CN111586100B - Target object message sending device and method - Google Patents
Target object message sending device and method Download PDFInfo
- Publication number
- CN111586100B CN111586100B CN202010254039.1A CN202010254039A CN111586100B CN 111586100 B CN111586100 B CN 111586100B CN 202010254039 A CN202010254039 A CN 202010254039A CN 111586100 B CN111586100 B CN 111586100B
- Authority
- CN
- China
- Prior art keywords
- value
- target object
- unit time
- determining
- transaction amount
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
-
- 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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Abstract
The embodiment of the invention relates to a target object message sending device and a method, wherein the device comprises: the data collection module is used for collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions conducted aiming at the target object; the statistical module is used for determining a preference value of a user for a target object according to the first historical operation behavior data; the measurement benchmark calculation module is used for determining a lower limit benchmark value according to the preference value, the target object transaction amount of each unit time in a plurality of unit times and all single transaction amounts; and the screening module screens the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value and sends the transaction messages to the front end. Through the mode, personalized pushing is achieved. It can also be advantageous to achieve the desired degree of user awareness of the target object without unduly disturbing the user.
Description
Technical Field
The embodiment of the invention relates to the technical field of computers, in particular to a target object message sending device and a target object message sending method.
Background
The traditional market has a simple calculation mode for pricing a large order and lacks a pricing mode for various stock trading characteristics. For example, the following two technical schemes are adopted:
the first technical scheme is as follows: aiming at the stock A, small bills, medium bills, large bills and extra large bills of the data are divided according to the amount of money; the small bill is less than 4 ten thousand yuan, the medium bill is 4 to 50 ten thousand yuan, the large bill is 50 to 100 ten thousand yuan, and the large bill is more than 100 ten thousand yuan.
The second technical scheme is as follows: aiming at the fact that any stock has volume of transaction within sixty minutes, the total number of transaction strokes within sixty minutes is not less than 800, the volume of transaction stroke by stroke within sixty minutes is taken as a calculated value, the mean value and the variance are calculated, the mean value plus 50 x variance is taken as a dividing line, the volume of transaction stroke by stroke in the next minute is calculated, if the mean value is larger than the dividing line, the main force is defined as single, and the dividing line is calculated once every minute.
The problem of the above two schemes is that as shown in fig. 1, fig. 1 shows that the large order in the user's stock is bought and sold, and according to the original pricing method, the user can see that many self-selected stocks are clearly seen, but the large order bought and sold sent on one trading day is concentrated on several stocks, and other stocks which trade infrequently do not have any large order trading records. Large orders are priced during the day, either horizontally (at the same time as the previous day) or vertically (within the day), with excessive fluctuations and less relevant relationships to the volume of trades, and are prone to missing large orders for certain stocks that are not actively traded. For example, a stock may not be active for a certain period of time, and the volume of transactions in 60 minutes is less than 800 pens, in which case even if a large order appears in 60 minutes, it will be ignored directly.
As shown in fig. 2, in the main power purchase and sale data, 300 main power purchases are made at a unit price of 280.15, 9100 main power purchases are made at a unit price of 280.12. That is to say, the amount of the stock transaction is more than 280.15 times 300 is the big order, and it is conceivable that the number of orders of the transaction amount between 280.15 times 300 to 280.12 times 9100 is large, all the orders are sent to the user, and then the sent messages received by the user every day may be hundreds, which greatly interferes with the normal life of the user.
The two modes are that information pushing is carried out by setting a certain fixed numerical value, the pushing mode is single, reasonable pushing cannot be carried out according to the actual condition of a client, the practicability is low, and the user experience is poor.
Disclosure of Invention
In view of the above, embodiments of the present invention provide a target object message sending apparatus and method to solve the above technical problem.
In a first aspect, an embodiment of the present invention provides a target object message sending apparatus, where the apparatus includes:
the data collection module is used for collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions conducted aiming at the target object;
the statistical module is used for receiving the first historical operation behavior data sent by the data collection module and determining the preference value of the user to the target object according to the first historical operation behavior data;
the measuring reference calculation module receives the preference value of the user to the target object, the target object transaction amount of each unit time in a plurality of unit times and all single transaction amounts for the target object, wherein the preference value is sent by the statistics module;
determining a lower limit reference value according to the preference value, the target object transaction amount of each unit time in a plurality of unit times and all single transaction amounts;
and the screening module screens the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value sent by the measurement reference calculation module, and sends the transaction messages to the front end.
In a possible implementation manner, the measurement reference calculation module is specifically configured to determine a first numerical value according to the preference value and the target object transaction amount of each unit time in the plurality of unit times;
extracting single transaction amounts ordered to a first value from the target object transactions in each unit time;
and determining the lower limit reference value according to all single transaction amounts which are sequenced to be the first numerical value in the unit time.
In one possible embodiment, the data collection module is further configured to collect second historical operation behavior data executed by the user for all operable objects including the target object;
and the measuring reference calculation module receives the second historical operation behavior data sent by the data collection module and corrects the first numerical value according to the second historical operation behavior data.
In one possible implementation, the measurement reference calculation module calculates a mean value and a standard deviation according to all single transaction amounts which are ranked as a first value in a plurality of unit times;
and determining the lower limit reference value according to all the single transaction amount, the average value and the standard deviation which are sequenced into the first numerical value.
In a possible implementation, the metric calculation module is specifically configured to:
calculating the kurtosis value and skewness value of normal distribution according to the single transaction amount, the mean value and the standard deviation which are sequenced into the first numerical value;
when the kurtosis value and the skewness value are determined to be zero, determining that all single transaction amounts sequenced into a first numerical value accord with standard normal distribution, and determining a lower limit reference value according to the mean value and the standard deviation;
or when the kurtosis value is determined to be not zero and the skewness value is zero, determining that all single transaction amounts sequenced into a first numerical value accord with the post-peak tail distribution, and determining a lower-limit reference value according to the mean value, the standard deviation and the kurtosis value;
or when the kurtosis value is determined to be zero, the skewness value is not zero, or the kurtosis value and the skewness value are not zero, determining that all single transaction amounts sequenced into a first numerical value accord with the post-spike distribution, determining a left quantile point alpha of a confidence interval by using an integral method, and determining a lower limit reference value according to the left quantile point alpha.
In a possible embodiment, the measurement reference calculation module is specifically configured to determine a lower-limit reference value according to a mean value, a standard deviation, and a coefficient n, where the coefficient n is a coefficient determined when the mean value and the standard deviation are normally distributed and a preset confidence level is reached;
or the measurement reference calculation module is specifically configured to obtain a coefficient n according to the kurtosis value adjustment, and determine the lower-limit reference value according to the mean value, the standard deviation, and the coefficient n.
In one possible embodiment, the apparatus further comprises: an acquisition module;
the acquisition module acquires a preset minimum reference value;
the measurement reference calculation module receives the lowest reference value sent by the acquisition module;
and when the lower limit reference value is determined to be smaller than the lowest reference value according to all the single transaction amounts which are sequenced to be the first numerical value in the unit time, determining the lowest reference value as the final lower limit reference value.
In one possible implementation, the metric calculation module is further configured to count a target object transaction amount of a first unit time in the plurality of unit times;
when the target object transaction amount in a first unit time in a plurality of unit times is larger than a preset threshold value, reducing the target object transaction amount in the first unit time according to a preset rule;
and sorting the single transaction amount of the reduced target object according to the transaction amount, wherein the first unit time is any one of a plurality of unit times.
In a second aspect, an embodiment of the present invention provides a target object message sending method, where the method includes:
collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions aiming at the target object;
determining a preference value of a user for a target object according to the first historical operation behavior data;
determining a lower limit reference value according to the preference value, the target object transaction amount of each unit time in a plurality of unit times and all single transaction amounts;
and screening the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value, and sending the transaction messages to the front end.
In a possible embodiment, the determining the lower limit reference value according to the preference value, the target object transaction amount of each unit time in the plurality of unit times, and all single transaction amounts specifically includes:
determining a first numerical value according to the preference value and the target object transaction amount of each unit time in the plurality of unit times;
extracting single transaction amounts ordered to a first value from the target object transactions in each unit time;
and determining the lower limit reference value according to all single transaction amounts which are sequenced to be the first numerical value in the unit time.
The target object message sending device provided by the embodiment of the invention has the following beneficial effects: the reference lower limit value determined by the embodiment is different from the mechanical means of 'one-time cutting' in the first prior art. The difference exists in consideration of the difference of the self-attribute of the target object. That is, a reference lower limit value is determined for each transaction of the target object. The difference from the second prior art solution is that the reference lower limit determined by the present solution does not need to consider whether certain transactions are active for a certain period of time, and naturally does not cause the reference lower limit to always fluctuate in the lateral and longitudinal directions. In addition, in the scheme of the application, when the reference lower limit value is determined, the factor of historical operation behavior data of the user on the target object is added, and the transaction amount of the target object in the actual situation and the single transaction amount of all transactions conducted on the target object are combined. Therefore, the degree of interest of the user for each target object can be determined (the degree of interest uses a lower limit reference value as an evaluation standard), and according to the degree of interest of the user for the target object, the transaction messages meeting the preset conditions are screened from the transaction messages in the subsequent unit time and sent to the front end. Namely, the message push standards for different target objects are respectively formulated for each user, so that personalized push is realized. The method can be more beneficial to achieving the expectation degree of the user for knowing the target object under the condition of not disturbing the user too much.
Drawings
FIG. 1 is a prior art screenshot of a stock manifest;
FIG. 2 is another prior art screenshot of a stock manifest;
fig. 3 is a schematic structural diagram of a target object message sending apparatus according to an embodiment of the present invention;
fig. 4 is a flowchart illustrating a method for sending a target object message according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
For the convenience of understanding of the embodiments of the present invention, the following description will be further explained with reference to specific embodiments, which are not to be construed as limiting the embodiments of the present invention.
Fig. 3 is a device for sending a target object message according to an embodiment of the present invention, where the device includes: a data collection module 310, a statistics module 320, a metric calculation module 330, and a screening module 340.
The data collection module 310 is used for collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions conducted aiming at the target object;
the statistical module 320 is used for receiving the first historical operation behavior data sent by the data collection module 310 and determining a preference value of the user for the target object according to the first historical operation behavior data;
the measurement reference calculation module 330 receives the user preference value for the target object sent by the statistics module 320, the target object transaction amount per unit time in the plurality of unit times sent by the data collection module 310, and all single transaction amounts for transaction for the target object;
determining a lower limit reference value according to the preference value, the target object transaction amount of each unit time in a plurality of unit times and all single transaction amounts;
the screening module 340 screens out the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value sent by the measurement reference calculation module 330, and sends the transaction messages to the front end.
Specifically, the target object may be an object that is interested by the public in a certain field, and the interest level of different users for the target object may be different. Such as stocks, bonds or other transaction objects in the financial field, for which different users are interested to a different extent. And the first historical operating behavior data may include some historical operating behaviors performed by the user with respect to the target object. For simplicity, the stock exchange will be described as an example in this embodiment.
The first historical actions performed by the user on the stock may include attention, query understanding, self-selection, purchase, sell, etc.
The plurality of unit times are manually set, and the unit time may be one hour, one day, one month, one quarter, or the like. Specifically, according to the actual situation, in the present embodiment, for example, 200 days are set among the plurality of unit times, and the unit time is 1 day. Then, the target trading volume per unit time in the unit time is the trading volume for a stock on each day of 200 days. The single transaction amount is the amount of each transaction of the stock.
The data collection module 310 collects the data and sends the data to the statistics module 320 and the metric calculation module 330, respectively.
Specifically, after receiving the first historical operation behavior data sent by the data collection module 310, the statistics module 320 determines a preference value of the user for the target object according to the first historical operation behavior data. The specific value of the preference value is determined according to an empirically set preset model, and is in one-to-one correspondence with the first historical operating behavior data. In a specific example, if the first historical operational behavior data collected by the data collection module 310 is that a user bought a stock, the statistic module 320 may determine the preference value to be 15 according to the first historical operational behavior data by using a preset model. If the first historical operational behavior data collected by the data collection module 310 is that the user pays attention to a certain stock, the statistic module 320 may determine the preference value to be 10 according to the first historical operational behavior data by using a preset model. The preference value is used to indicate that the device determines the number of pieces that can be used to push a stock exchange message to the user according to a preset model without other operations. The value range of the preference value is generally set to be within the range of 5-30.
The measurement reference calculation module 330 determines a lower reference value according to the preference value, the target object transaction amount in each unit time of the plurality of unit times, and the amount of all single transactions.
In an alternative embodiment, the metric calculation module 330 may first determine the first value according to the preference value and the target object transaction amount per unit time in each of the plurality of unit times.
Then, a single transaction amount ordered to a first value is extracted from the target object transaction for each unit time.
Finally, a lower limit benchmark value is determined based on all single-transaction amounts sorted into a first value in a plurality of unit times.
Optionally, the measurement reference calculating module 330 determines the first numerical value according to the preference value and the target object transaction amount in each unit time of the plurality of unit times, and the specific implementation process is as follows:
determining a value based on the target object transaction amount for each of the plurality of units of time;
the preference value is then compared to this value. One of the smaller values is selected as the first value.
For example, the preference value is 10, and the stock trading volume is 5 per single day. Then the first value per unit time is determined to be 5. Alternatively, the preference value is 10, and the daily transaction amount is greater than 10, then the first value is selected to be 10.
Of course, the above-described case is preferable. But in reality, the stock trading volume is very unstable. For example, the first day is 10 transactions and the second day is 100 transactions. This first value cannot be determined according to the method described above. Therefore, when a plurality of unit times are selected, the unit times can be obtained according to the preset model. For example, the criteria for selecting the unit time entered in the device is that the transaction amount must not fall below a certain value. For example, the number of transactions per unit time cannot be less than 100. The criterion may also be that the maximum value of all preference values set cannot be undershot. For example, the first historical operation behavior data collected by the data collection module 310 is that the user pays attention to a certain stock, and the preference value determined by the statistical module 320 by using the preset model is 5; the first historical operation behavior data collected by the data collection module 310 is that a user purchases a certain stock, and the preference value determined by the statistical module by using a preset model is 10; then, the data collection module 310 may follow the rule that the transaction amount per unit time is not lower than 10 at the lowest when selecting a plurality of unit times. In this way, no matter what the first historical operation behavior is, the corresponding number of the pieces is less than or equal to the transaction number per unit time. In this way, the first value selected in each unit time can be unified into the preference value determined according to the first historical operation behavior data. Of course, if this is the case, the 200 days described above may not be consecutive 200 days, but 200 days purposefully selected according to the predetermined model.
After the stock transaction of each single day is sorted according to the transaction amount, the transaction amount sorted into the first numerical value is extracted. If the plurality of unit time amounts is 200 days, each day is a unit time, then the single transaction amount ordered as the first value includes 200.
Finally, the lower limit reference value is calculated based on the 200 transaction amounts.
Alternatively, in one case, if the total actual trading amount of the target trading object (stock) per unit time is less than a preset value (for example, 1000), the number of stock trading messages sent for the user in advance is the total actual trading amount. For example, if the actual trading volume is 5, the number of stock trading messages to be sent to the user is determined to be 5.
In another case, if the total number of actual trades of the target object (stock) is greater than or equal to a preset value (e.g., 1000), a certain process is required for the total number of actual trades. The purpose of this is to make the subsequent calculation result more close to the user's expectation.
For example, the number of trading messages of a certain stock in one day may be 10 ten thousand, and if the actual total trading number of the stock is directly used as the number of stock trading messages which are sent for the user in advance, after the first value is determined subsequently, 10 ten thousand stock trading data are all sorted according to the trading amount. And then the transaction amount corresponding to the first numerical value is extracted and sorted. This transaction amount may be a large transaction amount, and there may be some deviation from the subsequent determination of the large single transaction amount.
Therefore, further optionally, the metric calculating module 330 is further configured to count the target object transaction amount of the first unit time in the plurality of unit times;
when the target object transaction amount in a first unit time in a plurality of unit times is larger than a preset threshold value, reducing the target object transaction amount in the first unit time according to a preset rule;
and sorting the single transaction amount of the reduced target object according to the transaction amount, wherein the first unit time is any one of a plurality of unit times.
Specifically, a large number of transactions are quantitatively reduced, for example, using a function. The specific degree of reduction is set empirically based on a large amount of experimental data. For example, the reduced value is set to 6. Then the log10 is used as the base log for the rank of all transaction amounts, and the final extracted data is actually the data for the original transaction amounts ranked 1, 10, 100, 1000, 10000 and 100000.
Or, the reduced value is determined to be set as other values according to a large amount of experimental data, so that the data can be reduced in an interval value-taking mode, and then the reduced stock transactions are sorted according to the transaction amount. The specific reduction of the transaction amount according to which preset rule may be set according to the actual situation, which is not necessarily the case.
Further optionally, in order to further improve the accuracy of the first numerical value, the data collection module 310 is further configured to collect second historical operation behavior data executed by the user for all operable objects including the target object;
the metric calculation module 330 receives the second historical operation behavior data sent by the data collection module 310, and corrects the first value according to the second historical operation behavior data.
Specifically, the second historical operating behavior data including the first historical operating behavior data may be obtained by crawler capture means. The acquired data may include: a user has queried about a stock over the network, has asked questions about the stock, has paid attention to the stock, has often bought or sold the stock, and so on. And determining a coefficient according to the second historical operating behavior data, wherein the specific coefficient is also set by experience obtained by a large amount of experimental data, is stored in the device and has a certain mapping relation with the second historical operating behavior data. For example, if the metric calculation module 330 counts that the number of times that the user bought a stock is greater than a certain number, for example, greater than 5 times, based on the second historical operating behavior data collected by the data collection module 310, the coefficient may be set to 1.5 according to the criterion that the number of times is greater than the certain number. The first value is multiplied by a coefficient as a final correction result of the first value. Subsequently, other operations are also performed based on the correction result.
Optionally, the metric calculating module 330 is specifically configured to calculate a mean and a standard deviation according to all single transaction amounts ranked as the first numerical value in the plurality of unit times.
And determining the lower limit reference value according to all the single transaction amount, the average value and the standard deviation which are sequenced into the first numerical value.
Specifically, the kurtosis value and the skewness value of the normal distribution are calculated according to all single transaction amounts, the mean value and the standard deviation which are sorted into the first numerical value. The kurtosis is also called as a kurtosis coefficient and represents the characteristic number of the probability density distribution curve of the peak value height at the average value. Skewness is expressed as bs, which is the average of n measurements of a sample, and represents the characteristic number of the degree of asymmetry of the probability distribution density curve with respect to the average. The methods for calculating the kurtosis value and the skewness value are all the prior art, and are not described in detail here. In this embodiment, the peakedness value and the skewness value are calculated because ideally, all the single-transaction amounts sorted into the first value should conform to the standard normal distribution, and further, the reference lower limit value is calculated in a manner of conforming to the standard normal distribution. In practice, however, this is likely not to be the case. And whether the specific measurement accords with the standard normal distribution or not can be determined according to the kurtosis value and the skewness value, and when all the single-transaction money ordered into the first numerical value does not accord with the standard normal distribution, the reference lower limit value is calculated in a preset mode. See below for details:
in the first case, when the metric calculation module 330 determines that the kurtosis value and the skewness value are both zero, it is determined that all the single-transaction amounts sorted into the first numerical value conform to the standard normal distribution, and the metric calculation module 330 determines the lower-limit reference value according to the mean value and the standard deviation.
Specifically, the metric calculating module 330 may determine the lower-limit reference value according to a mean, a standard deviation, and a coefficient n, where n is a coefficient determined when the mean and the standard deviation are normally distributed and a preset confidence level (e.g., 95%) is reached.
In the second case, when the metric calculation module 330 determines that the kurtosis value is not zero and the skewness value is zero, it determines that all the single-transaction amounts sorted into the first value conform to the post-spike distribution, and the metric calculation module 330 determines the lower-limit metric value according to the mean value, the standard deviation and the kurtosis value.
Specifically, the measurement reference calculation module 330 obtains a coefficient n according to the kurtosis value adjustment, and determines a lower-limit reference value according to the mean value, the standard deviation, and the coefficient n.
Or, when the measurement reference calculation module 330 determines that the kurtosis value is zero, the skewness value is not zero, or both the kurtosis value and the skewness value are not zero, it determines that all the single-transaction amounts ordered as the first numerical value conform to the post-spike tail distribution, determines the left quantile α of the confidence interval by using an integral method, and determines the lower limit reference value according to the left quantile α. In one specific example, such as 200 days out of a plurality of unit times. Then there are 200 transaction amounts ordered as the first value. For example, all the unit times correspond to the first value of 5. Then, there are 200 stock trading amounts ranked 5 th per day.
The mean μ and standard deviation σ of the 200 rank 5 stock trading amounts were calculated. And then respectively calculating the peak value and the deviation value according to the 200 transaction amounts, the mean value mu and the standard deviation sigma.
When the measurement reference calculation module 330 determines that the kurtosis value and the skewness value are both zero, all single transaction amounts ordered as the first numerical value are determined to be in accordance with the standard normal distribution, and then a coefficient n is obtained according to the steps. And the coefficient n is a coefficient determined by performing normal distribution by using the mean value and the standard deviation and reaching the preset confidence level. For example, according to the above method, the coefficient n is obtained as 2, and then the reference lower limit value is μ -2 σ.
Or, when the measurement reference calculating module 330 determines that the kurtosis value is not zero and the skewness value is zero, the measurement reference calculating module 330 adjusts the kurtosis value to obtain the coefficient n. And determining a lower limit reference value according to the mean value, the standard deviation and the coefficient n.
Or, when the measurement reference calculation module 330 determines that the kurtosis value is zero, the skewness value is not zero, or both the kurtosis value and the skewness value are not zero, the measurement reference calculation module 330 determines the left quantile α of the confidence interval by using an integration method, and determines the lower limit reference value according to the left quantile α.
The lower limit reference value determined by the normal distribution mode is more stable compared with the prior art. After the lower limit reference value is determined in this way, the target object transaction message is screened with the lower limit reference value as a measure and sent to the front end. The pushed messages are closer to the transaction messages expected by the user, the number of the pushed messages is more consistent with the expected degree of the user, the situation that the user is disturbed by too many messages is avoided, and the situation that the user knows little about the target object which is interested in the transaction messages due to too few pushed messages is avoided.
Further optionally, since the data is obtained by the measurement reference calculation module 330 through big data statistics, calculation, and the like, special situations are considered in actual situations, for example, the calculated reference lower limit value is too small. The device therefore also comprises: an acquisition module 350;
the obtaining module 350 obtains a preset minimum reference value;
the measurement reference calculation module 330 receives the lowest reference value sent by the acquisition module 350;
and when the lower limit reference value is determined to be smaller than the lowest reference value according to all the single transaction amounts which are sequenced to be the first numerical value in the unit time, determining the lowest reference value as the final lower limit reference value.
The target object message sending device provided by the embodiment of the invention has the following beneficial effects: the reference lower limit value determined by the embodiment is different from the mechanical means of 'one-time cutting' in the first prior art. The difference exists in consideration of the difference of the self-attribute of the target object. That is, a reference lower limit value is determined for each transaction of the target object. The difference from the second prior art solution is that the reference lower limit determined by the present solution does not need to consider whether certain transactions are active for a certain period of time, and naturally does not cause the reference lower limit to always fluctuate in the lateral and longitudinal directions. In addition, in the scheme of the application, when the reference lower limit value is determined, the factor of historical operation behavior data of the user on the target object is added, and the transaction amount of the target object in the actual situation and the single transaction amount of all transactions conducted on the target object are combined. Therefore, the degree of interest of the user for each target object can be determined (the degree of interest uses a lower limit reference value as an evaluation standard), and according to the degree of interest of the user for the target object, the transaction messages meeting the preset conditions are screened from the transaction messages in the subsequent unit time and sent to the front end. Namely, the message push standards for different target objects are respectively formulated for each user, so that personalized push is realized. The method can be more beneficial to achieving the expectation degree of the user for knowing the target object under the condition of not disturbing the user too much.
Fig. 4 is a schematic flowchart of a target object message sending method according to an embodiment of the present invention, and as shown in fig. 4, the method includes:
And step 420, determining a preference value of the user for the target object according to the first historical operation behavior data.
And step 440, screening the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value, and sending the transaction messages to the front end.
Optionally, determining the lower limit reference value according to the preference value, the target object transaction amount in each unit time of the plurality of unit times, and all single transaction amounts, specifically including:
determining a first numerical value according to the preference value and the target object transaction amount of each unit time in the plurality of unit times;
extracting single transaction amounts ordered to a first value from the target object transactions in each unit time;
and determining the lower limit reference value according to all single transaction amounts which are sequenced to be the first numerical value in the unit time.
Further optionally, after determining the first value according to the preference value and the target object transaction amount per unit time in each of the plurality of unit times, the method may further include:
and collecting second historical operation behavior data executed by the user aiming at all operable objects including the target object.
And correcting the first numerical value according to the second historical operation behavior data.
Further optionally, the determining the lower limit benchmark value according to all single transaction amounts ranked as the first numerical value in the plurality of unit times specifically includes:
calculating a mean and a standard deviation from all single transactions ranked as a first value in a plurality of unit times;
and determining the lower limit reference value according to all the single transaction amount, the average value and the standard deviation which are sequenced into the first numerical value.
Further optionally, the determining the lower limit reference value according to all the single transaction amounts, the mean values and the standard deviations which are sorted into the first numerical value specifically includes:
calculating the kurtosis value and skewness value of normal distribution according to the single transaction amount, the mean value and the standard deviation which are sequenced into the first numerical value;
when the kurtosis value and the skewness value are determined to be zero, determining that all single transaction amounts sequenced into a first numerical value accord with standard normal distribution, and determining a lower limit reference value according to the mean value and the standard deviation;
or when the kurtosis value is determined to be not zero and the skewness value is zero, determining that all single transaction amounts sequenced into a first numerical value accord with the post-peak tail distribution, and determining a lower-limit reference value according to the mean value, the standard deviation and the kurtosis value;
or when the kurtosis value is determined to be zero, the skewness value is not zero, or the kurtosis value and the skewness value are not zero, determining that all single transaction amounts which are sequenced into a first numerical value accord with the post-spike distribution, determining a left quantile point alpha of a confidence interval by using an integral method, and determining a lower limit reference value n as a coefficient obtained by normal distribution by using a mean value and a standard deviation according to the left quantile point alpha.
Further optionally, when it is determined that the kurtosis value and the skewness value are both zero, it is determined that all the single-transaction amounts ordered as the first numerical value conform to the standard normal distribution, and the lower-limit reference value is determined according to the mean value and the standard deviation, and specifically includes: and determining a lower limit reference value according to the mean value, the standard deviation and a coefficient n, wherein the coefficient n is a coefficient determined by normally distributing by using the mean value and the standard deviation and reaching a preset confidence level.
Further optionally, when it is determined that the kurtosis value is not zero, determining that all single-transaction amounts sorted into the first numerical value conform to the post-spike-tail distribution, and determining a lower-limit reference value according to the mean value, the standard deviation, and the kurtosis value, specifically including:
and adjusting according to the kurtosis value to obtain a coefficient n, and determining a lower limit reference value according to the mean value, the standard deviation and the coefficient n.
Optionally, the method further includes: acquiring a preset minimum reference value;
and when the lower limit reference value is determined to be smaller than the lowest reference value according to all the single transaction amounts which are sequenced to be the first numerical value in the unit time, determining the lowest reference value as the final lower limit reference value.
Further optionally, before determining the first value according to the preference value and the target object transaction amount per unit time in each of the plurality of unit times, the method further includes: counting the target object transaction amount of a first unit time in a plurality of unit times;
when the target object transaction amount in a first unit time in a plurality of unit times is larger than a preset threshold value, reducing the target object transaction amount in the first unit time according to a preset rule;
and sorting the single transaction amount of the reduced target object according to the transaction amount, wherein the first unit time is any one of a plurality of unit times.
The method steps in the target object message sending method provided in this embodiment have been described in detail in the embodiment corresponding to fig. 3, and therefore are not described herein again.
The target object message sending method provided by the embodiment of the invention has the following beneficial effects: the reference lower limit value determined by the embodiment is different from the mechanical means of 'one-time cutting' in the first prior art. The difference exists in consideration of the difference of the self-attribute of the target object. That is, a reference lower limit value is determined for each transaction of the target object. The difference from the second prior art solution is that the reference lower limit determined by the present solution does not need to consider whether certain transactions are active for a certain period of time, and naturally does not cause the reference lower limit to always fluctuate in the lateral and longitudinal directions. In addition, in the scheme of the application, when the reference lower limit value is determined, the factor of historical operation behavior data of the user on the target object is added, and the transaction amount of the target object in the actual situation and the single transaction amount of all transactions conducted on the target object are combined. Therefore, the degree of interest of the user for each target object can be determined (the degree of interest uses a lower limit reference value as an evaluation standard), and according to the degree of interest of the user for the target object, the transaction messages meeting the preset conditions are screened from the transaction messages in the subsequent unit time and sent to the front end. Namely, the message push standards for different target objects are respectively formulated for each user, so that personalized push is realized. The method can be more beneficial to achieving the expectation degree of the user for knowing the target object under the condition of not disturbing the user too much.
Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above embodiments are provided to further explain the objects, technical solutions and advantages of the present invention in detail, it should be understood that the above embodiments are merely exemplary embodiments of the present invention and are not intended to limit the scope of the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (8)
1. A target object messaging apparatus, the apparatus comprising:
the data collection module is used for collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions aiming at the target object;
the statistical module is used for receiving the first historical operation behavior data sent by the data collection module and determining the preference value of the user to the target object according to the first historical operation behavior data;
the measuring benchmark calculation module receives the preference value of the user to the target object, the target object transaction amount of each unit time in the unit times and all single transaction amounts for the target object to be transacted, wherein the preference value is sent by the statistic module, and the target object transaction amount is sent by the data collection module;
determining a first numerical value according to the preference value and the target object transaction amount of each unit time in the unit times; extracting the single transaction amount ordered to the first value from the target object transaction in each unit time; determining the lower limit reference value according to all single transaction amounts sequenced to the first numerical value in the unit times;
and the screening module screens the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value sent by the measurement reference calculation module, and sends the transaction messages to the front end.
2. The apparatus according to claim 1, wherein the data collection module is further configured to collect second historical operation behavior data performed by the user for all operable objects including the target object;
and the measurement reference calculation module receives the second historical operation behavior data sent by the data collection module and corrects the first numerical value according to the second historical operation behavior data.
3. The apparatus according to claim 1 or 2, wherein the metric calculation module calculates a mean and a standard deviation from all single transactions in the plurality of unit times that are ordered as the first value;
and determining the lower limit reference value according to the single transaction amount, the mean value and the standard deviation which are sorted into the first numerical value.
4. The apparatus of claim 3, wherein the metric calculation module is specifically configured to:
calculating a kurtosis value and a skewness value of normal distribution according to the single transaction amount, the mean value and the standard deviation which are ranked as the first numerical value;
when the kurtosis value and the skewness value are both determined to be zero, determining that all the single transaction amounts ranked as the first numerical value conform to standard normal distribution, and determining the lower limit reference value according to the mean value and the standard deviation;
or when the kurtosis value is determined to be not zero and the deviation value is zero, determining that all single transaction amounts ranked as the first numerical value conform to the post-peak tail distribution, and determining the lower-limit reference value according to the mean value, the standard deviation and the kurtosis value;
or when the kurtosis value is determined to be zero, the deviation value is not zero, or the kurtosis value and the deviation value are not zero, determining that all single transaction amounts sequenced into the first numerical value accord with the post-spike tail distribution, determining a left quantile point alpha of a confidence interval by using an integral method, and determining the lower limit reference value according to the left quantile point alpha.
5. The apparatus according to claim 4, wherein the metric calculating module is specifically configured to determine the lower-limit reference value according to the mean, the standard deviation, and a coefficient n, where the coefficient n is a coefficient determined when the mean and the standard deviation are normally distributed and a preset confidence level is reached;
or, the measurement reference calculation module is specifically configured to obtain the coefficient n according to the kurtosis value adjustment, and determine the lower-limit reference value according to the mean value, the standard deviation, and the coefficient n.
6. The apparatus of any one of claims 1, 2 or 4, 5, further comprising: an acquisition module;
the acquisition module acquires a preset minimum reference value;
the measurement reference calculation module receives the lowest reference value sent by the acquisition module;
and when the lower limit reference value is determined to be smaller than the lowest reference value according to all the single transaction amounts which are sequenced to be the first numerical value in the unit time, determining the lowest reference value to be the final lower limit reference value.
7. The apparatus according to any one of claims 1, 2, 4 and 5, wherein the metric calculation module is further configured to count a target object transaction amount of a first unit time in the plurality of unit times;
when the target object transaction amount in a first unit time in the unit times is larger than a preset threshold value, reducing the target object transaction amount in the first unit time according to a preset rule;
and sorting the single transaction amount of the reduced target object according to the transaction amount, wherein the first unit time is any one of the unit times.
8. A target object messaging method, the method comprising:
collecting first historical operation behavior data executed by a user aiming at a target object, target object transaction amount of each unit time in a plurality of unit times and single transaction amount of all transactions aiming at the target object;
determining a preference value of the user for the target object according to the first historical operation behavior data;
determining a first numerical value according to the preference value and the target object transaction amount of each unit time in the unit times;
extracting the single transaction amount ordered to the first value from the target object transaction in each unit time;
determining the lower limit reference value according to all single transaction amounts sequenced to the first numerical value in the unit times;
and screening the transaction messages meeting the preset conditions from the transaction messages in the subsequent unit time according to the lower limit reference value, and sending the transaction messages to the front end.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010254039.1A CN111586100B (en) | 2020-04-01 | 2020-04-01 | Target object message sending device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010254039.1A CN111586100B (en) | 2020-04-01 | 2020-04-01 | Target object message sending device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111586100A CN111586100A (en) | 2020-08-25 |
| CN111586100B true CN111586100B (en) | 2022-04-22 |
Family
ID=72124270
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010254039.1A Active CN111586100B (en) | 2020-04-01 | 2020-04-01 | Target object message sending device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111586100B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113610644B (en) * | 2021-08-16 | 2023-10-13 | 中国农业银行股份有限公司 | System transaction screening method and device |
Family Cites Families (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8341158B2 (en) * | 2005-11-21 | 2012-12-25 | Sony Corporation | User's preference prediction from collective rating data |
| CN104092597A (en) * | 2014-07-03 | 2014-10-08 | 深圳市富途网络科技有限公司 | Instant communication method and system for financial transactions |
| US20160014057A1 (en) * | 2014-07-08 | 2016-01-14 | Samsung Electronics Co., Ltd. | Method and system for providing dynamically customized web push messages in a wireless network |
| GB2534849A (en) * | 2015-01-28 | 2016-08-10 | Canon Kk | Client-driven push of resources by a server device |
| CN106156122B (en) * | 2015-04-07 | 2020-02-07 | 阿里巴巴集团控股有限公司 | Transaction information acquisition method and device |
| CN107169029B (en) * | 2017-04-13 | 2020-06-23 | 深圳大学 | Recommendation method and device |
| CN108629045A (en) * | 2018-05-14 | 2018-10-09 | 平安科技（深圳）有限公司 | A kind of information push method, storage medium and server |
| CN109816424A (en) * | 2018-12-14 | 2019-05-28 | 深圳壹账通智能科技有限公司 | Credit card favor information recommended method and system based on customer consumption portrait |
| CN109934721A (en) * | 2019-01-18 | 2019-06-25 | 深圳壹账通智能科技有限公司 | Finance product recommended method, device, equipment and storage medium |
-
2020
- 2020-04-01 CN CN202010254039.1A patent/CN111586100B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111586100A (en) | 2020-08-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109509082B (en) | Monitoring method and device for bank application system | |
| CN111586100B (en) | Target object message sending device and method | |
| CN109583729B (en) | Data processing method and device for platform online model | |
| Wei et al. | Informed trading, flow toxicity and the impact on intraday trading factors | |
| CN116089401B (en) | User data management method and system | |
| CN113312578A (en) | Data index fluctuation attribution method, device, equipment and medium | |
| CN106156122B (en) | Transaction information acquisition method and device | |
| CN115174686B (en) | Method and device for dynamically adjusting weights of multiple service channels based on service efficiency | |
| Igan et al. | The premia on state-contingent sovereign debt instruments | |
| CN111724009A (en) | Risk assessment method, wind control system and risk assessment equipment | |
| Tran et al. | Value-at-risk and the global financial crisis | |
| CN113988639A (en) | Asset value dynamic management system | |
| US20130091072A1 (en) | Algorithm for post-trade analysis and formulation of optimized strategy for subsequent trades | |
| CN109544304B (en) | Method for carrying out early warning according to mobile terminal information | |
| CN113191857A (en) | Cloud platform e-commerce data processing method and system based on big data | |
| KR102517169B1 (en) | System and method for automatic stock trading | |
| CN116934182B (en) | Enterprise data collaborative observation method, system and medium | |
| KR101954129B1 (en) | Smart Communication FinTech Payment System | |
| CN113609454A (en) | Abnormal transaction detection method and device | |
| US20110035330A1 (en) | Quality Control of Investment Performance Calculations | |
| CN115170314A (en) | Method, system and device for automatically selecting identification channel in image identification field | |
| JP2003006431A (en) | System and method for risk calculation, and program | |
| Tran et al. | Improving Quantile Forecast Accuracy: An Application to Oil Market | |
| CN114518990A (en) | Service data early warning method, device, equipment and storage medium | |
| Cayton et al. | The Impact of News Sentiment on Financial Risk: An Extreme Value Approach |
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 |