CN112861980B - Calendar task table mining method based on big data and computer equipment - Google Patents
Calendar task table mining method based on big data and computer equipment Download PDFInfo
- Publication number
- CN112861980B CN112861980B CN202110194607.8A CN202110194607A CN112861980B CN 112861980 B CN112861980 B CN 112861980B CN 202110194607 A CN202110194607 A CN 202110194607A CN 112861980 B CN112861980 B CN 112861980B
- Authority
- CN
- China
- Prior art keywords
- calendar
- executed
- account
- target
- link
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Accounting & Taxation (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
Abstract
The application relates to the technical field of big data, and provides a calendar task table mining method, a calendar task table mining device, computer equipment and a computer readable storage medium based on big data. The calendar task list mining method based on big data comprises the steps of inputting calendar data of a target account into a trained calendar weight model for optimization link identification to obtain calendar link information to be optimized, wherein information in a target calendar pool is used for describing a corresponding relation between the target account and each calendar content in each calendar link to be optimized, so that a calendar set to be executed can be determined from the target calendar pool corresponding to the target account according to the calendar link information to be optimized, the calendar task list to be executed is output based on the calendar set to be executed, the content in the calendar task list is matched with the target account better, the mined content can be prevented from having characteristic skewness, and the effectiveness of a transaction list is improved. In addition, the method is also suitable for the technical field of block chains.
Description
Technical Field
The invention belongs to the technical field of big data, and particularly relates to a calendar task table mining method based on big data, a calendar task table mining device based on big data, computer equipment and a computer readable storage medium.
Background
With the wide use of big data analysis technology by various industries, no matter research and development of products, popularization of products, sales channels of products and the like, the support of the big data analysis technology is not opened for a long time.
However, in the conventional big data analysis means, feature extraction is mainly performed based on various information of a user, a user figure or a user tag of the user is generated, and operations such as classification and resource recommendation are performed based on the user figure or the user tag. For example, when a salesperson outputs a calendar task based on big data analysis, feature mining needs to be performed based on relevant information of the salesperson and performance data, where the relevant information may include: attribute information of the salesperson, such as a scholarly, age, sex, marital situation, etc., and performance data, such as sales volume of a certain type of product, etc. However, in the conventional feature mining, variables which have a small influence degree or are not related, such as the academic history of the salesperson, marital conditions, and the like, are easily introduced. Or, the excavated features are prone to be biased due to some abnormal data, for example, the sales data of a money-exploded product appearing in a certain period of time is higher and has a certain influence on the sales data of all salesmen, so that the excavated features are influenced by the sales data of the money-exploded product, and feature bias appears.
Disclosure of Invention
In view of this, embodiments of the present application provide a calendar task table mining method based on big data, a calendar task table mining device based on big data, a computer device, and a computer-readable storage medium, so as to solve the problem that the mined content has a characteristic bias in the existing data analysis scheme.
A first aspect of an embodiment of the present application provides a calendar task table mining method based on big data, including:
inputting calendar data of a target account into a trained calendar weight model for optimization link identification to obtain calendar link information to be optimized;
according to the calendar link information to be optimized, determining a calendar set to be executed from a target calendar pool corresponding to the target account; the target calendar pool is constructed based on the account type corresponding to the target account;
and outputting a calendar task table to be executed based on the calendar set to be executed.
A second aspect of the embodiments of the present application provides a calendar task table mining device based on big data, including:
the identification unit is used for inputting the calendar data of the target account into the trained calendar weight model for optimization link identification to obtain calendar link information to be optimized;
the determining unit is used for determining a calendar set to be executed from a target calendar pool corresponding to the target account according to the calendar link information to be optimized; the target calendar pool is constructed based on the account type corresponding to the target account;
and the output unit is used for outputting the calendar task table to be executed based on the calendar set to be executed.
A third aspect of embodiments of the present application provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the first aspect when executing the computer program.
A fourth aspect of embodiments of the present application provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the first aspect.
The calendar task table mining method, the calendar task table mining device, the computer equipment and the computer readable storage medium based on the big data have the following beneficial effects that:
in the embodiment of the application, the calendar data of the target account is input into the trained calendar weight model for optimization link recognition, because the calendar data of the target account represents the actual calendar content of the target account, the trained calendar weight model is used for optimization link recognition according to the calendar data of the target account, calendar links needing optimization can be determined, namely calendar link information to be optimized is obtained, because the information in the target calendar pool is used for describing the corresponding relation between the target account and the calendar content in each calendar link to be optimized, a calendar set to be executed can be determined from the target calendar pool corresponding to the target account according to the calendar link information to be optimized, so a calendar task table to be executed is output based on the calendar set to be executed, and the content in the calendar task table is matched with the target account better, the effectiveness of outputting the transaction list is improved, and the condition that the mined content has characteristic bias is avoided.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or the prior art descriptions 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 illustrating an implementation of a big data-based calendar task table mining method according to an embodiment of the present application;
FIG. 2 is a flowchart illustrating an implementation of step S11 in an embodiment of the present application;
FIG. 3 is a flowchart illustrating an implementation of a big data-based calendar task table mining method according to another embodiment of the present application;
fig. 4 is a block diagram illustrating a structure of a big data-based calendar task table mining device according to an embodiment of the present application;
fig. 5 is a block diagram of a computer device according to an embodiment of the present disclosure.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In the method for outputting the calendar task table based on big data provided by this embodiment, the execution subject is a server or a terminal, and specifically, the execution subject may be a server or a terminal configured with the method or the function, where the server may be any server in a server cluster. The server cluster may be a cluster composed of a plurality of servers, and a distributed system is constructed based on the server cluster, so that data sharing or data synchronization can be realized among the plurality of servers in the server cluster. On the basis, an object script file is configured to any server in the server cluster, and the object script file describes the output method of the calendar task table based on big data provided by the embodiment, so that the server configured with the object script file can execute each step in the method for outputting the calendar task table by executing the object script file. When the terminal executes the method for outputting the calendar task table, a corresponding application program may be installed on the terminal, and since the application program is configured with an object script file for describing each execution step of the output method of the calendar task table based on the big data, the terminal can execute the object script file through the application program, and further execute each step of the output method of the calendar task table based on the big data.
When the optimization is realized, the server or the terminal inputs the calendar data of the target account into the trained calendar weight model for optimization link recognition, because the calendar data of the target account represents the actual calendar content of the target account, the trained calendar weight model is utilized to perform the optimization link recognition according to the calendar data of the target account, calendar links needing to be optimized can be determined, namely calendar link information to be optimized is obtained, because the target calendar pool is constructed based on the account type corresponding to the target account, and the information in the target calendar pool is used for describing the corresponding relation between the target account and each calendar content in each calendar link to be optimized, a calendar set to be executed can be determined from the target calendar pool corresponding to the target account according to the calendar link information to be optimized, and a calendar task table to be executed is output based on the calendar set to be executed, the contents in the calendar task list are matched with the target account, the effectiveness of outputting the transaction list is improved, and the output efficiency of the calendar task is improved.
For example, the target account corresponds to a "salesperson", the calendar data is taken as the behavior data of the "salesperson" in the actual sales activity process as an example, the terminal or the server inputs the calendar data of the target account corresponding to the "salesperson" into the trained calendar weight model for performing optimization link identification, since the calendar data of the "salesperson" represents the actual calendar content of the "salesperson", the optimization link identification is performed according to the calendar data of the "salesperson" by using the trained calendar weight model, so that the calendar link needing to be optimized can be determined, that is, the information of the calendar link to be optimized corresponding to the "salesperson" is obtained, because the target calendar pool is constructed based on the account type corresponding to the target account of the "salesperson", and the information in the target calendar pool is used for describing the corresponding relationship between the "salesperson" and each calendar content in each calendar link to be optimized, the calendar content suitable for the salesperson to improve in each calendar link to be optimized is described, so that the calendar set to be executed can be determined from the target calendar pool corresponding to the salesperson according to the calendar link information to be optimized, and the terminal or the server outputs the calendar task table to be executed based on the calendar set to be executed, so that the content in the calendar task table is more matched with the salesperson, the effectiveness of outputting the transaction list is improved, and the output efficiency of the calendar task is further improved.
The calendar task table mining method based on big data provided by the embodiment is described in detail through a specific implementation manner.
Fig. 1 shows a flowchart of an implementation of a big data-based calendar task table mining method provided in an embodiment of the present application, which is detailed as follows:
s11: and inputting the calendar data of the target account into the trained calendar weight model for optimization link identification to obtain calendar link information to be optimized.
In step S11, there is a unique correspondence between the target account and the actual salesperson, i.e., different salespersons may correspond to different accounts. Here, the target account broadly refers to an account used by a salesperson who needs to output a calendar task form for him through a terminal or a server.
In implementation, the calendar data may include editing content and recording content, where the editing content is obtained by a salesperson through self-editing, and the recording content is obtained by recording according to actual operations of the salesperson, for example, after the salesperson actually executes a calendar, for example, after transmitting publicity information to a client for interaction, the salesperson edits the calendar content and resource index data in an account of the salesperson, and further obtains the editing content in the calendar data; after the salesperson completes the product transaction in the process of propaganda interaction with the customer, all information of the transaction is used as record content in the calendar data.
In other embodiments, the editing content and the recording content in the calendar data may be obtained by configuring a plurality of calendar links in an account corresponding to the salesperson, each calendar link corresponding to a plurality of calendar contents, and the user automatically checks the calendar contents as the editing content and fills the resource index data as the recording content to form the calendar data. For example, calendar links may include: the system comprises a calendar link, a database, a.
In all embodiments of the present application, the trained calendar weight model is used to describe a corresponding relationship between calendar data and calendar link information to be optimized. Because the calendar data is the content of the salesman's actual execution calendar identified by the target account and the corresponding resource index data thereof, and the resource index data can represent the beneficial effect generated by the actual execution calendar, the trained calendar weight model is utilized to perform optimization link identification on the content of the salesman's actual execution calendar and the resource index data. The optimization link identification refers to mining a link to be optimized based on the content of the actual calendar executed by the salesman, and the trained calendar weight model is used for describing the corresponding relationship between the calendar data and the link information of the calendar to be optimized, so that the trained calendar weight model is used for performing optimization link identification on the content of the actual calendar executed by the salesman and the resource index data, omission or deficiency of the salesman corresponding to the target account in the actual execution of the corresponding calendar content can be mined, and the obtained link information of the calendar to be optimized can be used for describing the link of the calendar which is corresponding to the content of the calendar required to be executed by the salesman corresponding to the target account in the future.
In implementation, the trained calendar weight model may be constructed and trained based on a known LR logistic regression model, actual executed calendar content and index data thereof in the calendar data are used as input information of the model, the trained calendar weight model is input, whether future index data of the target account will be promoted is predicted by using the trained calendar weight model, and finally, the probability representing the potential value is converted into a score to form an activity score of the target account, wherein the probability is higher to indicate that the probability of promotion is higher, that is, the space capable of being promoted and the calendar content capable of being executed in the future are more, and the probability and the activity score may be in direct proportion or in inverse proportion. Splitting calendar links corresponding to activity score of the target account to obtain a weight score corresponding to each calendar link, and judging whether the corresponding calendar link is a calendar link to be optimized or not based on the weight score to further obtain calendar link information to be optimized.
It can be understood that, in practical application, the trained calendar weight model may be constructed and trained based on a known LR logistic regression model, so those skilled in the art know that other models having the same function as the calendar weight model in this embodiment can be constructed according to actual requirements by using other model frames similar to the LR logistic regression model, and then the trained calendar weight model can be obtained by training the model frames by using the sample data set including the calendar data samples and the calendar link samples, so that how to train or construct the trained calendar weight model is not described herein.
Fig. 2 shows a flowchart of the implementation of step S11 in this embodiment. As shown in fig. 2, as an embodiment, the calendar data includes an actually executed calendar link of the target account and resource index data corresponding to the actually executed calendar link; step S11 specifically includes:
s111: determining a liftable probability value of the actually executed calendar link according to the resource index data by using the trained calendar weight model;
s112: converting the liftable probability value into an activity amount score according to a preset conversion strategy, wherein the activity amount score is used for representing the degree to be optimized of the actual calendar execution link;
s113: and obtaining calendar link information to be optimized based on the activity score.
In this embodiment, the calendar data includes actual execution calendar content of the salesperson, and index data corresponding to the actual execution calendar, where the actual execution calendar content is used to describe the transactions related to the sales activities that the salesperson has performed in the past; the index data is used to describe resource index data obtained by the salesperson after conducting a transaction related to the sales activity. As one example, the resource indicator data may include a number of customer information recorded, a number of content required, a number of product requirements, a product sales amount, a product revenue, and the like.
It should be noted that, in this embodiment, the preset conversion policy is used to describe the logical content for converting the liftable probability value into the activity score, where the preset conversion policy may be configured as a script file, or configured as another branch module of the model embedded in the calendar weight model, such as a score card model unit of LR logistic regression.
In the implementation process, the resource index data can be used as an input quantity X, the trained calendar weight model is input for binning processing, the resource index data in the calendar data is divided into 0-1 scalar interval, 1-3 scalar intervals or 3-5 scalar intervals, and then the index data is converted into discrete variables, specifically divided nodes can be tested by chi-square, and adjacent areas with the minimum chi-square value are combined together. After the evidence weight WOE value of each resource index data is calculated, the WOE value can be used as the input of the model instead of the input quantity X. Therefore, the nonlinearity of the model can be increased, the expression capability of the model is improved, and the calculation is easy.
As one example, through a score card model of LR logistic regression, it is predicted whether the performance of the target account is improving, and the model will output the probability of the agent's performance improvement. For example, the probability is converted into a score, which can be measured and calculated by the following formula;
score ═ a + B × ln (odds); wherein A is a basic coefficient; b is a proportionality coefficient; odds ═ p/(1-p); and p is the predicted probability output by the trained calendar weight model.
Further, by solving the equation: p0 ═ a + B × ln (odds); p0+ PDO + a + B + ln (2 + Odds); the basic coefficient a and the scaling coefficient B can be solved. In the above equation set, P0 is a preset initial score point; PDO represents the fraction of the desired boost when the probability is doubled; the specific score can be adjusted within a set score interval range according to actual conditions. Score scores may be calculated to indicate the Score for each agent's performance improvement. The score of the total score can be assigned to each variable X in turn, by the formula of logistic regression:
it can be calculated that:
ln(Odds)＝w0+w1*x1+w2*x2+…；Score＝A+B*(w0+w1*x1+w2*x2+…)。
in this embodiment, the total score is divided into scores of each index, a score corresponding to the resource index data of each calendar link of the target account is calculated, and the score is compared with a preset reference value, where the preset reference value may be an average score of all other accounts, and the calendar link with the largest difference and significantly lower than the average division condition is found out and used as the calendar link to be optimized.
S12: according to the calendar link information to be optimized, determining a calendar set to be executed from a target calendar pool corresponding to the target account; the target calendar pool is constructed based on the account type corresponding to the target account.
In step S12, the information in the target calendar pool is used to describe the correspondence between the link information to be optimized and the calendar set to be executed, and there is an association between the target calendar pool and the target account. The calendar set to be executed comprises at least one calendar to be executed.
In this embodiment, the calendar link information to be optimized describes a calendar link corresponding to calendar content that needs to be executed by a salesperson corresponding to a target account in the future, and a plurality of transaction contents are provided in the transaction link, and not all the transaction contents are matched with the target account, so that different target calendar pools can be configured for different accounts. Here, because there is a unique correspondence between the target account and the actual salesperson, that is, different salespersons correspond to different accounts, all the accounts can be classified according to the characteristics or attributes of the different salespersons, so as to distinguish the account types between the different accounts, and then configure different target calendar pools according to the different account types.
In practical application, when different target calendar pools are configured for accounts of different account types, all calendar contents in each calendar link can be correspondingly divided according to the classification result of the accounts, namely according to each classified account type, and calendar contents which are more matched with the account type are selected from the calendar links, so that the target calendar pool corresponding to the account type is formed.
For example, calendar links may include: the calendar comprises client accumulation and interactive development, wherein in the calendar link of client accumulation, the calendar content comprises client contact information checking, product information sending to the client and offline publicity activities; under the condition that the calendar link is 'interactive exploitation', the calendar contents are 'mail push', 'WeChat push', 'gift giving'; the target calendar pool corresponding to the account a may include: the calendar link is all calendar contents accumulated by the client, and the calendar link is interactive development of part calendar contents, namely 'checking client contact information', 'sending product information to the client', 'off-line propaganda activity' and 'WeChat pushing'; the target calendar pool corresponding to the account B may include: the calendar link is part of calendar contents accumulated by the client, and the calendar link is used for interactively developing all calendar contents, namely 'checking the contact way of the client', 'sending product information to the client', 'pushing mail' and 'pushing WeChat'.
As an example, step S12 may include:
determining a target calendar pool corresponding to the account type according to the account type corresponding to the target account; and determining a calendar set to be executed from the target calendar pool according to the calendar link information to be optimized.
In this embodiment, the account types are used to distinguish or identify the target calendar pools corresponding to the target accounts, where the account sets corresponding to all the sales staff may be classified in advance to obtain a plurality of account subsets, so that all the accounts in the account sets can be accurately classified, and then a corresponding target calendar pool is allocated to each account subset, so that when an applicable calendar is queried for the target account, all the unrecorded or unprocessed calendars do not need to be recommended, and only a more accurate target calendar needs to be matched from the target calendar pool.
It should be appreciated that, in implementation, since the classifying of the account set into a plurality of account subsets and the allocating of the corresponding target calendar pool to each account subset may be performed before step S12, after the allocating of the corresponding target calendar pool to each account subset, each time S12 is performed thereafter, the allocating of the corresponding target calendar pool to each account subset is not required until the reclassifying of the account set is performed.
Based on the above embodiment, as an embodiment of the present application, the calendar task table mining method based on big data further includes, before step S12, the following steps:
classifying the account set based on attribute information carried by each account in the account set to obtain a plurality of account subsets; wherein the plurality of account subsets correspond to a plurality of account types;
and selecting calendar samples from the calendar sample pool to construct a target calendar pool for each account subset.
In this embodiment, the attribute information is personal information of the salesperson corresponding to the account, such as age, sex, school calendar, and the like as features. The classification processing of the account set is performed based on the attribute information carried by each account, and may be performed by screening out a plurality of most significant or representative accounts based on the attribute information, and then by comparing similarities between the plurality of accounts and other accounts, further implementing the classification processing of the account set, and obtaining a plurality of account subsets.
In implementation, the method may first perform clustering operation on the account set based on the attribute information carried by each account in the account set by using a Kmeans + + clustering method, and then find out the most significant features of each class, such as young, female, and the like, by using the statistical values of the features. After key features are extracted for each account group, a plurality of accounts which best meet the features are screened out, and the rest other accounts can be calculated by a KNN algorithm to be most similar to the account in which account group, so that the accounts can be divided into the corresponding account groups. After the account set is classified, for each account subset, calendar data corresponding to each account is input as a factor, and a classification prediction model is used to predict the future resource index promotion probability of the account, for example, an XGBoost model, a LightGBM model, or the like is used to predict whether the performance of a salesman corresponding to the account will be promoted in the next three months. The importance of each calendar data can be directly output by utilizing a classification prediction model, and calendar contents with higher accumulated importance are selected to form a corresponding target calendar pool after the importance is ranked from high to low.
As a possible implementation manner of this embodiment, before the step of selecting, for each account subset, a calendar sample from the calendar sample pool to construct the target calendar pool, the method may further include: and constructing a calendar sample pool based on calendar data corresponding to each account in the account set.
In this embodiment, the calendar sample pool is constructed based on the calendar data corresponding to each account in the account set, so as to configure an accurate calendar sample pool for all known accounts. The calendar samples in the calendar sample pool are obtained by screening calendar contents actually recorded in all accounts, and the calendars which are more accordant with actual execution conditions or have wider application range and better application effect are determined as the calendar samples by screening all the calendar contents.
During implementation, the saturation, the relevance and the like of each calendar content in the calendar link can be measured and calculated according to the calendar data corresponding to each account. The saturation is used to describe the size of the execution range of the calendar contents, and the correlation is used to characterize the similarity between the calendar contents.
For example, 100 salespeople see how many salespeople have done an action, and if less than 10 people have done the action, the calendar content can be culled by the lower saturation.
As another example, calendar content 1: view customer contact, with calendar content 2: when the call is made, the agent basically checking the contact address also makes the call, the correlation coefficient calculated by the completion times of the two calendar contents of the 100 salespeople is very high, and if the correlation coefficient is higher than 80%, the repeated features are selected to be eliminated.
In this embodiment, since the calendar samples in the calendar sample pool are obtained by screening calendar contents actually recorded in all accounts, and all the calendar contents are screened to determine calendars which are more suitable for actual execution conditions or have wider application range and better application effect as the calendar samples, the calendar sample pool can be optimized before the target calendar pool is constructed for each account subset, so that a better sample selection environment is provided for selecting calendar samples from the calendar sample pool.
S13: and outputting a calendar task table to be executed based on the calendar set to be executed.
In step S13, the calendar set to be executed includes at least one calendar to be executed, which may be a calendar link or one or more calendar contents in the calendar link.
In this embodiment, the calendar task table to be executed is used for the data list displayed on the terminal or the server, and when the calendar set to be executed includes a plurality of calendar sets to be executed, the calendar sets to be executed may be sorted by comparing the difficulty level or the importance level of each calendar set to be executed, and the sorted result between the terminals or the servers is output as the calendar task table to be executed.
As an example, step S13 may include:
if the total number N of the calendars to be executed in the calendar set to be executed is greater than a preset threshold value, calling a sorting tool to sort the N calendars to be executed in the calendar set to be executed to obtain a sorted calendar set to be executed; wherein N is an integer greater than 2; and outputting the sorted calendar set to be executed as a calendar task table to be executed.
In this embodiment, the calendar sets to be executed that need to be output for the target account are sorted, and the calendar content that should be recommended with priority, is valid, and is most suitable for the salesperson corresponding to the target account is found. The sequencing tool can be a sequencing model constructed by adopting algorithms such as FM, FFM, Wide & Deep, DCN and the like.
During implementation, taking deep fm as an example, the deep fm algorithm effectively combines the advantages of a factorization machine and a neural network in feature learning: and simultaneously extracting the low-order combination features and the high-order combination features. In deep FM, an FM algorithm is responsible for extracting the first-order features and second-order features formed by pairwise combination of the first-order features; the DNN algorithm is responsible for extracting features of high-order features formed by fully connecting input first-order features and the like. The output of the final ranking model is also composed of two parts, and the specific representation is as the formula: y ═ sigmoid (yFM + yDNN);
wherein, FM part exports:
here, w is a weight coefficient for measuring the importance of the feature x, where x is associated data with the calendar to be executed in the calendar data of the target account, such as the number of times of executing the calendar to be executed, time, and the like. And in the DNN part, expanding the imbedding representation of the features into a full-join vector, and performing several layers of full-join calculation.
For example, 50 calendars to be executed are found from the target calendar pool as a calendar set to be executed, and the probability of the increase of the resource index data in the scene of the calendar set to be executed is predicted through a recommendation model such as deep fm. By sorting the probability values output by the 50 calendars to be executed of the target account, the top 5 calendars to be executed can be selected.
In the above scheme, the calendar data of the target account is input into the trained calendar weight model for optimization link recognition, and since the calendar data of the target account represents the actual calendar content of the target account, the trained calendar weight model is used for optimization link recognition according to the calendar data of the target account, so that calendar links needing optimization can be determined, that is, calendar link information to be optimized is obtained, since the information in the target calendar pool is used for describing the corresponding relationship between the target account and each calendar content in each calendar link to be optimized, a calendar set to be executed can be determined from the target calendar pool corresponding to the target account according to the calendar link information to be optimized, so that a calendar task table to be executed is output based on the calendar set to be executed, and the content in the calendar task table is more matched with the target account, the characteristic bias of the mined content is avoided.
In addition, by using the calendar task table mining method based on big data provided by the embodiment, the contents in the output calendar task table are more matched with the target account, so that the effectiveness of outputting the transaction list can be improved, and the output efficiency of the calendar task is improved.
Fig. 3 shows a flowchart of an implementation of a big data-based calendar task table mining method according to another embodiment of the present application. Referring to fig. 3, with respect to the embodiment shown in fig. 1, in the calendar task table mining method based on big data provided in this embodiment, after the step of outputting the calendar task table to be executed based on the calendar set to be executed, the method further includes: s21, detailed description is as follows:
in this embodiment, after the step of outputting the calendar task table to be executed based on the calendar set to be executed, the method further includes:
s21: and deploying the calendar task table to be executed into block link points.
In this embodiment, in order to share the calendar task table to be executed, the calendar task table to be executed is deployed into the blockchain, so as to prevent the content of the calendar task table to be executed from being tampered.
In all embodiments of the present application, deploying the calendar task table to be executed to the block link point enables the block link point to obtain the calendar task table to be executed by calling the calendar task table to be executed, that is, provide the target account with the future work plan of the corresponding salesman. Meanwhile, the safety and the fair transparency to the user of the calendar task list to be executed can be ensured. The blockchain referred to in this example is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, consensus mechanism, encryption algorithm, and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
In the above scheme, the calendar data of the target account is input into the trained calendar weight model for optimization link recognition, and since the calendar data of the target account represents the actual calendar content of the target account, the trained calendar weight model is used for optimization link recognition according to the calendar data of the target account, so that calendar links needing optimization can be determined, that is, calendar link information to be optimized is obtained, since the information in the target calendar pool is used for describing the corresponding relationship between the target account and each calendar content in each calendar link to be optimized, a calendar set to be executed can be determined from the target calendar pool corresponding to the target account according to the calendar link information to be optimized, so that a calendar task table to be executed is output based on the calendar set to be executed, and the content in the calendar task table is more matched with the target account, the characteristic bias of the mined content is avoided.
In addition, by using the calendar task table mining method based on big data provided by the embodiment, the contents in the output calendar task table are more matched with the target account, so that the effectiveness of outputting the transaction list can be improved, and the output efficiency of the calendar task is improved.
In addition, the calendar task table to be executed is deployed to the block chain node, so that the calendar task table to be executed can be acquired and used by other nodes with calling authority in the block chain, and the utilization rate of the calendar task table to be executed can be further improved.
Referring to fig. 4, fig. 4 is a block diagram illustrating a calendar task table mining device based on big data according to an embodiment of the present disclosure. The mobile terminal in this embodiment includes units for executing the steps in the embodiments corresponding to fig. 1 and fig. 3. Please refer to fig. 1 and fig. 3, and fig. 1 and fig. 3 for the corresponding embodiments. For convenience of explanation, only the portions related to the present embodiment are shown. Referring to fig. 4, the calendar task table mining device 40 includes: a recognition unit 41, a determination unit 42, and an output unit 43. Specifically, the method comprises the following steps:
the identification unit 41 is configured to input calendar data of the target account into the trained calendar weight model to perform optimization link identification, so as to obtain calendar link information to be optimized;
a determining unit 42, configured to determine, according to the calendar link information to be optimized, a calendar set to be executed from a target calendar pool corresponding to the target account; the target calendar pool is constructed based on the account type corresponding to the target account;
an output unit 43, configured to output the calendar task table to be executed based on the calendar set to be executed.
As an embodiment of the present application, the calendar data includes an actual calendar execution link of the target account and resource index data corresponding to the actual calendar execution link; the identifying unit 41 is specifically configured to determine, according to the resource index data, a liftable probability value of the actually executed calendar link by using the trained calendar weight model; converting the liftable probability value into an activity amount score according to a preset conversion strategy, wherein the activity amount score is used for representing the degree to be optimized of the actual calendar execution link; and obtaining calendar link information to be optimized based on the activity score.
As an embodiment of the present application, the determining unit 42 is specifically configured to determine, according to an account type corresponding to the target account, a target calendar pool corresponding to the account type; and determining a calendar set to be executed from the target calendar pool according to the calendar link information to be optimized.
As an embodiment of the present application, the output unit 43 is specifically configured to, if the total number N of calendars to be executed in the calendar set to be executed is greater than a preset threshold, invoke a sorting tool to sort the N calendars to be executed in the calendar set to be executed, so as to obtain a sorted calendar set to be executed; wherein N is an integer greater than 2; and outputting the sorted calendar set to be executed as a calendar task table to be executed.
As an embodiment of the present application, the calendar task table mining device 40 based on big data further includes:
the classification unit is used for classifying the account set based on the attribute information carried by each account in the account set to obtain a plurality of account subsets; wherein the plurality of account subsets correspond to a plurality of account types;
and the first construction unit is used for selecting the calendar samples from the calendar sample pool to construct a target calendar pool for each account subset.
As an embodiment of the present application, the calendar task table mining device 40 based on big data further includes:
and the second construction unit is used for constructing the calendar sample pool based on the calendar data corresponding to each account in the account set.
As an embodiment of the present application, the calendar task table mining device 40 based on big data further includes:
and the deployment unit 44 is used for deploying the calendar task table to be executed into the block link point.
It should be understood that, in the structural block diagram of the calendar task table mining device based on big data shown in fig. 4, each unit is used to execute each step in the embodiment corresponding to fig. 1 and 3, and for each step in the embodiment corresponding to fig. 1 and 3, the above embodiment is explained in detail, specifically please refer to the description in the embodiment corresponding to fig. 1 and 3 and fig. 1 and 3, and no further description is provided here.
Fig. 5 is a block diagram of a computer device according to an embodiment of the present disclosure. As shown in fig. 5, the computer apparatus 50 of this embodiment includes: a processor 51, a memory 52 and a computer program 53 stored in said memory 52 and executable on said processor 51, such as a program of a big data based calendar task table mining method. The processor 51, when executing the computer program 53, implements the steps in the embodiments of the big data-based calendar task table mining method described above, such as S11 to S13 shown in fig. 1, or S11 to S21 shown in fig. 3. Alternatively, when the processor 51 executes the computer program 53, the functions of the units in the embodiment corresponding to fig. 4, for example, the functions of the units 41 to 44 shown in fig. 4, are implemented, for which reference is specifically made to the relevant description in the embodiment corresponding to fig. 4, which is not repeated herein.
Illustratively, the computer program 53 may be divided into one or more units, which are stored in the memory 52 and executed by the processor 51 to accomplish the present application. The one or more units may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution of the computer program 53 in the computer device 50. For example, the computer program 53 may be divided into an identification unit, a determination unit, and an output unit, each unit functioning specifically as described above.
The turntable device may include, but is not limited to, a processor 51, a memory 52. Those skilled in the art will appreciate that fig. 5 is merely an example of a computer device 50 and is not intended to limit the computer device 50 and may include more or fewer components than shown, or some components may be combined, or different components, e.g., the turntable device may also include input output devices, network access devices, buses, etc.
The Processor 51 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 52 may be an internal storage unit of the computer device 50, such as a hard disk or a memory of the computer device 50. The memory 52 may also be an external storage device of the computer device 50, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, provided on the computer device 50. Further, the memory 52 may also include both internal storage units and external storage devices of the computer device 50. The memory 52 is used for storing the computer program and other programs and data required by the turntable device. The memory 52 may also be used to temporarily store data that has been output or is to be output.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; such modifications and substitutions do not substantially depart from the spirit and scope of the embodiments of the present application and are intended to be included within the scope of the present application.
Claims (9)
1. A calendar task table mining method based on big data is characterized by comprising the following steps:
inputting calendar data of a target account into a trained calendar weight model for optimization link identification to obtain calendar link information to be optimized; the calendar data comprises an actual calendar execution link of the target account and resource index data corresponding to the actual calendar execution link;
according to the calendar link information to be optimized, determining a calendar set to be executed from a target calendar pool corresponding to the target account; the target calendar pool is constructed based on the account type corresponding to the target account;
outputting a calendar task table to be executed based on the calendar set to be executed;
the method for inputting the calendar data of the target account into the trained calendar weight model for optimization link identification to obtain calendar link information to be optimized comprises the following steps:
determining a liftable probability value of the actually executed calendar link according to the resource index data by using the trained calendar weight model;
converting the liftable probability value into an activity amount score according to a preset conversion strategy, wherein the activity amount score is used for representing the degree to be optimized of the actual calendar execution link;
and obtaining calendar link information to be optimized based on the activity score.
2. The calendar task table mining method based on big data according to claim 1, wherein the determining a calendar set to be executed from a target calendar pool corresponding to the target account according to the calendar link information to be optimized includes:
determining a target calendar pool corresponding to the account type according to the account type corresponding to the target account;
and determining a calendar set to be executed from the target calendar pool according to the calendar link information to be optimized.
3. The big data-based calendar task table mining method according to claim 1, wherein outputting a calendar task table to be executed based on the calendar set to be executed comprises:
if the total number N of the calendars to be executed in the calendar set to be executed is greater than a preset threshold value, calling a sorting tool to sort the N calendars to be executed in the calendar set to be executed to obtain a sorted calendar set to be executed; wherein N is an integer greater than 2;
and outputting the sorted calendar set to be executed as a calendar task table to be executed.
4. The calendar task table mining method based on big data according to claim 1, wherein before the step of determining a calendar set to be executed from a target calendar pool corresponding to the target account according to the calendar link information to be optimized, the method further comprises:
classifying the account set based on attribute information carried by each account in the account set to obtain a plurality of account subsets; wherein the plurality of account subsets correspond to a plurality of account types;
and selecting calendar samples from the calendar sample pool to construct a target calendar pool for each account subset.
5. The big-data-based calendar task table mining method according to claim 4, further comprising, before the step of selecting a calendar sample from a calendar sample pool to construct a target calendar pool for each of the account subsets, respectively:
and constructing a calendar sample pool based on calendar data corresponding to each account in the account set.
6. The big data-based calendar task table mining method according to any one of claims 1 to 5, further comprising, after the step of outputting a calendar task table to be executed based on the calendar set to be executed:
and deploying the calendar task table to be executed into block link points.
7. A calendar task table mining device based on big data is characterized by comprising:
the identification unit is used for inputting the calendar data of the target account into the trained calendar weight model for optimization link identification to obtain calendar link information to be optimized; the calendar data comprises an actual calendar execution link of the target account and resource index data corresponding to the actual calendar execution link;
the determining unit is used for determining a calendar set to be executed from a target calendar pool corresponding to the target account according to the calendar link information to be optimized; the target calendar pool is constructed based on the account type corresponding to the target account;
the output unit is used for outputting a calendar task table to be executed based on the calendar set to be executed;
the identification unit is specifically configured to determine a liftable probability value of the actually executed calendar link according to the resource index data by using a trained calendar weight model; converting the liftable probability value into an activity amount score according to a preset conversion strategy, wherein the activity amount score is used for representing the degree to be optimized of the actual calendar execution link; and obtaining calendar link information to be optimized based on the activity score.
8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor executing the computer program to perform the steps of the method according to any one of claims 1 to 6.
9. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110194607.8A CN112861980B (en) | 2021-02-21 | 2021-02-21 | Calendar task table mining method based on big data and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110194607.8A CN112861980B (en) | 2021-02-21 | 2021-02-21 | Calendar task table mining method based on big data and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112861980A CN112861980A (en) | 2021-05-28 |
| CN112861980B true CN112861980B (en) | 2021-09-28 |
Family
ID=75988368
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110194607.8A Active CN112861980B (en) | 2021-02-21 | 2021-02-21 | Calendar task table mining method based on big data and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112861980B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113835847B (en) * | 2021-08-10 | 2023-11-24 | 复旦大学 | Transaction processing optimization method of distributed account book platform based on snapshot |
| CN113761521A (en) * | 2021-09-02 | 2021-12-07 | 恒安嘉新（北京）科技股份公司 | Script file detection method, device, equipment and storage medium based on machine learning |
| CN117252309B (en) * | 2023-11-15 | 2024-02-13 | 青岛海信信息科技股份有限公司 | Automatic timing execution method for office business process |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107451249A (en) * | 2017-07-28 | 2017-12-08 | 成都澳海川科技有限公司 | The Forecasting Methodology and device of event development trend |
| CN108475365A (en) * | 2015-12-30 | 2018-08-31 | 微软技术许可有限责任公司 | By management role classification and divide priority |
| CN109063920A (en) * | 2018-08-20 | 2018-12-21 | 阿里巴巴集团控股有限公司 | A kind of transaction risk recognition methods, device and computer equipment |
| CN110322093A (en) * | 2018-03-30 | 2019-10-11 | 阿里巴巴集团控股有限公司 | Information processing method, information display method, device and calculating equipment |
| CN111199440A (en) * | 2018-11-16 | 2020-05-26 | 北京嘀嘀无限科技发展有限公司 | Event prediction method and device and electronic equipment |
| CN111930805A (en) * | 2020-08-10 | 2020-11-13 | 中国平安人寿保险股份有限公司 | Information mining method and computer equipment |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101477656A (en) * | 2009-02-11 | 2009-07-08 | 浪潮集团山东通用软件有限公司 | Implementation process for scheduling production tasks in accordance with production capability gradually in production management process |
| CN106156257A (en) * | 2015-04-28 | 2016-11-23 | 北大方正集团有限公司 | A kind of Tendency Prediction method of microblogging public sentiment event |
| CN111291589B (en) * | 2018-12-06 | 2024-08-02 | 西安光启智能技术有限公司 | Information association analysis method and device, storage medium and electronic device |
| CN111583010A (en) * | 2019-02-18 | 2020-08-25 | 北京奇虎科技有限公司 | Data processing method, device, equipment and storage medium |
-
2021
- 2021-02-21 CN CN202110194607.8A patent/CN112861980B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108475365A (en) * | 2015-12-30 | 2018-08-31 | 微软技术许可有限责任公司 | By management role classification and divide priority |
| CN107451249A (en) * | 2017-07-28 | 2017-12-08 | 成都澳海川科技有限公司 | The Forecasting Methodology and device of event development trend |
| CN110322093A (en) * | 2018-03-30 | 2019-10-11 | 阿里巴巴集团控股有限公司 | Information processing method, information display method, device and calculating equipment |
| CN109063920A (en) * | 2018-08-20 | 2018-12-21 | 阿里巴巴集团控股有限公司 | A kind of transaction risk recognition methods, device and computer equipment |
| CN111199440A (en) * | 2018-11-16 | 2020-05-26 | 北京嘀嘀无限科技发展有限公司 | Event prediction method and device and electronic equipment |
| CN111930805A (en) * | 2020-08-10 | 2020-11-13 | 中国平安人寿保险股份有限公司 | Information mining method and computer equipment |
Non-Patent Citations (1)
| Title |
|---|
| 针对大规模网络访问数据的相关性及用户行为预测分析;李淑婷;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180315(第3期);正文全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112861980A (en) | 2021-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zohair et al. | Prediction of Student’s performance by modelling small dataset size | |
| CN112861980B (en) | Calendar task table mining method based on big data and computer equipment | |
| Altuntas et al. | Analysis of patent documents with weighted association rules | |
| CN113051291A (en) | Work order information processing method, device, equipment and storage medium | |
| WO2022127339A1 (en) | Website registration-based user portrait generating method and apparatus, device and medium | |
| CN113420018A (en) | User behavior data analysis method, device, equipment and storage medium | |
| CN113424207B (en) | System and method for efficiently training understandable models | |
| Enouen et al. | Sparse interaction additive networks via feature interaction detection and sparse selection | |
| CN114328277A (en) | Software defect prediction and quality analysis method, device, equipment and medium | |
| CN114519508A (en) | Credit risk assessment method based on time sequence deep learning and legal document information | |
| Lee et al. | Smart Robust Feature Selection (SoFt) for imbalanced and heterogeneous data | |
| CN113591881A (en) | Intention recognition method and device based on model fusion, electronic equipment and medium | |
| CN113327132A (en) | Multimedia recommendation method, device, equipment and storage medium | |
| CN113343102A (en) | Data recommendation method and device based on feature screening, electronic equipment and medium | |
| CN117314586A (en) | Product recommendation method, device, computer equipment and storage medium | |
| US20100169106A1 (en) | System and method for profiling jurors | |
| CN110619564B (en) | Anti-fraud feature generation method and device | |
| Mints | Classification of tasks of data mining and data processing in the economy | |
| CN112926989B (en) | Bank loan risk assessment method and equipment based on multi-view integrated learning | |
| CN115713424A (en) | Risk assessment method, risk assessment device, equipment and storage medium | |
| Bhuvaneswari et al. | Predicting periodical sales of products using a machine learning algorithm | |
| Mahalle et al. | Data Acquisition and Preparation | |
| CN113888265A (en) | Product recommendation method, device, equipment and computer-readable storage medium | |
| CN113379212A (en) | Block chain-based logistics information platform default risk assessment method, device, equipment and medium | |
| CN113435741A (en) | Training plan generation method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |