Trade intelligent analysis method based on big data

Abstract

The invention relates to the technical field of intelligent trade analysis, and discloses an intelligent trade analysis method based on big data, and the method comprises the steps: obtaining a commodity type to which a presale commodity belongs, and recording the commodity type as a presale type; setting a collection time period, wherein the collection time period is a time period from the beginning of collection to the end of collection of the pre-sale type of commodities before the buyer purchases; executing the first habit statistical strategy to obtain a collection time period of the buyer; according to the historical order record, executing a second habit statistical strategy to obtain a time period from the end of collection to the purchase of the pre-sold type of commodities by the buyer, and recording the time period as an execution time period; executing the first prediction strategy, predicting the time that the buyer needs to experience after the execution time point to finish collection of the pre-sold type of commodities, and recording the time as a prediction time period; when the execution time is in the selling time period, marking the buyer as a target buyer; and executing the second prediction strategy to predict the sales of the pre-sold commodity. The relationship between the collection habit and the purchase habit of the buyer is determined, and the prediction accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of trade intelligent analysis, and in particular to a trade intelligent analysis method based on big data. Background Art

[0002] In today’s e-commerce environment, trade suppliers are responding to the dynamic changes in consumer behavior by leveraging advanced data analytics and machine learning technologies to optimize supply chain management and inventory strategies. This technical background is usually based on consumers’ interactions on online platforms, such as favorites lists and shopping cart contents. By analyzing this data, suppliers are able to predict demand trends for specific items and adjust their inventory and production plans accordingly. If a product is added to favorites or shopping carts by a large number of users, this may indicate high demand in the future, so suppliers can increase the inventory of the product in advance to avoid stock-outs. Suppliers can use these insights to optimize marketing strategies. Overall, this data-based supply chain management enables suppliers to respond to market demand more flexibly and efficiently, shorten product time to market, and reduce the risk of excess inventory, thereby achieving cost-effectiveness and increased customer satisfaction.

[0003] In the process of supplying goods, existing e-commerce companies only judge whether users will buy a product based on whether the product has been added to their favorites. This judgment method is too general and does not consider the relationship between users' favorite habits and purchasing habits. It cannot accurately estimate the demand for the product, resulting in inventory accumulation.

[0004] This solution proposes to execute the first habit statistical strategy and the second habit statistical strategy based on historical order records, obtain the buyer's collection period and execution period, determine whether the buyer is a target buyer, execute the second prediction strategy for the target buyer, and predict the sales of pre-sale goods. Summary of the invention

[0005] The present invention provides a trade intelligence analysis method based on big data, which is used to promote the solution of the problems mentioned in the above background technology.

[0006] The present invention provides the following technical solution: a trade intelligent analysis method based on big data, comprising:

[0007] Get the goods pre-sold by the seller and record them as pre-sold goods;

[0008] Placing the pre-sale products on the shopping platform for buyers to browse and collect;

[0009] Obtain the start and end time of the sale of the pre-sale product, and record the time period formed by the start and end time as the sale period;

[0010] Get the product category to which the pre-sale product belongs, and record it as the pre-sale category;

[0011] For each buyer:

[0012] Setting a collection period, which is the time period from when the buyer starts collecting to when the buyer stops collecting the pre-sale product before purchasing it;

[0013] Obtain the buyer's historical order records on the shopping platform;

[0014] According to the historical order records, the first habit statistical strategy is executed to obtain the buyer's collection period;

[0015] According to the historical order records, the second habit statistical strategy is executed to obtain the time period from the buyer's end of collection to the purchase of pre-sale products, which is recorded as the execution period;

[0016] If the buyer has collected the pre-sale item:

[0017] Obtain the time point when the buyer collects the pre-sale product, which is recorded as the execution time point;

[0018] Execute the first prediction strategy to predict the time it will take for the buyer to finish collecting the pre-sale products after the execution time point, which is recorded as the prediction period;

[0019] Calculate the forecast period + execution period, and record the result as the period to be purchased;

[0020] Get the time point of the purchase period after the execution time point, and record it as the execution time;

[0021] When the execution moment is within the selling period, the buyer is recorded as the target buyer;

[0022] The second forecasting strategy is implemented to forecast the sales of pre-sale products.

[0023] Optionally, executing the first habit statistics strategy according to the historical order records includes:

[0024] Setting a first interval for determining a collection period;

[0025] Get the time point of any purchase of pre-sale products in the historical order records, and record it as the first time point;

[0026] Get the time point of the first interval from the first time point forward, and record it as the second time point;

[0027] Recording the time point at which the buyer collects the pre-sale product during the period from the second time point to the first time point as the collection time point;

[0028] Get all collection time points;

[0029] Calculate the intervals between adjacent collection time points and form a collection interval set in chronological order from early to late.

[0030] Optionally, executing the first habit statistics strategy according to the historical order records includes:

[0031] Setting a determination value, the determination value is used to determine whether the collection interval set is used to calculate the collection period;

[0032] Get the number of elements in the collection interval set corresponding to each first time point as the number of collections at each first time point;

[0033] The first time point at which the number of collections is greater than the judgment value is recorded as the marked time point;

[0034] The collection interval set for each marked time point:

[0035] Setting a threshold number for determining collection periods;

[0036] Get the smallest interval in the collection interval set and record it as the marked interval;

[0037] Get the start time of the interval corresponding to the threshold number of intervals before the marked interval;

[0038] Obtain the end time of the interval corresponding to the threshold number of intervals after the marking interval;

[0039] The time period between the start time and the end time is recorded as the marked time period.

[0040] Optional, including:

[0041] Obtain the marked periods corresponding to all marked time points in the historical order records, which are recorded as the first historical marked period, the second historical marked period, ... the eth historical marked period respectively;

[0042] Execute the following formula:

[0043] (first historical marking period + second historical marking period + ... + e-th historical marking period) / e = collection period.

[0044] Optionally, executing a second habit statistics strategy according to historical order records includes:

[0045] For any marked time point in the historical order record:

[0046] Get the marked time period corresponding to the marked time point;

[0047] Calculate the interval from the end of the marked period to the marked time point;

[0048] Get the intervals corresponding to all marked time points, and record them as the first historical interval, the second historical interval, and so on, the nth historical interval;

[0049] Execute the following formula:

[0050] (first historical interval+second historical interval+...+nth historical interval) / n=execution period.

[0051] Optionally, executing the first prediction strategy includes:

[0052] Obtain the number of pre-sale products collected in the collection period from the execution time point forward, and record it as the first judgment number;

[0053] Obtain the number of pre-sale products collected during the collection period starting from the execution time point, and record it as the second judgment number;

[0054] When the first judgment quantity = 0:

[0055] Then, the length of the prediction period is the length of the collection period;

[0056] When the second judgment quantity = 0:

[0057] Then, the length of the prediction period is 0.

[0058] Optionally, executing the first prediction strategy includes:

[0059] When the first judgment quantity ≠ 0 and the second judgment quantity ≠ 0:

[0060] Calculate the first judgment quantity + the second judgment quantity = the total judgment quantity;

[0061] Calculate the second judgment quantity / the total judgment quantity = the collection ratio;

[0062] Then, calculate the collection ratio * collection period = length of the prediction period.

[0063] Optionally, executing the second prediction strategy includes:

[0064] Obtain the total number of pre-sale products collected by the target buyer during the collection period, and record it as the predicted number;

[0065] Calculate 1 / predicted quantity = predicted ratio;

[0066] Obtain the predicted ratios of all target buyers, which are recorded as the first predicted ratio, the second predicted ratio, ... the mth predicted ratio;

[0067] Execute the following formula:

[0068] First predicted ratio+second predicted ratio+…+mth predicted ratio=predicted sales of pre-sale goods.

[0069] The present invention has the following beneficial effects:

[0070] 1. The trade intelligent analysis method based on big data sets a first interval, finds the second time point corresponding to the first time point, obtains the collection time point in the time period from the second time point to the first time point, that is, obtains the collection time point close to the time point of purchasing the pre-sale goods. Such collection time points indicate that the buyer's purchasing sentiment is high, and best represent the buyer's collection of the pre-sale goods before purchasing the pre-sale goods. Pre-sale goods with a long collection time are excluded, which reduces the interference of calculation and makes the data more representative.

[0071] 2. This trade intelligent analysis method based on big data excludes the first time point when the collection interval concentration element is less than the judgment value through the judgment value. The number of pre-sale products collected at this first time point is small and cannot fully represent the collection situation of the buyer before purchasing the pre-sale products. Therefore, it needs to be excluded; reduce the interference of calculation and make the data more representative.

[0072] 3. The trade intelligent analysis method based on big data records the first time point corresponding to the buyer's collection situation before purchasing the pre-sale goods as the marked time point, obtains the smallest interval in the collection interval concentration, and obtains the marked interval. The marked interval represents the interval of the buyer's most frequent collection. Therefore, it is thought that the time before and after the time point corresponding to the marked interval are the times when the buyer frequently collects the pre-sale goods. The threshold number of intervals before and after the marked interval are obtained as the interval of the buyer's frequent collection, and the start time and end time are obtained, and then the marked period is obtained. The marked period corresponding to all the marked time points is calculated, and the average is calculated. The result is recorded as the buyer's collection period, that is, the time period when the buyer concentrates on collecting the goods before purchasing the pre-sale goods is determined, and the buyer's shopping habits are refined, providing a premise for subsequent judgment of whether the buyer is a target buyer.

[0073] 4. This intelligent trade analysis method based on big data divides buyers into authorized buyers and unauthorized buyers, obtains historical order records on the shopping platform of authorized buyers, protects the privacy of unauthorized buyers, and at the same time, uses authorized buyers as samples for analyzing the sales of pre-qualified goods to improve the accuracy of trade analysis.

[0074] 5. The intelligent trade analysis method based on big data calculates the interval between the end moment of the collection period corresponding to each marked time point and the marked time point, obtains all the intervals, calculates the mean, and obtains the execution period. The execution period indicates how much time it takes for the buyer to purchase the pre-sale goods after collecting them. The execution period corresponds to the time from the end of the buyer's frequent collection of pre-sale goods to the purchase of the pre-sale goods. The buyer's behavior is positioned on the two actions of frequent collection and purchase, which reduces the interference of unimportant data and simplifies the calculation process. The buyer's shopping habits are refined to provide a premise for subsequent judgment of whether the buyer is a target buyer.

[0075] 6. In the intelligent trade analysis method based on big data, when the buyer collects pre-sale goods, the execution time point is obtained, but the specific position of the buyer's collection period where the execution time point is located is uncertain. At this time, the number of pre-sale goods collected in the collection period before the execution time point is obtained to obtain the first judgment number; the number of pre-sale goods collected in the collection period after the execution time point is obtained to obtain the second judgment number; when the first judgment number is 0, it means that the pre-sale goods have not been frequently collected before the execution time point. Therefore, the predicted execution time point is the starting time of the collection period, and the predicted period = the collection period; when the second judgment number = 0, it means that the pre-sale goods have not been frequently collected after the execution time point. Therefore, the predicted execution time point is the end time of the collection period, and the predicted period = 0; first, it is determined whether there is a judgment number equal to 0, and the length of the prediction period is directly obtained to simplify the calculation.

[0076] 7. The trade intelligent analysis method based on big data, when the first judgment quantity ≠ 0 and the second judgment quantity ≠ 0: it means that the pre-sale products have been frequently collected before and after the execution time point, and the execution time point is within the collection period. The total judgment amount is calculated, and the second judgment quantity / the total judgment amount is divided to get the collection ratio. According to the collection ratio*collection period, the length of the prediction period is obtained. The number of collections in the collection period before and after the execution time point is calculated as the position of the execution time point in the collection period. Because the pre-sale products are pre-listed on the shopping platform and the quantity is limited, the buyer adds the pre-sale products to the collection, which means that the buyer is in high spirits to buy the pre-sale products, which means that the execution time point must be within the collection period. Then, the prediction period is obtained according to the first judgment quantity and the second judgment quantity, and the time for the buyer to continue to collect the pre-sale products is accurately determined, thereby providing prediction accuracy.

[0077] 8. This intelligent trade analysis method based on big data obtains the total number of pre-sale products collected during the collection period for a single buyer, and obtains the predicted number. Then, the probability that the buyer purchases the pre-sale product is equal to 1 / predicted number, and the predicted ratio is obtained. The predicted ratios of all target buyers are obtained, and the sum is calculated to obtain the predicted sales of the pre-sale products. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 Schematic diagram of the method of the present invention. DETAILED DESCRIPTION

[0079] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0080] Embodiment 1, refer to Figure 1 , trade intelligence analysis methods based on big data, including:

[0081] Get the goods pre-sold by the seller and record them as pre-sold goods;

[0082] Placing the pre-sale products on the shopping platform for buyers to browse and collect;

[0083] Obtain the start and end time of the sale of the pre-sale product, and record the time period formed by the start and end time as the sale period;

[0084] Get the product category to which the pre-sale product belongs, and record it as the pre-sale category;

[0085] For each buyer:

[0086] Setting a collection period, which is the time period from when the buyer starts collecting to when the buyer stops collecting the pre-sale product before purchasing it;

[0087] Obtain the buyer's historical order records on the shopping platform;

[0088] By dividing buyers into authorized buyers and unauthorized buyers, obtaining historical order records on the shopping platform of authorized buyers, and protecting the privacy of unauthorized buyers, authorized buyers are used as samples for analyzing the sales of pre-qualified goods to improve the accuracy of trade analysis.

[0089] According to the historical order records, the first habit statistical strategy is executed to obtain the buyer's collection period;

[0090] According to the historical order records, the second habit statistical strategy is executed to obtain the time period from the buyer's end of collection to the purchase of pre-sale products, which is recorded as the execution period;

[0091] If the buyer has collected the pre-sale item:

[0092] Obtain the time point when the buyer collects the pre-sale product, which is recorded as the execution time point;

[0093] Execute the first prediction strategy to predict the time it will take for the buyer to finish collecting the pre-sale products after the execution time point, which is recorded as the prediction period;

[0094] Calculate the forecast period + execution period, and record the result as the period to be purchased;

[0095] Get the time point of the purchase period after the execution time point, and record it as the execution time;

[0096] When the execution moment is within the selling period, the buyer is recorded as the target buyer;

[0097] The second forecasting strategy is implemented to forecast the sales of pre-sale products.

[0098] The executing of the first habit statistics strategy according to the historical order records includes:

[0099] Setting a first interval for determining a collection period;

[0100] Get the time point of any purchase of pre-sale products in the historical order records, and record it as the first time point;

[0101] Get the time point of the first interval from the first time point forward, and record it as the second time point;

[0102] Recording the time point at which the buyer collects the pre-sale product during the period from the second time point to the first time point as the collection time point;

[0103] In this embodiment, the length of the first interval is 3 months;

[0104] By setting a first interval, finding the second time point corresponding to the first time point, obtaining the collection time point in the time period from the second time point to the first time point, that is, obtaining the collection time point close to the time point of purchasing the pre-sale goods. Such collection time points indicate that the buyer's purchasing sentiment is high, and can best represent the buyer's collection of the pre-sale goods before purchasing the pre-sale goods. Pre-sale goods with a long collection time are excluded, which reduces the interference of calculation and makes the data more representative.

[0105] Get all collection time points;

[0106] Calculate the intervals between adjacent collection time points and form a collection interval set in chronological order from early to late.

[0107] The executing of the first habit statistics strategy according to the historical order records includes:

[0108] Setting a determination value, the determination value is used to determine whether the collection interval set is used to calculate the collection period;

[0109] Get the number of elements in the collection interval set corresponding to each first time point as the number of collections at each first time point;

[0110] The first time point at which the number of collections is greater than the judgment value is recorded as the marked time point;

[0111] In this embodiment, the determination value = 15;

[0112] In this embodiment, the buyer has a total of 5 purchase behaviors of pre-sale commodities, and 5 first time points are obtained. Each first time point corresponds to a collection interval set, which are recorded as the first collection interval set, the second collection interval set, the third collection interval set, the fourth collection interval set and the fifth collection interval set in chronological order.

[0113] Compare each collection interval set with the determination value, wherein the number of elements in the first collection interval set and the third collection interval set is greater than the determination value;

[0114] Therefore, the first time point corresponding to the first collection interval set and the third collection interval set is recorded as the marked time point;

[0115] By judging the value, the first time point whose elements in the collection interval set are smaller than the judgment value is excluded. The number of pre-sale products collected at this first time point is small and cannot fully represent the collection situation of the buyer before purchasing the pre-sale products. Therefore, it needs to be excluded; reducing the interference of calculations and making the data more representative.

[0116] The collection interval set for each marked time point:

[0117] Setting a threshold number for determining collection periods;

[0118] Get the smallest interval in the collection interval set and record it as the marked interval;

[0119] Get the start time of the interval corresponding to the threshold number of intervals before the marked interval;

[0120] Obtain the end time of the interval corresponding to the threshold number of intervals after the marking interval;

[0121] The time period between the start time and the end time is recorded as the marked time period.

[0122] Obtain the marked periods corresponding to all marked time points in the historical order records, which are recorded as the first historical marked period, the second historical marked period, ... the eth historical marked period respectively;

[0123] Execute the following formula:

[0124] (first historical marking period + second historical marking period + ... + e-th historical marking period) / e = collection period.

[0125] In this embodiment, the threshold number is 5;

[0126] For each favorite interval in the first favorite interval set and the third favorite interval set:

[0127] Get the smallest interval in the collection interval set and get the marked interval;

[0128] Get the start time of the interval corresponding to the 5 intervals before the marked interval;

[0129] Get the end time of the interval corresponding to 5 intervals after the marked interval;

[0130] Get the marked time period;

[0131] Among them, the marking period of the first collection interval set is 20 days;

[0132] The marking period for the third collection interval set is 18 days;

[0133] Then, the first historical marking period = 20 days; the second historical marking period = 18 days;

[0134] Calculate (first historical marking period + second historical marking period) / 2 = (20+18) / 2 = 19 days = collection period;

[0135] The first time point that can represent the buyer's collection situation before purchasing the pre-sale goods is recorded as the marked time point, and the smallest interval in the collection interval set is obtained to obtain the marked interval. The marked interval represents the interval when the buyer collects the pre-sale goods most frequently. Therefore, it is thought that the time before and after the time point corresponding to the marked interval are the times when the buyer frequently collects the pre-sale goods. The threshold number of intervals before and after the marked interval are obtained as the intervals when the buyer frequently collects, and the start time and end time are obtained, and then the marked period is obtained. The marked period corresponding to all the marked time points is calculated, and the average is calculated. The result is recorded as the buyer's collection period, which determines the time period when the buyer concentrates on collecting the goods before purchasing the pre-sale goods, refines the buyer's shopping habits, and provides a premise for subsequent judgment of whether the buyer is a target buyer.

[0136] The second habit statistics strategy is executed according to the historical order records, including:

[0137] For any marked time point in the historical order record:

[0138] Get the marked time period corresponding to the marked time point;

[0139] Calculate the interval from the end of the marked period to the marked time point;

[0140] Get the intervals corresponding to all marked time points, and record them as the first historical interval, the second historical interval, and so on, the nth historical interval;

[0141] Execute the following formula:

[0142] (first historical interval+second historical interval+...+nth historical interval) / n=execution period.

[0143] In this embodiment, the interval between the end time of the first historical marking period and the corresponding marking time point is equal to 4 days, and the first historical interval is obtained;

[0144] The interval between the end time of the second historical marking period and the corresponding marking time point is equal to 4 days, and the second historical interval is obtained;

[0145] Calculate (first historical interval + second historical interval) / 2 = (4+4) / 2 = 4 days = execution period;

[0146] By calculating the interval between the end moment of the collection period corresponding to each marked time point and the marked time point, all intervals are obtained, the mean is calculated, and the execution period is obtained. The execution period indicates how much time it takes for the buyer to purchase the pre-sale goods after collecting them. The execution period corresponds to the time from the buyer ending the frequent collection of pre-sale goods to the purchase of the pre-sale goods. The buyer's behavior is positioned on the two actions of frequent collection and purchase, which reduces the interference of unimportant data and simplifies the calculation process. The buyer's shopping habits are refined to provide a premise for subsequent judgment of whether the buyer is a target buyer.

[0147] The executing the first prediction strategy comprises:

[0148] Obtain the number of pre-sale products collected in the collection period from the execution time point forward, and record it as the first judgment number;

[0149] Obtain the number of pre-sale products collected during the collection period starting from the execution time point, and record it as the second judgment number;

[0150] When the first judgment quantity = 0:

[0151] Then, the length of the prediction period is the length of the collection period;

[0152] When the second judgment quantity = 0:

[0153] Then, the length of the prediction period is 0.

[0154] When the buyer collects pre-sale goods, the execution time point is obtained, but the specific position of the execution time point in the buyer's collection period is uncertain. At this time, the number of pre-sale goods collected in the collection period before the execution time point is obtained to obtain the first judgment number; the number of pre-sale goods collected in the collection period after the execution time point is obtained to obtain the second judgment number; when the first judgment number is 0, it means that the pre-sale goods have not been frequently collected before the execution time point. Therefore, the predicted execution time point is the starting time of the collection period, and the predicted period = collection period; when the second judgment number = 0, it means that the pre-sale goods have not been frequently collected after the execution time point. Therefore, the predicted execution time point is the end time of the collection period, and the predicted period = 0; first, it is determined whether there is a judgment number equal to 0, and the length of the prediction period is directly obtained to simplify the calculation.

[0155] The executing the first prediction strategy comprises:

[0156] When the first judgment quantity ≠ 0 and the second judgment quantity ≠ 0:

[0157] Calculate the first judgment quantity + the second judgment quantity = the total judgment quantity;

[0158] Calculate the second judgment quantity / the total judgment quantity = the collection ratio;

[0159] Then, calculate the collection ratio * collection period = length of the prediction period.

[0160] In this embodiment, the first judgment quantity = 6, the second judgment quantity = 6;

[0161] Calculate the first judgment number + the second judgment number = the total number of judgments = 12;

[0162] Calculate the second judgment quantity / the total judgment quantity = the collection ratio = 6 / 12 = 0.5;

[0163] Calculate the collection ratio * collection period = length of the prediction period = 0.5*19≈10 days.

[0164] When the first judgment quantity ≠ 0 and the second judgment quantity ≠ 0: it means that the pre-sale products have been frequently collected before and after the execution time point, and the execution time point is within the collection period. Calculate the total judgment amount, use the second judgment quantity / total judgment amount to get the collection ratio, and according to the collection ratio*collection period, get the length of the prediction period. Calculate the number of collections in the collection period before and after the execution time point to judge the position of the execution time point in the collection period. Because pre-sale products are pre-listed on the shopping platform and are limited in quantity, the buyer adds the pre-sale products to the collection, which means that the buyer is enthusiastic about purchasing the pre-sale products, which means that the execution time point must be within the collection period. Then, the prediction period is obtained according to the first judgment quantity and the second judgment quantity, and the time for the buyer to continue collecting the pre-sale products is accurately determined, thereby providing prediction accuracy.

[0165] The executing the second prediction strategy comprises:

[0166] Obtain the total number of pre-sale products collected by the target buyer during the collection period, and record it as the predicted number;

[0167] Calculate 1 / predicted quantity = predicted ratio;

[0168] Obtain the predicted ratios of all target buyers, which are recorded as the first predicted ratio, the second predicted ratio, ... the mth predicted ratio;

[0169] Execute the following formula:

[0170] First predicted ratio+second predicted ratio+…+mth predicted ratio=predicted sales of pre-sale goods.

[0171] In this embodiment, the buyer is a target buyer, and the total number of pre-sale commodities collected by the target buyer during the collection period is 10, that is, the predicted number = 10;

[0172] Calculate 1 / predicted quantity = predicted ratio = 0.1;

[0173] Get the predicted proportions of all target buyers, calculate the sum, and get the predicted sales of pre-sale goods equal to 5,000 units.

[0174] For a single buyer, obtain the total number of pre-sale products collected during the collection period to get the predicted number. Then, the probability of the buyer purchasing the pre-sale product is equal to 1 / predicted number, and the predicted ratio is obtained. Obtain the predicted ratios of all target buyers, calculate the sum, and get the predicted sales of the pre-sale products.

[0175] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0176] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A trade intelligence analysis method based on big data, comprising: Get the goods pre-sold by the seller and record them as pre-sold goods; Placing the pre-sale products on the shopping platform for buyers to browse and collect; Obtain the start and end time of the sale of the pre-sale product, and record the time period formed by the start and end time as the sale period; Get the product category to which the pre-sale product belongs, and record it as the pre-sale category; For each buyer: Setting a collection period, which is the time period from when the buyer starts collecting to when the buyer stops collecting the pre-sale product before purchasing it; Obtain the buyer's historical order records on the shopping platform; According to the historical order records, the first habit statistical strategy is executed to obtain the buyer's collection period; According to the historical order records, the second habit statistical strategy is executed to obtain the time period from the buyer's end of collection to the purchase of pre-sale products, which is recorded as the execution period; If the buyer has collected the pre-sale item: Obtain the time point when the buyer collects the pre-sale product, which is recorded as the execution time point; Execute the first prediction strategy to predict the time it will take for the buyer to finish collecting the pre-sale products after the execution time point, which is recorded as the prediction period; Calculate the forecast period + execution period, and record the result as the period to be purchased; Get the time point of the purchase period after the execution time point, and record it as the execution time; When the execution moment is within the selling period, the buyer is recorded as the target buyer; The second forecasting strategy is implemented to forecast the sales of pre-sale products.

2. The trade intelligent analysis method based on big data according to claim 1 is characterized by: The executing of the first habit statistics strategy according to the historical order records includes: Setting a first interval for determining a collection period; Get the time point of any purchase of pre-sale products in the historical order records, and record it as the first time point; Get the time point of the first interval from the first time point forward, and record it as the second time point; Recording the time point at which the buyer collects the pre-sale product during the period from the second time point to the first time point as the collection time point; Get all collection time points; Calculate the intervals between adjacent collection time points and form a collection interval set in chronological order from early to late.

3. The trade intelligent analysis method based on big data according to claim 2 is characterized by: The executing of the first habit statistics strategy according to the historical order records includes: Setting a determination value, the determination value is used to determine whether the collection interval set is used to calculate the collection period; Get the number of elements in the collection interval set corresponding to each first time point as the number of collections at each first time point; The first time point at which the number of collections is greater than the judgment value is recorded as the marked time point; The collection interval set for each marked time point: Setting a threshold number for determining collection periods; Get the smallest interval in the collection interval set and record it as the marked interval; Get the start time of the interval corresponding to the threshold number of intervals before the marked interval; Obtain the end time of the interval corresponding to the threshold number of intervals after the marking interval; The time period between the start time and the end time is recorded as the marked time period.

4. The trade intelligent analysis method based on big data according to claim 3 is characterized by: include: Obtain the marked periods corresponding to all marked time points in the historical order records, which are recorded as the first historical marked period, the second historical marked period, ... the eth historical marked period respectively; Execute the following formula: (first historical marking period + second historical marking period + ... + e-th historical marking period) / e = collection period.

5. The trade intelligent analysis method based on big data according to claim 4 is characterized by: The second habit statistics strategy is executed according to the historical order records, including: For any marked time point in the historical order record: Get the marked time period corresponding to the marked time point; Calculate the interval from the end of the marked period to the marked time point; Get the intervals corresponding to all marked time points, and record them as the first historical interval, the second historical interval, and so on, the nth historical interval; Execute the following formula: (first historical interval+second historical interval+...+nth historical interval) / n=execution period.

6. The trade intelligent analysis method based on big data according to claim 1 is characterized by: The executing the first prediction strategy comprises: Obtain the number of pre-sale products collected in the collection period from the execution time point forward, and record it as the first judgment number; Obtain the number of pre-sale products collected during the collection period starting from the execution time point, and record it as the second judgment number; When the first judgment quantity = 0: Then, the length of the prediction period is the length of the collection period; When the second judgment quantity = 0: Then, the length of the prediction period is 0.

7. The trade intelligent analysis method based on big data according to claim 6 is characterized by: The executing the first prediction strategy comprises: When the first judgment quantity ≠ 0 and the second judgment quantity ≠ 0: Calculate the first judgment quantity + the second judgment quantity = the total judgment quantity; Calculate the second judgment quantity / the total judgment quantity = the collection ratio; Then, calculate the collection ratio * collection period = length of the prediction period.

8. The trade intelligent analysis method based on big data according to claim 1 is characterized by: The executing the second prediction strategy comprises: Obtain the total number of pre-sale products collected by the target buyer during the collection period, and record it as the predicted number; Calculate 1 / predicted quantity = predicted ratio; Obtain the predicted ratios of all target buyers, which are recorded as the first predicted ratio, the second predicted ratio, ... the mth predicted ratio; Execute the following formula: First predicted ratio+second predicted ratio+…+mth predicted ratio=predicted sales of pre-sale goods.

Images