server by using the terminal application corresponding to the server, and specify an evaluation subject identifier to be evaluated, where the identifier of the evaluation subject may be an enterprise identifier or an individual user identifier.
  • the server pre-builds at least one social network map, each of which includes a plurality of nodes having an association relationship with each other, and each node corresponds to one node identifier.
  • the terminal selects a social network map according to the social network map tag list given in the request page.
  • the terminal will expand the node identifier list of the social network graph in the request page, and the user selects one of the node identifiers or the plurality of node identifiers from the node identifier list as the evaluation subject identifier to be evaluated.
  • the terminal sends a risk assessment request to the server, the terminal carries the selected evaluation subject identifier.
  • Step S204 Search for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, an evaluation subject node corresponding to the evaluation subject identifier, and each node.
  • the edge set between the edges is the directed edge with weights.
  • the server looks up a social network map that includes an assessment subject identity. This method of determining the corresponding social network map is suitable for the selected evaluation subject identifier to exist in a unique social network map. That is, the evaluation subject is only a node in a social network map.
  • the terminal sends the risk assessment request, and the identifier of the social network map corresponding to the evaluation subject is required to be carried by the identifier, so that the server can find the identifier for the server.
  • Social network data for risk assessment
  • the social network map is a directed weighted map generated according to the association data of each node.
  • the relationship here may be a secured loan relationship, the corresponding associated node is a borrower, a guarantee room and a lender; it may also be a shareholder relationship, such as a business and a corporate legal person; an industry relationship, such as between an enterprise and an enterprise Relationships; or other associations, such as information associations, phone associations, address associations, and industry chain associations.
  • the related parties in the above relationship data constitute nodes in the social network graph, and the association relationship constitutes a set of edges between the nodes. On this basis, according to the set weight configuration rules, the weight of the edge between each node is configured.
  • the weight allocation rule may adopt the following method: the shareholder related party may determine the assignment rule according to the shareholding ratio; the guarantee related party may formulate the assignment rule according to the guarantee amount; the related party of the industry association may determine a value by trial mode; the information related party, Weighted valuation rules can be based on internal and external differences, impact levels, and timeliness; upstream and downstream industry chain stakeholders can choose valuation rules based on the average degree of upstream and downstream dependence of the industry.
  • the sum of the weights of all directed edges belonging to the node contained in each node should be less than 1.
  • FIG. 1 An example of a constructed weighted directed graph is shown in FIG.
  • the example diagram includes four nodes, the node N1 has an association relationship with the node N2 and the node N3, the influence weight of the node N1 on the node N2 is 0.7, the influence weight of the node N1 on the node N3 is 0.3, and the node N2 and the node N3 can mutually The impact, but the influence weight is different; the node N2 is also associated with the node N4, the node N2 has an influence weight of 0.6 on the node N4; the node N3 has a relationship with the node N4, and the influence weight of the node N2 on the node N4 is 0.4.
  • Step S206 Monitor the risk factor, determine the activated first-level active node in the social network map according to the risk factor, and the first-level active node includes at least one associated node.
  • the risk factor in this embodiment refers to a risk factor that can affect one or more of the associated nodes in the social network map searched for in step S204.
  • the risk factor can be intelligence information obtained by the server monitoring and analyzing big data.
  • one node in the social network map is the oil industry, and the intelligence information can be that the oil price rises to a set limit, and the number of oil companies in a certain area declares bankruptcy and so on.
  • the risk factor can also be a risk that the user assumes, such as a shortage of enterprise A funds in the social network map, and the departure of the corporate B legal person.
  • the risk factor may also be one of the associated nodes in the social network graph, and when the risk factor is detected, the associated node corresponding to the risk factor is activated. If the corporate B legal person leaves the company, the corporate node of the enterprise B is activated.
  • the node associated with the risk factor is searched, and the node having the relationship with the risk factor is activated, and the activated node is the first active node. If the oil price rises to a set limit, and the number of oil companies in a certain area declares bankruptcy, the nodes of the oil industry are activated.
  • Step S208 Calculate the cumulative weight of the first-level active node to the adjacent associated node according to the social network graph, and determine whether the accumulated weight exceeds the preset activation threshold of the adjacent associated node to determine the second-level active node.
  • the influence of the first-level active node on the adjacent node is determined by the relationship between the nodes in the social network map and the influence weight of the configuration. Specifically, the cumulative weight of the first-level active node to the adjacent-related node is calculated, and it is determined whether the cumulative weight exceeds the preset activation threshold of the node, and if so, the adjacent-related node is also activated.
  • the activation threshold of the pre-configured node N1 is 0.6
  • the activation threshold of the node N2 is 0.5
  • the activation threshold of the node N3 is 0.7
  • the activation threshold of the node N4 is 0.9.
  • the cumulative weight of the node N2 is the influence weight of the node N1, which is 0.7
  • the cumulative weight of the node N2 is greater than the activation weight of the node N2, and the node N2 is activated.
  • the node N1 cannot activate the node N3 (0.3 ⁇ 0.7). Therefore, the second level active node is N2.
  • Step S210 Determine the next-level active node according to the first-level activation node and the second-level activation node, until the cumulative weight of the activated associated node is transmitted to the evaluation subject node, and obtain the cumulative weight of the evaluation subject node, thereby obtaining the evaluation subject node.
  • the risk assessment results are sent to the terminal.
  • the risk transmission path is analyzed step by step based on the social network graph until the risk is transmitted to the evaluation subject, and the cumulative weight of the evaluation subject is calculated, and the cumulative weight of the evaluation subject is compared with the preset activation threshold of the evaluation subject, if the cumulative weight of the evaluation subject is greater than
  • the preset activation threshold is used to obtain the risk assessment result that the activation subject is activated, and the obtained risk assessment result is fed back to the terminal requesting the analysis.
  • the linear threshold model is used to determine step by step the social network is activated step by step (affected by the risk).
  • Correlation nodes in the end, determine the degree of risk transmitted to the assessment subject to determine whether the assessment subject is affected and then obtain the risk assessment results for the assessment subject, and achieve a risk assessment method to dynamically assess the risk.
  • the assessment path is traceable.
  • the related risk assessment including social relations, credit relations, industry attribution, and other comprehensive analysis and evaluation, the evaluation results are more reliable; based on the linear threshold model for risk degree judgment, the evaluation efficiency is higher.
  • step S206 monitoring a risk factor, determining an activated first-level active node in the social network map according to the risk factor, where the first-level active node includes at least one associated node, specifically including The following steps:
  • Step S302 Acquire a hot word set configured in advance for the associated node in the social network graph.
  • the server pre-configures hot words for at least one associated node in the social network map, and the hot words configured for the associated nodes may be one or a set of multiple hot words.
  • the hot word of a node is essentially a collection of words associated with a node that are frequently searched for. For example, a set of hot words for the node “ ⁇ Steel” is “iron ore”, “steel enterprise”, “special steel”, “bag dust removal” and “steel enterprise”; configured for the node “credit bank”
  • the collection of hot words is “non-performing assets”, “deposit-to-deposit ratio”, “stock market” and so on.
  • the configured hot word set is a risk factor that the terminal needs to monitor.
  • the risk index of the risk factor is reflected by the hot word lyric information corresponding to the hot word set.
  • Step S304 Collect hot word public opinion information according to the hot word set, wherein the hot word public opinion information is information capable of reflecting the trend dynamic of the associated node.
  • the hot words and lyric information are searched on the setting website or setting the platform, for example, the price of iron ore rises, and the special breakthrough of special steel research.
  • These hot words lyric information can directly or indirectly reflect the credit status and business status of the corresponding associated nodes.
  • Step S306 When the trend dynamic change of the associated node indicated by the hot word public opinion information reaches a set threshold, the indicated associated node is activated, and the activated associated node is the first-level active node.
  • the iron ore supply appears to set a large gap or the research of new substances occupy the iron In the market with more than 50% of ore, the “ ⁇ Steel” node corresponding to iron ore is activated.
  • the hot word lyrics may also be used as a node, and the influence weights of the hot word sensation node and the corresponding associated node may be configured.
  • the hot word sensation node is activated to determine whether the influence weight of the hot word sensation node on the associated node exceeds the preset activation threshold of the associated node, and if so, the associated node is activated, and if not, the individual hot word sensation node cannot be activated. The associated node.
  • the hot word lyrics is used as a node, and the hot word lyrics corresponds to a set number of hot words.
  • the hot word lyrics corresponds to a set number of hot words.
  • the hot word sensation node is activated.
  • the M node is a hot word sensation node, assuming that there are 5 keywords in the M node, that is, the M node threshold is 0.5.
  • the M node Based on the functions provided by the real-time monitoring platform of the microblog hotspot event, and monitoring the heat of the five keywords in real time, if three of the five words exceed the monitoring hotspot threshold, then the M node is not activated, if at the same time 5 The nodes all exceed the threshold and the M node is activated.
  • the risk factors that may affect the nodes can be automatically monitored, and risk monitoring and risk prediction can be automatically triggered to realize credit risk.
  • Real-time discovery and feedback provide a strong guarantee.
  • a risk assessment method is provided, and the method specifically includes the following steps:
  • Step S402 Receive a risk assessment request sent by the terminal, and extract an evaluation subject identifier carried in the risk assessment request.
  • Step S404 Search for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, an evaluation subject node corresponding to the evaluation subject identifier, and each node.
  • the edge set between the edges is the directed edge with weights.
  • Step S406 Receive at least one associated node specified by the terminal, and the designated associated node is a first-level active node.
  • Step S408 Calculate the cumulative weight of the first-level active node to the adjacent associated node according to the social network graph, and determine whether the accumulated weight exceeds the preset activation threshold of the adjacent associated node to determine the second-level active node.
  • Step S410 Determine the next-level active node according to the first-level activation node and the second-level activation node, until the cumulative weight of the activated associated node is transmitted to the evaluation subject node, and obtain the cumulative weight of the evaluation subject node, thereby obtaining the evaluation subject node.
  • the risk assessment results are sent to the terminal.
  • the activation node is specified by the end user to evaluate the risk status of the assessment subject when the designated node is activated.
  • the activation node By specifying the activation node by the user, it is possible to predict the impact of any node being activated on the evaluation subject, and the risk prediction is more flexible.
  • the risk propagation path is also recorded while assessing the assessment subject.
  • the risk assessment method further includes the following steps:
  • Step S502 Record the sequence in which the associated nodes brought by different risk factors are activated.
  • Different risk factors will trigger different first-level activation nodes, which in turn will result in different activation sequences from the first-level activation node to the other association nodes and the evaluation subject. As different risk factors are detected, different activation orders of the associated nodes will be generated according to the social network map.
  • the activation order is: N1-N2-(N3, N5)-N4.
  • the activation sequence is: N2-N5-N4. This example can only produce two activation sequences. When a more complex social network map is built according to the actual environment, different risk factors will generate multiple node activation sequences.
  • Step S504 Generate a risk propagation path map according to a sequence in which the associated nodes are activated, and push the generated risk propagation path map to the terminal.
  • a risk propagation path map is generated according to the sequence in which the recorded associated nodes are activated.
  • the risk propagation path map includes multiple branches to indicate different node activation sequences generated by different risk factors. Push the generated risk propagation path map to the terminal display to grasp the risk dynamics from a macro perspective.
  • Step S506 Determine at least one key associated node in the associated node according to the risk propagation path map, and monitor an activation state of the key associated node, and send an alarm message to the terminal when monitoring that the key associated node is activated.
  • the server will determine at least one key associated node based on the generated propagation path map.
  • the determined key association node may be a node with a large number of branches in the propagation path graph, that is, a node with the largest number of activation nodes, and may also be a node directly causing the evaluation subject node to be activated.
  • the key associated nodes may also be designated by the user. That is, after the server sends the generated propagation path map to the terminal, the receiving instruction of the key associated node sent by the terminal is received, and the key associated node identifier selected by the user carried in the instruction is extracted.
  • the server After determining the key associated nodes, the server will monitor the risk factors of the key associated nodes or the hot words of the key associated nodes in real time to control the activation status of the key associated nodes in time. When it is detected that the key associated node is activated, the alarm information is sent to the terminal in real time to implement the risk response measure in the first time.
  • a key monitoring method can achieve better monitoring results on the basis of reducing monitoring resources.
  • a risk assessment device comprising:
  • the evaluation request module 602 is configured to receive the risk assessment request sent by the terminal, and extract the evaluation subject identifier carried in the risk assessment request.
  • the social network map finding module 604 is configured to search for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes an association node directly or indirectly associated with the evaluation subject identifier, and the evaluation subject identifier The subject node and the edge set between the nodes are evaluated, and the edges in the edge set are directed edges with weights.
  • the risk factor monitoring module 606 is configured to monitor the risk factor, determine the activated first-level activation node in the social network map according to the risk factor, and the first-level activation node includes at least one associated node.
  • the risk delivery module 608 is configured to calculate, according to the social network graph, an accumulated weight of the first-level active node to its neighboring associated node, and determine whether the accumulated weight exceeds a preset activation threshold of the adjacent associated node to determine the second-level active node.
  • the evaluation result feedback module 610 is configured to determine, according to the first level activation node and the second level activation node, the next level activation node, until the cumulative weight of the activated association node is transmitted to the evaluation subject node, and the cumulative weight of the evaluation subject node is obtained, and further The risk assessment result of the evaluation subject node is obtained, and the risk assessment result is sent to the terminal.
  • the risk factor monitoring module 606 includes:
  • the hot word set obtaining module 702 is configured to obtain a hot word set configured in advance for the associated node in the social network graph.
  • the hot word public opinion collection module 704 is configured to collect hot word public opinion information according to the hot word set, wherein the hot word public opinion information is information capable of reflecting the trend dynamic of the associated node.
  • the first level activation node determining module 706 is configured to: when the trend dynamic change of the associated node indicated by the hot word public opinion information reaches a set threshold, the indicated associated node is activated, and the activated associated node is the first level active node. .
  • the risk factor monitoring module may further be replaced by a first level activation node specifying module 802, where the first level activation node specifying module 802 is configured to receive at least one associated node specified by the terminal.
  • the specified associated node is the first-level active node.
  • the risk assessment apparatus further includes: a risk propagation path map generation module 902, configured to record a sequence in which the associated nodes are activated by different risk factors; The risk propagation path map is generated in sequence, and the risk propagation path map is pushed to the terminal.
  • a risk propagation path map generation module 902 configured to record a sequence in which the associated nodes are activated by different risk factors; The risk propagation path map is generated in sequence, and the risk propagation path map is pushed to the terminal.
  • the risk assessment apparatus further includes: a focus monitoring module 904, configured to determine at least one key associated node in the associated node according to the risk propagation path map, and perform an activation state of the key associated node. Monitoring, when the monitored key associated node is activated, sends an alarm message to the terminal.
  • a focus monitoring module 904 configured to determine at least one key associated node in the associated node according to the risk propagation path map, and perform an activation state of the key associated node. Monitoring, when the monitored key associated node is activated, sends an alarm message to the terminal.
  • the network interface may be an Ethernet card or a wireless network card.
  • the above modules may be embedded in the hardware in the processor or in the memory in the server, or may be stored in the memory in the server, so that the processor calls the corresponding operations of the above modules.
  • the processor can be a central processing unit (CPU), a microprocessor, a microcontroller, or the like.
  • a server including a memory, a processor, and computer executable instructions stored on the memory and executable on the processor, the processor executing computer executable instructions
  • the following steps are implemented: receiving a risk assessment request sent by the terminal, extracting an evaluation subject identifier carried in the risk assessment request, and searching for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes the assessment entity identifier directly Or an indirectly associated associated node, an evaluation subject node corresponding to the evaluation subject identifier, and an edge set between the nodes, the edges of the edge set are directed edges with weights; the monitoring risk factor is based on the risk factor in the social network.
  • Determining the activated first-level activation node, the first-level activation node includes at least one association node; calculating the cumulative weight of the first-level activation node to its adjacent association node according to the social network graph, and determining whether the cumulative weight exceeds the adjacent association
  • the preset activation threshold of the node to determine the second-level activation node;
  • the first-level activation node and the second-level activation node determine the next-level activation node until the cumulative weight of the activated associated node is transmitted to the evaluation subject node, and the cumulative weight of the evaluation subject node is obtained, thereby obtaining the risk assessment result of the evaluation subject node.
  • Send the risk assessment results to the terminal Send the risk assessment results to the terminal.
  • the monitoring risk factor performed by the processor of the server determines the activated first-level activation node in the social network map according to the risk factor, and the first-level activation node includes at least one associated node, including:
  • the hot word lyric information is information that can reflect the trend dynamics of the associated nodes
  • the associated node that is indicated is activated, and the activated associated node is the first-level active node.
  • the processor of the server further performs the step of receiving at least one associated node specified by the terminal, the designated associated node being a first level active node.
  • the processor of the server further performs the steps of: recording a sequence in which the associated nodes are activated by different risk factors; generating a risk propagation path map according to the sequence in which the associated nodes are activated, and transmitting the risk The route map is pushed to the terminal.
  • the processor of the server further performs the steps of: determining at least one key associated node in the associated node according to the risk propagation path map, and monitoring an activation state of the key associated node, when monitoring that the key associated node is activated At the time, an alarm message is sent to the terminal.
  • the linear threshold model is used to determine the associated nodes that are activated step by step (affected by the risk) in the social network, and finally the risk is transmitted to the evaluation subject, and the risk is propagated dynamically.
  • the assessment, the assessment path is traceable, and the assessment is more efficient; and the association risk assessment is conducted through the social network map, and the assessment results are more reliable.
  • one or more non-volatile readable storage media storing computer-executable instructions are provided that, when executed by one or more processors, cause one or more processors The following steps are performed: receiving a risk assessment request sent by the terminal, extracting an evaluation subject identifier carried in the risk assessment request, and searching for a pre-generated social network map corresponding to the evaluation subject identifier, where the social network graph includes the assessment entity identifier directly or The indirectly associated associated node, the evaluation subject node corresponding to the evaluation subject identifier, and the edge set between the nodes, the edges in the edge set are all directed edges with weights; the monitoring risk factor is based on the risk factor in the social network diagram Determining the activated first-level activation node, the first-level activation node includes at least one association node; calculating the cumulative weight of the first-level activation node to its adjacent association node according to the social network graph, and determining whether the cumulative weight exceeds the adjacent association node Preset activation threshold to determine the second level activation no
  • the monitoring risk factor performed by the one or more processors determines the activated first level active node in the social network map according to the risk factor, and the first level activation node includes at least one associated node step
  • the method includes: acquiring a hot word set configured in advance for the associated node in the social network graph; the hot word public opinion information is information capable of reflecting the trend dynamic of the associated node; and the trend dynamic change of the associated node indicated by the hot word public opinion information reaches a set threshold And the associated node is activated, and the activated associated node is the first-level active node.
  • the one or more processors further perform the step of receiving at least one associated node specified by the terminal, the designated associated node being a first level active node.
  • the one or more processors further perform the steps of: recording a sequence in which the associated nodes that are determined by different risk factors are activated; generating a risk propagation path map according to the order in which the associated nodes are activated, and The risk propagation path map is pushed to the terminal.
  • the one or more processors further perform the steps of: determining at least one key associated node in the associated node according to the risk propagation path map, and monitoring an activation state of the key associated node, when monitoring the key associated node When activated, an alarm message is sent to the terminal.
  • the linear threshold model is used to determine the associated nodes that are activated step by step (affected by the risk) in the social network, and finally the risk is transmitted to the evaluation subject, and the risk is propagated dynamically.
  • the assessment, the assessment path is traceable, and the assessment is more efficient; and the association risk assessment is conducted through the social network map, and the assessment results are more reliable.
  • the program may be stored in a storage medium of a computer system and executed by at least one processor in the computer system to implement a process comprising an embodiment of the methods as described above.
  • the storage medium may be a magnetic disk, an optical disk, or a read-only storage memory (Read-Only) Memory, ROM) or Random Access Memory (RAM).

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Development Economics (AREA)
  • Marketing (AREA)
  • Educational Administration (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种风险评估方法，包括：接收终端发送的风险评估请求；查找社交网络图；监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点；根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点；根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体，得到评估主体的累计权重，进而得到评估主体风险评估结果，将风险评估结果发送至终端。

Description

风险评估方法、装置、服务器和存储介质
本申请要求于2017年05月10日提交中国专利局、申请号为201710326779.X、发明名称为“风险评估方法、装置、服务器及存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
【技术领域】
本发明涉及计算机技术领域，特别是涉及一种风险评估方法、装置、服务器和存储介质。
【背景技术】
风险评估是指在风险事件发生之前或之后（但还没有结束），该事件对想要评估的企业或者个人带来的可能的影响和风险的量化的评估。即，风险评估就是量化测评某一事件或事物带来的影响或损失的可能程度。
传统的风险评估常用的评估方法主要有：专家预测法，是指由多个专家组成专家组通过座谈会的方式进行预测讨论，这种方式不仅评估效率低，而且不能确保能够客观反映现实；LEC风险评价法（该方法用与系统风险有关的三种因素指标值的乘积来评价操作人员伤亡风险大小，这三种因素分别是：L（likelihood，事故发生的可能性)、E（exposure，人员暴露于危险环境中的频繁程度）和C（consequence，一旦发生事故可能造成的后果）），是一种危险定量计算方法。依据危险源辨识记录，定量计算每种危险源所带来的分先，确定出最大风险，列出清单下发，有计划的控制风险。LEC风险评估法在对危险等级的划分，一定程度上是凭借经验判断，应用时具有局限性，不能普遍适用。也就是，传统的风险评估方法多为经验性的静态决策评估方法，评估效率低，评估过程不具有追溯性。
【发明内容】
根据本申请公开的各种实施例，提供一种风险评估方法和装置、服务器和存储介质。
一种风险评估方法，所述方法包括：
接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；
根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
一种风险评估装置，所述装置包括：
评估请求模块，用于接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
社交网络图查找模块，用于查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
风险因子监控模块，用于监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
风险传递模块，用于根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；
评估结果反馈模块，用于根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
一种服务器，所述服务器包括存储器和处理器，所述存储器中储存有计算机可执行指令，所述指令被所述处理器执行时，使得所述处理器执行以下步骤：
接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；
根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
一个或多个存储有计算机可执行指令的非易失性可读存储介质，所述计算机可执行指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：
接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；
根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
本申请的一个或多个实施例的细节在下面的附图和描述中提出。本申请的其它特征、目的和优点将从说明书、附图以及权利要求书变得明显。
【附图说明】
图1为一个实施例中风险评估方法的应用环境图；
图2为一个实施例中服务器的内部结构示意图；
图3为一个实施例中风险评估方法的流程图；
图4为一个实施例中带权重社交网络图；
图5为一个实施例中确定第一级激活节点所涉及的流程图；
图6为另一个实施例中风险评估方法的流程图；
图7为一个实施例中生成风险传播路径图所涉及的流程图；
图8为另一个实施例中带权重社交网络图；
图9为一个实施例中风险评估装置的结构框图；
图10为一个实施例中风险因子监控模块的结构框图；
图11为另一个实施例中风险评估装置的结构框图；
图12为又一个实施例中风险评估装置的结构框图；
图13为再一个实施例中风险评估装置的结构框图。
【具体实施方式】
为了使本发明的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本发明进行进一步详细说明。应当理解，此处所描述的具体实施例仅仅用以解释本发明，并不用于限定本发明。
如图1所示，在一个实施例中，提供了一种风险评估方法的应用环境图，该应用环境图包括终端110和服务器120。终端110可通过网络与服务器120通信。终端110可以是智能手机、平板电脑、笔记本电脑、台式计算机中的至少一种，但并不局限于此。服务器120可以是独立的物理服务器，也可以是多个物理服务器构成的服务器集群。终端110向服务器120发送风险评估请求，该风险评估请求中指定想要评估的评估主体标识。服务器接收终端发送的风险评估请求，并查找包含该评估主体标识的社交网络图，该社交网络图还包括与评估主体直接或者间接相关联的关联节点，以及各节点（包括评估主体节点和关联节点）之间的关联权重。当检测到风险因子时，根据社交网络图确定该风险因子对各节点带来的影响权重，判断该影响权重是否达到受影响的节点的预设激活权重，若是，则确定第一级激活节点。再以第一级激活节点为风险因子确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体，最终，确定评估主体的风险评估结果。服务器将得到的风险评估结果反馈至终端。
如图2所示，在一个实施例中，提供了一种服务器120，该服务器120包括通过系统总线连接的处理器、非易失性可读存储介质、内存储器和网络接口。其中，该服务器120的非易失性可读存储介质存储有操作系统、数据库和至少一条计算机可执行指令。该计算机可执行指令被处理器执行时，可使得处理器执行一种如图3所示的风险评估方法。数据库用于存储数据，如存储该风险评估方法执行过程中涉及的社交网络图等数据。处理器用于提供计算和控制能力，支撑整个服务器120的运行。服务器中的内存储器为非易失性可读存储介质中的操作系统、数据库和计算机可执行指令提供高速缓存的运行环境。网络接口用于与终端110进行通信连接。本领域技术人员可以理解，图2中示出的服务器的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的服务器的限定，具体的服务器可以包括比图中所示更多或更少的部件，或者组合某些部件，或者具有不同的部件布置。
如图3所示，在一个实施例中，提供了一种风险评估方法，该方法以应用于如图2所示的服务器进行举例说明，具体包括如下步骤：
步骤S202：接收终端发送的风险评估请求，提取风险评估请求中携带的评估主体标识。
终端可通过于服务器对应的终端应用向服务器发送风险评估请求，并指定待评估的评估主体标识，这里的带评估主体标识可以是企业标识，也可以是个人用户标识。
在一个实施例中，服务器预先构建至少一个社交网络图，每个社交网络图中均包含有多个相互之间具有关联关系的节点，每个节点对应一个节点标识。终端在向服务器发送风险评估请求前，根据请求页面中给出的社交网络图标签列表，选定一个社交网络图。响应于对社交网络图的选定，终端将在请求页面中展开该社交网络图的节点标识列表，用户从节点标识列表中选定其中一个节点标识或者多个节点标识作为待评估的评估主体标识。终端在向服务器发送风险评估请求时，携带该选定的评估主体标识。
步骤S204：查找预先生成的与评估主体标识对应的社交网络图，其中，社交网络图中包括与评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，边集中的边为带权重的有向边。
服务器查找包括评估主体标识的社交网络图。这种确定对应社交网络图的查找方式适用于选定的评估主体标识存在于唯一的社交网络图中。即评估主体仅是一个社交网络图中的节点。
当服务器预先构建的多个社交网络图存在节点重叠时，终端在发送风险评估请求时，除需携带评估主体标识还需携带该评估主体对应的社交网络图标识，以使服务器能够查找到用于进行风险评估的社交网络数据。
本实施例中，社交网络图是根据各节点的关联关系数据生成的有向带权图谱。这里的关联关系可以是担保贷款关系，对应的关联节点为借款方、担保房和贷款方；还可以是股东类关联关系，如企业与企业法人；行业关联关系，如企业与企业所属行业之间的关联关系；或者其他关联关系，如资讯关联、电话关联、地址关联和产业链关联等。上述关联关系数据中的关联方即构成了社交网络图中的节点，关联关系构成了各节点之间的边集。在此基础上，再根据设定的权重配置规则，配置各节点之间的边的权重。
权重配置规则可以采用如下方法：股东类关联方，可以根据持股比例确定赋值规则；担保关联方，可以根据担保金额制定赋值规则；行业协会关联方，可以试用方式确定一个值；资讯关联方，可以根据企业内外部差异、影响程度和时效等加权赋值规则；上下游产业链关联方，可以根据行业对上下游依赖的平均程度选定赋值规则。
需要注意的是，每个节点包含的所有指向该节点的有向边的权重总和应该小于1。在按照上述赋值规则进行赋值后，对不满足每个节点包含的所有指向该节点的有向边的权重总和小于1这一条件的节点所对应的赋值进行调整。
如图4所示为构建的带权重的有向图的一个示例。该示例图中包括四个节点，节点N1与节点N2和节点N3具有关联关系，节点N1对节点N2的影响权重为0.7，节点N1对节点N3的影响权重为0.3；节点N2与节点N3能够相互影响，但影响权重有所不同；节点N2还与节点N4具有关联关系，节点N2对节点N4的影响权重为0.6；节点N3与节点N4具有关联关系，节点N2对节点N4的影响权重为0.4。
步骤S206：监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点。
本实施例中的风险因子是指能够为步骤S204查找的社交网络图中的各关联节点中的一个或者多个带来影响的风险因子。
风险因子可以是服务器监控分析大数据得到的情报信息。例如，社交网络图中的一个节点为石油行业，情报信息可以是油价上涨达设定幅度，某一区域内设定数量的石油企业宣布破产等等。风险因子还可以是用户假设的可能发生的风险，如社交网络图中的企业A资金短缺，企业B法人离职等。风险因子还可以是社交网络图中的其中一个关联节点，当检测到风险因子，则该风险因子对应的关联节点被激活。如监控到企业B法人离职，则企业B的法人节点被激活。
监控到风险因子后，查找与风险因子具有关联的节点，与风险因子具有关系的节点被激活，被激活的节点即为第一激活节点。如油价上涨达设定幅度，某一区域内设定数量的石油企业宣布破产，则石油行业的节点被激活。
步骤S208：根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点。
确定第一级激活节点后，通过社交网络图中各节点之间的关联关系以及配置的影响权重，判断第一级激活节点对邻接节点的影响。具体为，计算第一级激活节点对邻接关联节点的累计权重，并判断累计权重是否超过该节点的自身预设激活阈值，若是，则该邻接关联节点也被激活。
根据图4中的社交网络图进行举例说明。假设预先配置的节点N1的激活阈值是0.6，节点N2的激活阈值是0.5，节点N3的激活阈值是0.7，节点N4的激活阈值是0.9。当风险因子被激活的第一级激活节点为N1时，节点N2的累计权重即为节点N1对其的影响权重，为0.7，节点N2的累计权重大于节点N2激活权重，节点N2被激活。此时，单独的节点N1无法激活节点N3（0.3<0.7）。因此，第二级激活节点为N2。同样的原理，根据第一级激活节点和第二级激活节点确定下一级激活节点，也就是，当节点N1和节点N2被激活后，其邻接的关联节点N3的累计权重为0.3+0.5=0.8>0.3，因此，节点N3亦被激活，同理，N4节点在N2和N3被激活的情况下，N4也被激活。
步骤S210：根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体节点，得到评估主体节点的累计权重，进而得到评估主体节点的风险评估结果，将风险评估结果发送至终端。
基于社交网络图逐级分析风险的传递路径，直至风险传递至评估主体，计算评估主体的累计权重，将评估主体的累计权重与评估主体的预设激活阈值进行对比，若评估主体的累计权重大于其预设激活阈值，则得到激活主体被激活的风险评估结果，将得到的风险评估结果反馈至请求分析的终端。
本实施例，通过构建待评估评估主体的有向且带影响权重的社交网络图，根据引起风险的风险因子，利用线性阈值模型逐级确定社交网络中逐级被激活（受风险影响的）的关联节点，最终，确定传递到评估主体的风险程度，以确定评估主体是否受到影响进而得到对评估主体的风险评估结果，达到了以风险传播的方式动态进行风险的评估，评估路径具有可追溯性；且通过社交网络图进行了关联性风险评估，包括社会关系、信贷关系、行业归属关系等全面的分析和评估，评估结果更加可靠；基于线性阈值模型进行风险程度判断，评估效率更高。
在一个实施例中，如图5所示，步骤S206：监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点，具体包括如下步骤：
步骤S302：获取预先为社交网络图中的关联节点配置的热词集合。
服务器预先为社交网络图中至少一个关联节点配置热词，为关联节点配置的热词可以是一个，也可以是多个热词的集合。节点的热词实质上是与节点相关的被搜索频次高的词语的集合。如为节点“××钢铁”配置的一组热词集合为“铁矿石”、“钢铁企业”、“特钢”、“布袋除尘”和“钢企”；为节点“信贷银行”配置的热词集合为“不良资产”、“存贷比”、“股市行情”等。
本实施例中，配置的热词集合即为终端需要监控的风险因子。风险因子的风险指数通过热词集合对应的热词舆情信息来体现。
步骤S304：根据热词集合收集热词舆情信息，其中，热词舆情信息为能够反映关联节点趋势动态的信息。
根据为关联节点设置的热词或者热词集合在设定网站或者设定平台搜索热词舆情信息，如，铁矿石的价格上涨，特钢研究新突破等。这些热词舆情信息能够直接或者间接的反映对应关联节点的信用状态、经营状态。
步骤S306：当热词舆情信息所指示的关联节点的趋势动态变化达到设定阈值，则指示的关联节点被激活，被激活的关联节点为第一级激活节点。
对收集的热词舆情信息进行统计分析，如分析热词舆情指示的趋势动态变化，如铁矿石价格上涨设定倍数，铁矿石供应出现设定大的缺口或者研究的新物质占据了铁矿石50%以上的市场，则与铁矿石对应的“××钢铁”节点被激活。
在另一个实施例中，还可以将热词舆情作为一个节点，并配置热词舆情节点与对应的关联节点的影响权重，当热词舆情节点所指示的热词趋势变化达到设定阈值，则该热词舆情节点被激活，判断热词舆情节点对关联节点的影响权重是否超过该关联节点的预设激活阈值，若是，则该关联节点被激活，若否，单独的热词舆情节点不能激活该关联节点。
在又一个实施例中，将热词舆情作为一个节点，热词舆情对应设定数量的热词，当检测到设定数量的热词中全部热词或者设定比例的热词在设定平台的搜索热度达到设定阈值，则该热词舆情节点被激活。举例来说，M节点为热词舆情节点，假设入选M节点的关键词共5个，即M节点阈值为0.5。基于微博热点事件实时监控平台提供的功能，同时实时监控这五个关键词微博热度情况，如果这五个词中有三个超出了监控热点阈值，那么，M节点不被激活，如果同时5个节点都超过阈值，那么M节点被激活。
本实施例中，根据设置的热词，可自动监控可能影响节点的风险因子（如自动监控与节点相关的财经新闻资讯、社交平台舆情），可自动触发风险监控和风险预测，为实现信用风险的实时发现与反馈提供了强有力的保障。
在一个实施例中，如图6所示，提供了一种风险评估方法，该方法具体包括如下步骤：
步骤S402：接收终端发送的风险评估请求，提取风险评估请求中携带的评估主体标识。
步骤S404：查找预先生成的与评估主体标识对应的社交网络图，其中，社交网络图中包括与评估主体标直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，边集中的边均为带权重的有向边。
步骤S406：接收终端指定的至少一个关联节点，指定的关联节点为第一级激活节点。
步骤S408：根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点。
步骤S410：根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体节点，得到评估主体节点的累计权重，进而得到评估主体节点的风险评估结果，将风险评估结果发送至终端。
本实施例中，由终端用户指定激活节点，以评估指定节点被激活时评估主体的风险状态。通过用户指定激活节点，可预测任意节点被激活对评估主体带来的影响，风险预测更加灵活。
在一个实施例中，如图7所示，在对评估主体进行风险评估的同时，还记录风险传播的路径。具体的，风险评估方法还包括如下步骤：
步骤S502：记录不同的风险因子带来的关联节点被激活的先后顺序。
不同的风险因子，将激发不同的第一级激活节点，进而将产生自第一级激活节点到其他关联节点和评估主体的不同的激活顺序。随着检测到不同的风险因子，根据社交网络图将产生关联节点不同的激活顺序。
参考图8，当风险因子使N1节点首先被激活时，激活顺序为：N1-N2-（N3、N5）-N4。当风险因子使N2节点首先被激活时，激活顺序为：N2-N5-N4。该示例只能产生两种激活顺序。当根据实际环境建立比较复杂的社交网络图时，不同的风险因子，将产生多种节点激活顺序。
步骤S504：根据关联节点被激活的先后顺序生成风险传播路径图，并将生成的风险传播路径图推送至终端。
根据记录的关联节点被激活的先后顺序生成风险传播路径图，风险传播路径图包括多个分支，以表示不同的风险因子所产生的不同的节点激活顺序。将生成的风险传播路径图推送至终端显示，以从宏观的角度把握风险动态。
步骤S506：根据风险传播路径图在关联节点中确定至少一个关键关联节点，并对关键关联节点的激活状态进行监控，当监控到关键关联节点被激活时，向终端发送报警信息。
服务器将基于生成的传播路径图确定至少一个关键关联节点。其中，确定的关键关联节点可以是传播路径图中的分支数量之多的节点，也就是，激活节点数量最多的节点，还可以是直接导致评估主体节点被激活的节点。
在另一个实施例中，关键关联节点还可以由用户进行指定。即服务器将生成的传播路径图发送至终端后，接收终端发送的对关键关联节点的选择指令，提取指令中携带的用户选择的关键关联节点标识。
确定关键关联节点后，服务器将对能够激活关键关联节点的风险因子或者关键关联节点的热词舆情进行实时监测，以便及时把控关键关联节点的激活状态。当检测到关键关联节点被激活时，实时向终端发送警报信息，以便第一时间实施风险应对措施。此外，在监控资源有限的情况下，采取重点监控的方式在减少监控资源的基础上能够取得较好的监控效果。
在一个实施例中，如图9所示，提供了一种风险评估装置，该装置包括：
评估请求模块602，用于接收终端发送的风险评估请求，提取风险评估请求中携带的评估主体标识。
社交网络图查找模块604，用于查找预先生成的与评估主体标识对应的社交网络图，其中，社交网络图中包括与评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，边集中的边均为带权重的有向边。
风险因子监控模块606，用于监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点。
风险传递模块608，用于根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点。
评估结果反馈模块610，用于根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体节点，得到评估主体节点的累计权重，进而得到评估主体节点的风险评估结果，将风险评估结果发送至终端。
在一个实施例中，如图10所示，风险因子监控模块606，包括：
热词集合获取模块702，用于获取预先为社交网络图中的关联节点配置的热词集合。
热词舆情收集模块704，用于根据热词集合收集热词舆情信息，其中，热词舆情信息为能够反映关联节点趋势动态的信息。
第一级激活节点确定模块706，用于当热词舆情信息所指示的关联节点的趋势动态变化达到设定阈值，则所指示的关联节点被激活，被激活的关联节点为第一级激活节点。
在一个实施例中，如图11所示，所述风险因子监控模块还可以用第一级激活节点指定模块802替代，第一级激活节点指定模块802，用于接收终端指定的至少一个关联节点，指定的关联节点为第一级激活节点。
在一个实施例中，如图12所示，风险评估装置还包括：风险传播路径图生成模块902，用于记录不同的风险因子带来的关联节点被激活的先后顺序；根据关联节点被激活的先后顺序生成风险传播路径图，并将风险传播路径图推送至终端。
在一个实施例中，如图13所示，风险评估装置还包括：重点监控模块904，用于根据风险传播路径图在关联节点中确定至少一个关键关联节点，并对关键关联节点的激活状态进行监控，当监控到关键关联节点被激活时，向终端发送报警信息。
上述网络访问行为识别装置中的各个模块可全部或部分通过软件、硬件及其组合来实现。其中，网络接口可以是以太网卡或无线网卡等。上述各模块可以硬件形式内嵌于或独立于服务器中的处理器中，也可以以软件形式存储于服务器中的存储器中，以便于处理器调用执行以上各个模块对应的操作。该处理器可以为中央处理单元（CPU)、微处理器、单片机等。
在一个实施例中，提供了一种服务器，如图2所示，该服务器包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机可执行指令，处理器执行计算机可执行指令时实现以下步骤：接收终端发送的风险评估请求，提取风险评估请求中携带的评估主体标识；查找预先生成的与评估主体标识对应的社交网络图，其中，社交网络图中包括与评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，边集中的边均为带权重的有向边；监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点；根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点；根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体节点，得到评估主体节点的累计权重，进而得到评估主体节点风险评估结果，将风险评估结果发送至终端。
在一个实施例中，服务器的处理器所执行的监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点的步骤，包括：
获取预先为社交网络图中的关联节点配置的热词集合；
热词舆情信息为能够反映关联节点趋势动态的信息；
当热词舆情信息所指示的关联节点的趋势动态变化达到设定阈值，则与所指示的关联节点被激活，被激活的关联节点为第一级激活节点。
在一个实施例中，服务器的处理器还执行如下步骤：接收终端指定的至少一个的关联节点，指定的关联节点为第一级激活节点。
在一个实施例中，服务器的处理器还执行如下步骤：记录确定不同的风险因子带来的关联节点被激活的先后顺序；根据关联节点被激活的先后顺序生成风险传播路径图，并将风险传播路径图推送至终端。
在一个实施例中，服务器的处理器还执行如下步骤：根据风险传播路径图在关联节点中确定至少一个关键关联节点，并对关键关联节点的激活状态进行监控，当监控到关键关联节点被激活时，向终端发送报警信息。
本实施例中，利用线性阈值模型逐级确定社交网络中逐级被激活（受风险影响的）的关联节点，最终得到风险传递到评估主体的评估，实现了以风险传播的方式动态进行风险的评估，评估路径具有可追溯性，评估效率更高；且通过社交网络图进行了关联性风险评估，评估结果更加可靠。
在一个实施例中，提供了一个或多个存储有计算机可执行指令的非易失性可读存储介质，该计算机可执行指令被一个或多个处理器执行时，使得一个或多个处理器执行以下步骤：接收终端发送的风险评估请求，提取风险评估请求中携带的评估主体标识；查找预先生成的与评估主体标识对应的社交网络图，其中，社交网络图中包括与评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，边集中的边均为带权重的有向边；监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点；根据社交网络图计算第一级激活节点对其邻接关联节点的累计权重，并判断累计权重是否超过邻接关联节点的预设激活阈值以确定第二级激活节点；根据第一级激活节点和第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到评估主体节点，得到评估主体节点的累计权重，进而得到评估主体节点的风险评估结果，将风险评估结果发送至终端。
在一个实施例中，一个或多个处理器所执行的监控风险因子，根据风险因子在社交网络图中确定被激活的第一级激活节点，第一级激活节点至少包括一个关联节点的步骤，包括：获取预先为社交网络图中的关联节点配置的热词集合；热词舆情信息为能够反映关联节点趋势动态的信息；当热词舆情信息所指示的关联节点的趋势动态变化达到设定阈值，则与所指示的关联节点被激活，被激活的关联节点为第一级激活节点。
在一个实施例中，一个或多个处理器还执行如下步骤：接收终端指定的至少一个关联节点，指定的关联节点为第一级激活节点。
在一个实施例中，一个或多个处理器还执行如下步骤：记录确定不同的风险因子带来的关联节点被激活的先后顺序；根据关联节点被激活的先后顺序生成风险传播路径图，并将风险传播路径图推送至终端。
在一个实施例中，一个或多个处理器还执行如下步骤：根据风险传播路径图在关联节点中确定至少一个关键关联节点，并对关键关联节点的激活状态进行监控，当监控到关键关联节点被激活时，向终端发送报警信息。
本实施例中，利用线性阈值模型逐级确定社交网络中逐级被激活（受风险影响的）的关联节点，最终得到风险传递到评估主体的评估，实现了以风险传播的方式动态进行风险的评估，评估路径具有可追溯性，评估效率更高；且通过社交网络图进行了关联性风险评估，评估结果更加可靠。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，程序可存储于一计算机可读取存储介质中，如本发明实施例中，该程序可存储于计算机系统的存储介质中，并被该计算机系统中的至少一个处理器执行，以实现包括如上述各方法的实施例的流程。其中，存储介质可为磁碟、光盘、只读存储记忆体（Read-Only Memory，ROM）或随机存储记忆体（Random Access Memory，RAM）等。
以上实施例的各技术特征可以进行任意的组合，为使描述简洁，未对上述实施例中的各个技术特征所有可能的组合都进行描述，然而，只要这些技术特征的组合不存在矛盾，都应当认为是本说明书记载的范围。
以上实施例仅表达了本发明的几种实施方式，其描述较为具体和详细，但并不能因此而理解为对发明专利范围的限制。应当指出的是，对于本领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干变形和改进，这些都属于本发明的保护范围。因此，本发明专利的保护范围应以所附权利要求为准。

Claims (20)

  1. 一种风险评估方法，包括：
    接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
    查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
    监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
    根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；及
    根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
  2. 根据权利要求1所述的方法，其特征在于，所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点包括：
    获取预先为所述社交网络图中的关联节点配置的热词集合；
    根据所述热词集合收集热词舆情信息，其中，所述热词舆情信息为能够反映所述关联节点趋势动态的信息；
    当所述热词舆情信息所指示的所述关联节点的趋势动态变化达到设定阈值，则所指示的关联节点被激活，被激活的所述关联节点为第一级激活节点。
  3. 根据权利要求1所述的方法，其特征在于，所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点的步骤替换为：
    接收所述终端指定的至少一个所述关联节点，指定的所述关联节点为第一级激活节点。
  4. 根据权利要求1所述的方法，其特征在于，还包括：
    记录不同的所述风险因子带来的所述关联节点被激活的先后顺序；
    根据所述关联节点被激活的先后顺序生成风险传播路径图，并将生成的所述风险传播路径图推送至所述终端。
  5. 根据权利要求4所述的方法，其特征在于，还包括：
    根据所述风险传播路径图在所述关联节点中确定至少一个关键关联节点，并对所述关键关联节点的激活状态进行监控，当监控到所述关键关联节点被激活时，向所述终端发送报警信息。
  6. 一种风险评估装置，包括：
    评估请求模块，用于接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
    社交网络图查找模块，用于查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
    风险因子监控模块，用于监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
    风险传递模块，用于根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；及
    评估结果反馈模块，用于根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
  7. 根据权利要求6所述的装置，其特征在于，所述风险因子监控模块，包括：
    热词集合获取模块，用于获取预先为所述社交网络图中的关联节点配置的热词集合；
    热词舆情收集模块，用于根据所述热词集合收集热词舆情信息，其中，所述热词舆情信息为能够反映所述关联节点趋势动态的信息；
    第一级激活节点确定模块，用于当所述热词舆情信息所指示的所述关联节点的趋势动态变化达到设定阈值，则所指示的关联节点被激活，被激活的所述关联节点为第一级激活节点。
  8. 根据权利要求6所述的装置，其特征在于，还包括：第一级激活节点指定模块，用于接收所述终端指定的至少一个所述关联节点，指定的所述关联节点为第一级激活节点。
  9. 根据权利要求6所述的装置，其特征在于，还包括：风险传播路径图生成模块，用于记录不同的所述风险因子带来的所述关联节点被激活的先后顺序；根据所述关联节点被激活的先后顺序生成风险传播路径图，并将生成的所述风险传播路径图推送至所述终端。
  10. 根据权利要求9所述的装置，其特征在于，还包括：重点监控模块，用于根据所述风险传播路径图在所述关联节点中确定至少一个关键关联节点，并对所述关键关联节点的激活状态进行监控，当监控到所述关键关联节点被激活时，向所述终端发送报警信息。
  11. 一种服务器，所述服务器包括存储器和处理器，所述存储器中储存有计算机可执行指令，所述指令被所述处理器执行时，使得所述处理器执行以下步骤：
    接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
    查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
    监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
    根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；及
    根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
  12. 根据权利要求11所述的服务器，其特征在于，所述处理器执行的所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点包括：
    获取预先为所述社交网络图中的关联节点配置的热词集合；
    根据所述热词集合收集热词舆情信息，其中，所述热词舆情信息为能够反映所述关联节点趋势动态的信息；
    当所述热词舆情信息所指示的所述关联节点的趋势动态变化达到设定阈值，则所指示的关联节点被激活，被激活的所述关联节点为第一级激活节点。
  13. 根据权利要求11所述的服务器，其特征在于，所述处理器所执行的所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点替换为：
    接收所述终端指定的至少一个所述关联节点，指定的所述关联节点为第一级激活节点。
  14. 根据权利要求11所述的服务器，其特征在于，所述处理器还执行如下步骤：
    记录不同的所述风险因子带来的所述关联节点被激活的先后顺序；
    根据所述关联节点被激活的先后顺序生成风险传播路径图，并将生成的所述风险传播路径图推送至所述终端。
  15. 根据权利要求14所述的服务器，其特征在于，所述处理器还执行如下步骤：
    根据所述风险传播路径图在所述关联节点中确定至少一个关键关联节点，并对所述关键关联节点的激活状态进行监控，当监控到所述关键关联节点被激活时，向所述终端发送报警信息。
  16. 一个或多个存储有计算机可执行指令的非易失性可读存储介质，所述计算机可执行指令被一个或多个处理器执行时，使得所述一个或多个处理器执行以下步骤：
    接收终端发送的风险评估请求，提取所述风险评估请求中携带的评估主体标识；
    查找预先生成的与所述评估主体标识对应的社交网络图，其中，所述社交网络图中包括与所述评估主体标识直接或者间接相关联的关联节点、所述评估主体标识对应的评估主体节点以及各节点之间的边集，所述边集中的边均为带权重的有向边；
    监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点；
    根据社交网络图计算所述第一级激活节点对其邻接关联节点的累计权重，并判断所述累计权重是否超过所述邻接关联节点的预设激活阈值以确定第二级激活节点；及
    根据所述第一级激活节点和所述第二级激活节点确定下一级激活节点，直至激活的关联节点的累计权重传递到所述评估主体节点，得到所述评估主体节点的累计权重，进而得到所述评估主体节点的风险评估结果，将所述风险评估结果发送至所述终端。
  17. 根据权利要求16所述的非易失性可读存储介质，其特征在于，所述处理器执行的所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点包括：
    获取预先为所述社交网络图中的关联节点配置的热词集合；
    根据所述热词集合收集热词舆情信息，其中，所述热词舆情信息为能够反映所述关联节点趋势动态的信息；
    当所述热词舆情信息所指示的所述关联节点的趋势动态变化达到设定阈值，则所指示的关联节点被激活，被激活的所述关联节点为第一级激活节点。
  18. 根据权利要求16所述的非易失性可读存储介质，其特征在于，所述处理器所执行的所述监控风险因子，根据所述风险因子在所述社交网络图中确定被激活的第一级激活节点，所述第一级激活节点至少包括一个所述关联节点替换为：
    接收所述终端指定的至少一个所述关联节点，指定的所述关联节点为第一级激活节点。
  19. 根据权利要求16所述的非易失性可读存储介质，其特征在于，所述处理器还执行如下步骤：
    记录不同的所述风险因子带来的所述关联节点被激活的先后顺序；
    根据所述关联节点被激活的先后顺序生成风险传播路径图，并将生成的所述风险传播路径图推送至所述终端。
  20. 根据权利要求19所述的非易失性可读存储介质，其特征在于，所述处理器还执行如下步骤：
    根据所述风险传播路径图在所述关联节点中确定至少一个关键关联节点，并对所述关键关联节点的激活状态进行监控，当监控到所述关键关联节点被激活时，向所述终端发送报警信息。
PCT/CN2017/090575 2017-05-10 2017-06-28 风险评估方法、装置、服务器和存储介质 Ceased WO2018205371A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710326779.XA CN107239882A (zh) 2017-05-10 2017-05-10 风险评估方法、装置、计算机设备及存储介质
CN201710326779.X 2017-05-10

Publications (1)

Publication Number Publication Date
WO2018205371A1 true WO2018205371A1 (zh) 2018-11-15

Family

ID=59984943

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/090575 Ceased WO2018205371A1 (zh) 2017-05-10 2017-06-28 风险评估方法、装置、服务器和存储介质

Country Status (3)

Country Link
CN (1) CN107239882A (zh)
TW (1) TWI634492B (zh)
WO (1) WO2018205371A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111815187A (zh) * 2020-07-14 2020-10-23 北京航空航天大学 一种基于方向加权网络的制造过程风险评价方法
CN113689291A (zh) * 2021-09-22 2021-11-23 杭银消费金融股份有限公司 基于异常移动的反欺诈识别方法及系统

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110019547B (zh) * 2017-11-10 2021-07-16 平安普惠企业管理有限公司 获取客户间的关联关系的方法、装置、设备及介质
CN107909274B (zh) * 2017-11-17 2023-02-28 平安科技（深圳）有限公司 企业投资风险评估方法、装置及存储介质
CN110209826A (zh) * 2018-02-06 2019-09-06 武汉观图信息科技有限公司 一种面向银行风险控制的金融图谱构建与分析方法
CN108062639A (zh) * 2018-02-23 2018-05-22 大连火眼征信管理有限公司 一种风险传播模型及适用于该模型的算法
CN109087163B (zh) * 2018-07-06 2021-07-09 创新先进技术有限公司 信用评估的方法及装置
CN108985072A (zh) 2018-07-16 2018-12-11 北京百度网讯科技有限公司 操作防御方法、装置、设备及计算机可读介质
CN109189821A (zh) * 2018-08-01 2019-01-11 成都数联铭品科技有限公司 基于关联方指标体系的数据分析系统
CN109583620B (zh) * 2018-10-11 2024-03-01 平安科技（深圳）有限公司 企业潜在风险预警方法、装置、计算机设备和存储介质
CN109657906B (zh) * 2018-11-12 2023-09-05 平安科技（深圳）有限公司 谣言传播风险的分析方法及装置、存储介质、计算机设备
CN109949154B (zh) * 2018-12-17 2024-07-02 深圳平安综合金融服务有限公司 客户信息分类方法、装置、计算机设备和存储介质
CN109345158A (zh) * 2018-12-19 2019-02-15 重庆百行智能数据科技研究院有限公司 企业风险识别方法、装置和计算机可读存储介质
CN111626887B (zh) * 2019-02-27 2024-12-27 北京奇虎科技有限公司 一种社交关系评估方法及装置
CN110069923A (zh) * 2019-03-13 2019-07-30 咪咕文化科技有限公司 一种识别风险用户的方法及相关装置
CN110659799A (zh) * 2019-08-14 2020-01-07 深圳壹账通智能科技有限公司 基于关系网络的属性信息处理方法、装置、计算机设备和存储介质
CN112580916B (zh) * 2019-09-30 2024-05-28 深圳无域科技技术有限公司 数据评估方法、装置、计算机设备和存储介质
CN111724250A (zh) * 2020-06-29 2020-09-29 深圳壹账通智能科技有限公司 风险传播的确定方法、装置、计算机系统及可读存储介质
CN111815448B (zh) * 2020-07-09 2021-05-04 睿智合创(北京)科技有限公司 基于关联网络的申请单确定方法
CN112734270B (zh) * 2021-01-19 2024-01-23 中国科学院地理科学与资源研究所 一种能源风险传导的测度方法、系统及数据平台
TWI807319B (zh) * 2021-05-10 2023-07-01 中國信託商業銀行股份有限公司 借貸風險偵測方法及其運算裝置
CN114781952B (zh) * 2022-06-23 2022-09-02 济宁市任城区畜牧兽医事业发展中心(济宁市任城区动物疫病预防控制中心、济宁市任城区动物卫生检疫中心) 一种畜牧业流行病防控的风险预警方法
CN116503163A (zh) * 2023-06-21 2023-07-28 建信金融科技有限责任公司 业务风险识别方法、装置、计算机设备、存储介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090164431A1 (en) * 2007-12-10 2009-06-25 Sprylogics International Inc. Analysis, Inference, and Visualization of Social Networks
CN106408423A (zh) * 2016-11-25 2017-02-15 泰康保险集团股份有限公司 用于风险评估的方法、系统及构建风险评估系统的方法
CN106570753A (zh) * 2015-10-08 2017-04-19 平安科技（深圳）有限公司 对违约风险进行控制的方法及服务器

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8854199B2 (en) * 2009-01-26 2014-10-07 Lytx, Inc. Driver risk assessment system and method employing automated driver log
EP2645297B1 (en) * 2012-03-29 2019-05-08 Sony Computer Entertainment America LLC Securing mobile devices

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090164431A1 (en) * 2007-12-10 2009-06-25 Sprylogics International Inc. Analysis, Inference, and Visualization of Social Networks
CN106570753A (zh) * 2015-10-08 2017-04-19 平安科技（深圳）有限公司 对违约风险进行控制的方法及服务器
CN106408423A (zh) * 2016-11-25 2017-02-15 泰康保险集团股份有限公司 用于风险评估的方法、系统及构建风险评估系统的方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
WANG, BIAO: "Analysis of User Influence in Social Networks", CHINESE MASTER'S THESES, no. 4, 15 April 2014 (2014-04-15), pages 41 - 50 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111815187A (zh) * 2020-07-14 2020-10-23 北京航空航天大学 一种基于方向加权网络的制造过程风险评价方法
CN113689291A (zh) * 2021-09-22 2021-11-23 杭银消费金融股份有限公司 基于异常移动的反欺诈识别方法及系统
CN113689291B (zh) * 2021-09-22 2022-11-01 杭银消费金融股份有限公司 基于异常移动的反欺诈识别方法及系统

Also Published As

Publication number Publication date
CN107239882A (zh) 2017-10-10
TW201901539A (zh) 2019-01-01
TWI634492B (zh) 2018-09-01

Similar Documents

Publication Publication Date Title
WO2018205371A1 (zh) 风险评估方法、装置、服务器和存储介质
CN112465411B (zh) 一种风险预测方法、装置及设备
Jia et al. A practical approach to constructing a knowledge graph for cybersecurity
WO2020153688A1 (ko) 암호화폐 거래 분석 방법 및 시스템
Fu et al. Enabling personalized search over encrypted outsourced data with efficiency improvement
US8046374B1 (en) Automatic training of a database intrusion detection system
CN114330322A (zh) 一种基于深度学习的威胁情报信息抽取方法
Puthal et al. A secure big data stream analytics framework for disaster management on the cloud
WO2019196224A1 (zh) 制度信息处理方法、装置、计算机设备和存储介质
US10749791B2 (en) System for rerouting electronic data transmissions based on generated solution data models
WO2018149083A1 (zh) 业务数据流转方法、装置、存储介质和终端
CN115309913A (zh) 一种基于深度学习的财务数据风险识别方法及系统
CN106599273A (zh) 一种基于社交网络分析的微博诈骗团体挖掘方法
Kamara Restructuring the NSA metadata program
Liu et al. Multi-source detection based on neighborhood entropy in social networks
CN119271704A (zh) 一种大数据实时处理系统及方法
Lejun et al. Redundant data detection and deletion to meet privacy protection requirements in blockchain-based edge computing environment
CN117609372A (zh) 一种区块链的dapps的分布系统
CN106411704A (zh) 一种分布式垃圾短信识别方法
CN121509073A (zh) 一种基于人工智能的大数据风险预警与评估方法
CN114297223B (zh) 一种基于大数据的中小企业信息化服务平台
CN110175333A (zh) 一种证据指引方法、装置、设备及存储介质
CN120090863A (zh) 一种满足标签和跳数双约束的加密图最短路径查询方法
Hashemi et al. Data leakage via access patterns of sparse features in deep learning-based recommendation systems
CN111526053A (zh) 一种基于置信度的网络安全告警处理方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17909040

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17909040

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 13/03/2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17909040

Country of ref document: EP

Kind code of ref document: A1