32results about How to "Meet actual business needs" patented technology

An article storage cabinet of fingerprint type is prepared as setting a numbers of doors on cabinet; arranging one ARM9 embedded computer host board in cabinet; connecting embedded computer host board to power supply, information-identifying module, infrared object sensing module, voice prompting module, liquid crystal display and multiple electromagnetic phase-lock; setting one infrared object sensing probe in front side of cabinet and erecting one or multiple infrared sensing probe separately in each door of cabinet.

This invention discloses a method for playing opera browser including: after a opera browser system sets up connection with an exchange system, it plays opera browser to a calling user, after the called user hangs up, the connection of the opera browser system and the switch system and the opera browser system continues playing the opera browser, in this invention, after the called user hangs up, the opera browser system continues playing the tones so as to secure that the calling user can listen to the tone completely, especially useful information.

The invention relates to the field of financial bill processing, and discloses a sequence segmentation method of a bill seal area in order to solve the problem that during OCR, the seal area is difficult to recognize or poor in recognition effect. The method comprises the steps that an ellipse detection method is utilized to quickly and effectively defect the seal area in an image so as to meet the instantaneity requirements for actual image recognition, and after the seal area is positioned, an irregular sequence of the seal area needs to be segmented. In the method, the area transformation method is adopted, and the target sequence in the seal area is transformed into a regular rectangular area. The method is suitable for the sequence segmentation of a financial bill.

The invention discloses a text generation method and device. According to the text generation method provided by the invention, the sample text and the business index corresponding to the sample textare trained in advance to obtain the extraction type text generation model and the generation type text generation model. During target text generation, the extraction type text generation model and the generation type text generation model are adopted to obtain a first target text candidate set and a second text candidate set, and the target text corresponding to the original text is seelected according to the prediction service index corresponding to each target text in the two candidate sets, so that the generated target text can meet the actual service requirement.

The invention discloses a data processing method and system, and relates to the technical field of computers. One specific embodiment of the method comprises the following steps that a configuration management center acquires configuration information generated for different types of data sources, and pushes the configuration information to a front-end processor; after the front-end processor encapsulates the configuration information into the configuration information entity class, data verification is carried out, and the configuration information in the configuration information entity class passing the verification is written into a local configuration file; after receiving the data extraction request, the front-mounted machine acquires a corresponding data source type and attribute information from the configuration information of the local configuration file according to a target data source identifier of the data extraction request; and the front-end processor loads a corresponding data engine according to the data source type so as to establish connection with the target data source based on the data engine and the attribute information, then acquires service data from thetarget data source, and returns the service data. According to the method, the hot deployment of the configuration information is realized, and the service data of the target data source can be flexibly and quickly acquired.

The invention provides a method and device for isolating a report application scene. The method includes the report task establishing step of establishing a report task and setting the basic properties of the report task; the report function setting step of setting report functions under the report task according to the type of the task and generating the rule of controlling the report functions corresponding to the service type; the report compiling system establishing step of establishing a report compiling system under the report task, namely, setting a unit system and a report template corresponding to the report task; the report task issuing step of issuing the report task to units at all levels and issuing reports through the units at all levels under the report task. A task management mode is adopted, the report task corresponding to the service type is set, the system automatically filters and displays unit information of the task and the report information of the task when a user conducts operation in the task, report data of all service organizations are effectively isolated, and the actual service requirements of a group enterprise for issuing group reports of different merge specifications at different time points are met.

The embodiment of the invention discloses a character recognition method, a character recognition device and a character recognition system. The method comprises the steps of obtaining a to-be-identified picture, wherein the to-be-identified picture at least comprises text content; inputting the picture to be recognized into a pre-constructed character recognition model, wherein the character recognition model is obtained by training based on a training sample of complex text information, the complex text information comprises information that at least one piece of attribute information of characters in the same text content has different values, and the attribute information at least comprises fonts, word sizes, colors and picture backgrounds; and obtaining a character recognition resultoutput by the character recognition model. According to the embodiment of the invention, the characters can be recognized accurately, efficiently and quickly, so that the working efficiency is improved. Meanwhile, the possibility of economic loss brought to customers and banks by human errors is greatly reduced.

The embodiment of the invention provides a service data processing method and device, electronic equipment and a computer storage medium, and the method comprises the steps: obtaining a business dataobtaining request of a client; based on the service type identifier and the service database, obtaining a node identifier and a node state identifier of each process node contained in the current service corresponding to the service type identifier; and sending the node identifier and the node state identifier of each process node contained in the current service to the client, so that the clientdisplays the node identifier and the node state identifier according to the service process of the current service. Scheme, the method comprises the following steps: acquiring a service data acquisition request of a client; according to the invention, the node identifier and the node state identifier of each process node included in the current service corresponding to the service type identifiercan be automatically obtained from the service database, so that a user can visually know the specific situation of each process node through the node identifier and the node state identifier, and theuser experience is improved.

The invention relates to an FMEA analysis method and system. The method comprises the following steps: determining a product structure and a current process route; dividing the devices in the product structure according to the functions of different devices in the product structure; dividing the structure into a virtual part, a black box part or a followed part; according to a product structure and a current process route, respectively determining a function tree of a virtual part and a function tree of a black box part; determining a failure tree of the virtual piece and a failure tree of the black box piece according to the function tree of the virtual piece and the function tree of the black box piece; determining the risk of completing the product according to the current process route and the risk reason according to the failure tree; and returning to the step of determining the product structure and the current process route, and updating the process route. According to the invention, the failure position and reason can be quickly positioned, and the product design efficiency and quality are improved.

The invention belongs to the technical field of IT (information technology) application, and discloses a method and a system for drug banning early warning research and judgment based on combination of report clues and multiple data sources, and the method for drug banning early warning research and judgment based on combination of report clues and multiple data sources comprises the following steps: collecting and summarizing clues; clue matching; early warning necessity analysis; early warning is initiated; determining a suspect trajectory; and early warning pushing. The anti-virus early warning research and judgment method based on the combination of the report clues and the multiple data sources is suitable for the field of anti-virus early warning research and judgment, data mining analysis and monitoring are carried out by means of informatization depending on the clues, early warning information is pushed to investigators once information changes of suspects are collected, and development and follow-up of subsequent capture actions are facilitated. Through a complete five-node process from clue acquisition, clue matching research and judgment, early warning initiation, early warning approval and early warning push, function intensification is realized, an information barrier is broken, multiple resources are integrated to meet actual service requirements, and early warning accuracy and timeliness are improved.

The invention relates to an information processing method and device, equipment and a medium. The information processing method comprises the following steps: acquiring target social entity information of a target social entity; extracting a target feature vector, related to registration, of the target social entity from the target social entity information; inputting the target feature vector into a pre-trained gradient boosting tree model to obtain a predicted credit score of the target social entity output by the pre-trained gradient boosting tree model; and if the predicted credit score of the target social entity meets a preset empty shell entity detection condition, determining that the target social entity is an empty shell entity. According to the embodiment of the invention, the accuracy of empty shell entity detection on the target social entity can be improved, so that the actual business requirements of departments such as tax inspection and risk control early warning can be met.

The invention relates to the technical field of information, and provides a control and management method and device for solving credit granting risks of small and micro enterprises. The method aims at solving the problems of data scarcity, incompleteness, insufficient historical information precipitation and the like in the credit business of the small and micro enterprises of a financial institution, obtaining the credit application requests of the small and micro enterprises, and forming a credit application data set Hive; requesting credit investigation mechanism data, performing cross validation on the credit investigation mechanism data and the credit application data set, and classifying customers according to verification information to obtain classified customer data; extracting classified customer data, and generating Hive of a training set and a test set after extracting customer application feature information; performing credit rating on the training set Hive by using an active learning technology to obtain a credit rating; after the credit rating is obtained, verifying the training set Hive through an asymmetric deviation analysis method; and issuing a credit line according to the verified credit rating result of the small and micro enterprises.

The invention relates to the technical field of informatization, provides an analytic hierarchy process and RFM fused commercial bank customer rating method and device, and aims to solve the problem of poor model rating effect caused by collinearity, single index and mismatching of actual business of RFM model variables in deposit business. The method mainly comprises the following implementation steps: determining a target layer and a criterion layer, wherein the target layer is used for grading a user, and the criterion layer is divided into three aspects of user activeness, user loyalty and user value; collecting user data sets in three aspects of user activeness, user loyalty and user value; carrying out data cleaning on the related data; respectively constructing scheme layer indexes of three criteria of user activity, user value and user loyalty by utilizing an RFM model; determining a scheme layer index weight based on an analytic hierarchy process, outputting a criterion layer index result, determining a criterion layer index weight, and outputting a score rating of a target layer customer; and based on the score rating of the target layer customer, finishing the formulation and popularization of the differentiated marketing strategy or activity.

The invention discloses an IEC61850 standard-based data online monitoring method, which comprises the following steps of: according to an IEC61850 standard, establishing an initial online monitoring model by taking an equipment type, a state monitoring type and a logic node of intelligent electronic equipment to be monitored as nodes; establishing a mapping relation corresponding to the unified state monitoring model to obtain a target online monitoring model, wherein the mapping relation comprises logic nodes of the online monitoring model, DOI nodes and remote units of the logic nodes respectively correspond to monitoring points of the unified state monitoring model and remote units associated with measurement and monitoring points; collecting monitoring data of the intelligent electronic equipment to be monitored, and matching the monitoring data with the logic nodes of the target online monitoring model to generate a corresponding data point file; and transmitting the data point file to a management information large area where the unified state monitoring model is located through cross-security-area transmission, and accessing the data point file periodically to the unified state monitoring model. The method provided by the invention meets the requirement of cross-regional data interaction, and is low in cost and easy to implement.

The invention relates to an invoice changing entity detection method and system. The method comprises the steps: acquiring the feature data of a plurality of entities, wherein the entities comprise to-be-detected entities and associated entities of the to-be-detected entities, and the feature data comprise invoicing features and attribute features; inputting the feature data into a pre-trained detection model to obtain invoice changing predictive scores of the to-be-detected entities output by the pre-trained detection model; and if the invoice changing predictive scores accord with a preset invoice changing entity detection condition, determining that the to-be-detected entities belong to invoice changing entities. According to the technical scheme, the trained detection model can be utilized, and the feature data of the to-be-detected entities and the associated entities are combined to carry out entity prediction with relatively high accuracy and robustness, so that the prediction accuracy of the invoice changing entity can be improved, and the actual business requirements of tax-related businesses such as tax inspection and collection and management can be further met.

The invention provides a stirrup calculation method and a stirrup calculation device for a horizontal haunched body of a frame beam. The method comprises the following steps of: obtaining a beam primitive and a column primitive of a target beam and a target column under a top view angle; determining whether a horizontal haunched body needs to be added between the target beam and the target column according to the position relationship between the beam primitive and the column primitive; under the condition that the horizontal haunched body needs to be added, position information of the horizontal haunched body is determined; calculating the reduced scale distance of the haunched body stirrups distributed on the horizontal haunched body according to a stirrup arrangement rule; the reduced scale distance is used for representing the vertical distance of the haunched body stirrup in the horizontal haunched body range; and calculating the reinforcement amount of the haunched body stirrup according to the reduced scale distance.

The invention relates to a scene resource construction method and a device, computer equipment and a storage medium. The method comprises the following steps: acquiring user data of each industry; respectively carrying out user characteristic analysis on the user data of each industry, and determining user characteristics of each industry; classifying the user data of each industry based on a preset user type classification standard to obtain user data of different user types in each industry; performing product attention point analysis on the user data of different user types in each industryto obtain product attention points of different user types in each industry; and based on the user characteristics of each industry and the product attention points of different user types in each industry, constructing different scene resources, and obtaining the scene resources of different user types in each industry. By adopting the method, corresponding scene resources can be compiled by pre-judging the product intention and chatting habit of each user type in each industry, so that the practice scene of the trainee can be more real, and the actual business requirement can be met.