CN102693502A - Consumer consumption behavior oriented time-lapse data analysis model establishment method - Google Patents
Consumer consumption behavior oriented time-lapse data analysis model establishment method Download PDFInfo
- Publication number
- CN102693502A CN102693502A CN2012101817114A CN201210181711A CN102693502A CN 102693502 A CN102693502 A CN 102693502A CN 2012101817114 A CN2012101817114 A CN 2012101817114A CN 201210181711 A CN201210181711 A CN 201210181711A CN 102693502 A CN102693502 A CN 102693502A
- Authority
- CN
- China
- Prior art keywords
- consumption
- time
- user
- commodity
- analysis
- 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.)
- Pending
Links
Abstract
The invention discloses a consumer consumption behavior oriented time-lapse data analysis model establishment method. Through establishing a time-lapse based algorithm model, the consumption data are subjected to quantification analysis in combination with the forgetting curve characteristics of people, so that the interests and tendency of consumer consumption in a long-term range can be obtained, meanwhile, the group consumer preference and the individual classification preference can be subjected to comprehensive analysis, so that the drift and drift rate of consumption tendency can be subjected to quantizable prediction. Through the establishment and implementation of the establishment method of the analysis model, the interests and tendency of consumer consumption can be subjected to effective and reasonable analysis, which helps businessmen to make effective responses and adjustments on future market strategies better so as to enhance the market competitiveness.
Description
Technical field
The method for building up of the passage of time Data Analysis Model of the user oriented consumer behavior that the present invention relates to; Mainly be to carry out modeling analysis to the customer consumption data; Through being that the method for building up of analytical model on basis is inferred consumption interest and the trend that the user is long-term with the passage of time, the realization of this model relates to the data trend modeling, the forgetting curve analysis; Fields such as statistical algorithms; The foundation of the method for building up through this analysis model and realize and can carry out effectively interest and the trend of customer consumption, reasonable analysis helps businessman and better the market strategy in future is made effective reaction and adjust to enhance one's market competitiveness.
Background technology
Present retail trade dog-eat-dog, along with the fast development of internet electronic business, traditional retail trade is constantly permeating to ecommerce, shopping at network aspect gradually also; The crowd of customer consumption extensively changes, the approach diversification of consumption, the traditional retail consumption data sharp increase that is; For the retailer, how can better hold customer consumption interest and tendency, thereby to make market reaction timely; Grasp the first market opportunities, become a problem of having to face.The business data analysis of adopting at present then mainly is the index property analysis to data itself; The index match condition that can only response data itself be reflected; More the forecast analysis according to passage of time often all is through the artificial secondary analysis of doing; Lack of science lacks reasoning property.So rapidly under the megatrend, the traditional data analysis has been difficult to adaptation, is embodied in following aspects at the professional sharp increase of traditional retail and web development:
1. be directed against the index property analysis of data itself; The index match condition that can only response data itself be reflected; The method for building up based on the analytical model of passage of time of neither one standard can not carry out one to consumption interest and tendency and effectively analyze and predict.
2. traditional analysis is based on the analysis of time dimension piece, does not form the method for building up of continuous analytical model.
3. do not combine people's characteristics such as forgetting curve itself to carry out analysis-by-synthesis, just carry out index analysis through data itself.The actual property reference value of analysis result is little.
4. be not directed to the method for building up of normative analytic model of the tendency of permanent consumption interest specially, standardization level with can be all very low with reference to the property degree, analysis result does not possess actual application value.
Summary of the invention
The objective of the invention is to some limitation in the present consumer consumption behavior data analysis; The method for building up of the passage of time Data Analysis Model of a kind of user oriented consumer behavior that proposes; This model mainly is through setting up based on the algorithm model on the passage of time basis; Forgetting curve characteristic in conjunction with people itself is carried out quantitative analysis to consumption data; Thereby obtain customer consumption interest and tendency in the long-term scope; The foundation of the method for building up through this analysis model and realize and can carry out effectively interest and the trend of customer consumption, reasonable analysis helps businessman and better the market strategy in future is made effective reaction and adjust to enhance one's market competitiveness.
Technical scheme of the present invention is:
A kind of method for building up of passage of time Data Analysis Model of user oriented consumer behavior, this method may further comprise the steps:
A, at first set up the customer consumption data model libraries, gather the customer consumption data in a period of time;
B, set up consumer consumption behavior analysis indexes storehouse;
Pass C, Time Created customer consumption trend analysis method;
D, require each item index of analyzing for customer-action analysis index storehouse, the passage of time customer consumption trend analysis method of integrating step C is analyzed, and obtains the analysis result of each item index, exports as the passage of time Data Analysis Model;
E, through analysis to customer consumption tendency in the continuous time section, the weight of adjustment each item index is further optimized model.
Among the step C of the present invention, passage of time customer consumption trend analysis is meant calculating user's in a period of time r short-term consumption propensity, adopts following formula:
Wherein, f (k, r) be the user at the short-term consumption propensity degree of short-term time r to arbitrary type of commodity, (k, the consumption propensity index of t) browsing at moment t (t ∈ r) that belongs to corresponding type of commodity for the user; If it is enough little that time interval r obtains, the interior user of different time intervals can regard as separate to the short-term consumption propensity degree of corresponding type commodity, calculate interest-degree and all start from scratch.Its short-term consumption propensity degree is all calculated in each classification, just can obtain user's short-term consumption propensity model.
User of the present invention is that customer consumption tendency degree index calculates with linear equation to the interest-degree of arbitrary consumption corresponding goods, for consumption time is provided with two threshold value T1, and T2 (T1 < T2), then the user is to interest-degree f (k) equation of certain commodity:
Wherein, f (k) is the consumption propensity degree index of user to these commodity; X1 representes the goods browse time; X2 representes that commodity select number of times; A1, b1, c1 are constant, a, b, c refer to certain commodity or consumer objects;
Shorter like the goods browse time, i.e. x1 < T1, then goods browse time heavier a1 of shared weights in the customer consumption attention rate>b1;
, i.e. x1 longer like the goods browse time>T2, then commodity are selected number of times shared heavier a3 <b3 of weights in the user interest degree;
If the goods browse time is moderate, be T2>x1>T1, represent that then the shared weights in the user interest degree of a2 and b2 are close.
Among the present invention; Pass in time; Long term data index according to customer consumption is carried out the general trend analysis, adopts following method, gives different weights with the user's in the continuous different time sections short-term consumption propensity according to the priority of time; Calculate user's permanent consumption tendency again, adopt following formula:
Wherein, F (k) is the permanent consumption interest of corresponding type commodity; (k r) is i short-term consumption interest of corresponding type commodity; Ai is a forgetting factor; Forgetting factor ai approaches with negative exponential function, and value is:
Beneficial effect of the present invention:
One, the present invention combines people's the characteristic of forgeing characteristic itself to come consumption data is analyzed, and analyzes science more comprehensively.
Two, the present invention adopts the model analysis based on passage of time, and precision of analysis is higher, and prediction can be bigger with reference to property.
Three, the present invention can consume preference and the individual segregation preference is comprehensively analyzed to colony.
Four, the present invention can carry out quantifiable prediction to the skew and the drift rate of consumption propensity simultaneously, can be higher with reference to property.
Five, the method for building up of the analytical model of the present invention through setting up standard makes the standardization of consumption interest trend analysis, and reproducibleization improves analysis efficiency greatly.
Six, help businessman better to the market strategy in future make effective reaction and the adjustment to enhance one's market competitiveness.
Embodiment
A kind of method for building up of passage of time Data Analysis Model of user oriented consumer behavior;
A, at first set up the customer consumption data model libraries;
Comprise
Set up the systematic parameter storage list
Set up consumption index storage list
Set up consumption data time window snapshot storage list
Set up the consumption raw data table
Set up the forgetting curve allocation list
B, set up consumer consumption behavior analysis indexes storehouse.
Mainly comprise:
Short-term consumption propensity
The permanent consumption tendency
Classification short-term consumption propensity
Classification permanent consumption tendency
Passage of time consumption trend based on price
Based on seasonal passage of time consumption trend
Retail commodity is sold viscosity
The Retail commodity pouplarity
Consumption fluctuation expection degree
Consumption fluctuating range expectation index
Or the like.
Pass C, Time Created customer consumption trend analysis algorithm,
User's consumption is pointed in a short period r.Can use following formula to calculate:
Wherein, (k is that the user is at the short-term consumption propensity degree of time r to arbitrary type of commodity r) to f; , (k, the consumption propensity index of t) browsing at moment t (t ∈ r) that belongs to corresponding type commodity for the user.If it is enough little that the time asks that separated r obtains, it is separate that different time asks that separated interior user can regard as the short-term consumption propensity degree of corresponding type commodity, calculates interest-degree and all start from scratch.Its short-term consumption propensity degree is all calculated in each classification, just can obtain user's short-term consumption propensity model.
D, for consumer behavior index storehouse, binding time is passed analytical algorithm and is analyzed, the output result.The user can calculate with one group of linear equation the interest-degree of certain commodity.For consumption time is provided with two wealthy value T1, T2 (T1 T2), then the user is to interest-degree f (k) equation of certain commodity
Wherein, f (k) is the consumption propensity degree index of user to these commodity; X1 representes the goods browse time; X2 representes that commodity select number of times; A1, b 1, and c 1, a2, b2, c2, a3, b3, c3 are constant.Within a short period of time (x1 < T1), goods browse time shared weights in the customer consumption attention rate heavier (a1>b1); Like goods browse time long (x1>T2), then commodity are selected number of times shared weights in the user interest degree heavier (a3 <b3);
E, through to the analysis in the continuous time section; Adjustment index weight makes further optimization of model carry out the general trend analysis according to the long term data index of customer consumption; Can analyze the general trend of customer consumption; This long-term propensity to consume is the relatively more fixing preference of user, is metastable.After the fixedly preference that analyzes the user; Fluctuation just might take place or move in consumption propensity; The fluctuation of user's permanent consumption tendency is less; Give different weights with the user's in the different time sections short-term consumption propensity according to the priority of time, calculate user's permanent consumption tendency again.Its calculating formula is following:
Wherein, F (k) is the long-term interest of c class; (k r) is i short-term interest of c class; Ai is a forgetting factor.Forgetting factor a i is different, and the pace of change of customer consumption tendency is just different.The ai decay is fast, and the customer consumption tendency changes also very fast, and it is big more that the permanent consumption tendency is influenced by current short-term consumption propensity; Vice versa.
In practical application, according to great this forgetting curve of Chinese mugwort guest, forgetting factor ai can approach with negative exponential function:
The present invention does not relate to all identical with the prior art prior art that maybe can adopt of part and realizes.
Claims (5)
1. the method for building up of the passage of time Data Analysis Model of a user oriented consumer behavior is characterized in that, this method may further comprise the steps:
A, at first set up the customer consumption data model libraries, gather the customer consumption data in a period of time;
B, set up consumer consumption behavior analysis indexes storehouse;
Pass C, Time Created customer consumption trend analysis method;
D, require each item index of analyzing for customer-action analysis index storehouse, the passage of time customer consumption trend analysis method of integrating step C is analyzed, and obtains the analysis result of each item index, exports as the passage of time Data Analysis Model;
E, through analysis to customer consumption tendency in the continuous time section, the weight of adjustment each item index is further optimized model.
2. the method for building up of the passage of time Data Analysis Model of user oriented consumer behavior according to claim 1; It is characterized in that among the described step C; Passage of time customer consumption trend analysis is meant calculating user's in a period of time r short-term consumption propensity, adopts following formula:
Wherein, f (k, r) be the user at the short-term consumption propensity degree of short-term time r to arbitrary type of commodity, (k, the consumption propensity index of t) browsing at moment t (t ∈ r) that belongs to corresponding type of commodity for the user; The interior user of different time intervals regards as separate to the short-term consumption propensity degree of corresponding type commodity, calculate interest-degree and all start from scratch;
Commodity to each classification all calculate its short-term consumption propensity degree, obtain user's short-term consumption propensity model.
3. the method for building up of the passage of time Data Analysis Model of user oriented consumer behavior according to claim 1; It is characterized in that among the step D; The user is that customer consumption tendency degree index calculates with linear equation to the interest-degree of arbitrary type of commodity; For consumption time is provided with two threshold value T1, T2, T1 T2, user are to interest-degree f (k) equation of certain commodity:
Wherein, f (k) is the consumption propensity degree index of user to such commodity; X1 representes the goods browse time; X2 representes that commodity select number of times; A1, a2, a3, b1, b2, b3, c1, c2, c3 are constant, refer to certain commodity or consumer objects;
Shorter like the goods browse time, i.e. x1 < T1, then goods browse time heavier a1 of shared weights in the customer consumption attention rate>b1;
, i.e. x1 longer like the goods browse time>T2, then commodity are selected number of times shared heavier a3 <b3 of weights in the user interest degree;
If the goods browse time is moderate, i.e. T2>x1>T1, represent that then the shared weights in the user interest degree of a2 and b2 are close.
4. the method for building up of the passage of time Data Analysis Model of user oriented consumer behavior according to claim 1; It is characterized in that in the step e, pass in time, carry out the general trend analysis according to the long term data index of customer consumption; Adopt following method; Give different weights with the user's in the continuous different time sections short-term consumption propensity according to the priority of time, calculate user's permanent consumption tendency again, adopt following formula:
Wherein, F (k) is the permanent consumption interest of corresponding type commodity; (k r) is i short-term consumption interest of corresponding type commodity; Ai is a forgetting factor.
5. the method for building up of the passage of time Data Analysis Model of user oriented consumer behavior according to claim 4 is characterized in that forgetting factor ai approaches with negative exponential function, and value is:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101817114A CN102693502A (en) | 2012-06-04 | 2012-06-04 | Consumer consumption behavior oriented time-lapse data analysis model establishment method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2012101817114A CN102693502A (en) | 2012-06-04 | 2012-06-04 | Consumer consumption behavior oriented time-lapse data analysis model establishment method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN102693502A true CN102693502A (en) | 2012-09-26 |
Family
ID=46858910
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN2012101817114A Pending CN102693502A (en) | 2012-06-04 | 2012-06-04 | Consumer consumption behavior oriented time-lapse data analysis model establishment method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102693502A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104346698A (en) * | 2014-11-05 | 2015-02-11 | 无锡雅座在线科技发展有限公司 | Catering member big data analysis and checking system based on cloud computing and data mining |
CN106548368A (en) * | 2016-10-14 | 2017-03-29 | 五邑大学 | Consumer's intension recognizing method based on user's forgetting curve |
CN106779843A (en) * | 2016-12-15 | 2017-05-31 | 中国银联股份有限公司 | A kind of competing method and apparatus for closing relationship analysis of trade company based on customer group's feature |
CN107004221A (en) * | 2014-11-28 | 2017-08-01 | Bc卡有限公司 | For predict using industry card use pattern analysis method and perform its server |
CN108664552A (en) * | 2018-04-02 | 2018-10-16 | 拉扎斯网络科技(上海)有限公司 | A kind of user preference method for digging and device |
CN110163686A (en) * | 2019-05-27 | 2019-08-23 | 成都魔方城科技有限公司 | Desired consumption portrait method and system based on consumer behaviour |
CN110874441A (en) * | 2020-01-19 | 2020-03-10 | 中国传媒大学 | User interest analysis method and system combining memory forgetting and memory enhancement |
CN112513898A (en) * | 2018-07-31 | 2021-03-16 | 株式会社彩 | Alcoholic drink information management system and management method |
CN112990445A (en) * | 2021-05-13 | 2021-06-18 | 国网浙江省电力有限公司金华供电公司 | Intelligent analysis machine learning method for monitoring information of power distribution network |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030158771A1 (en) * | 2002-01-16 | 2003-08-21 | Ncr Corporation | Retention modeling methodology for airlines |
CN101620706A (en) * | 2008-06-30 | 2010-01-06 | 上海全成通信技术有限公司 | Data mining and modeling method for incremental sales |
CN101706911A (en) * | 2009-11-23 | 2010-05-12 | 浪潮集团山东通用软件有限公司 | Method for implementing service-oriented index model in business intelligence system |
CN102456198A (en) * | 2010-10-21 | 2012-05-16 | 镇江金软计算机科技有限责任公司 | Business intelligence-based realization method of consumer consumption habit behavior analysis |
-
2012
- 2012-06-04 CN CN2012101817114A patent/CN102693502A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030158771A1 (en) * | 2002-01-16 | 2003-08-21 | Ncr Corporation | Retention modeling methodology for airlines |
CN101620706A (en) * | 2008-06-30 | 2010-01-06 | 上海全成通信技术有限公司 | Data mining and modeling method for incremental sales |
CN101706911A (en) * | 2009-11-23 | 2010-05-12 | 浪潮集团山东通用软件有限公司 | Method for implementing service-oriented index model in business intelligence system |
CN102456198A (en) * | 2010-10-21 | 2012-05-16 | 镇江金软计算机科技有限责任公司 | Business intelligence-based realization method of consumer consumption habit behavior analysis |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104346698B (en) * | 2014-11-05 | 2018-03-06 | 无锡雅座在线科技股份有限公司 | Based on the analysis of the food and drink member big data of cloud computing and data mining and checking system |
CN104346698A (en) * | 2014-11-05 | 2015-02-11 | 无锡雅座在线科技发展有限公司 | Catering member big data analysis and checking system based on cloud computing and data mining |
CN107004221A (en) * | 2014-11-28 | 2017-08-01 | Bc卡有限公司 | For predict using industry card use pattern analysis method and perform its server |
CN106548368A (en) * | 2016-10-14 | 2017-03-29 | 五邑大学 | Consumer's intension recognizing method based on user's forgetting curve |
CN106779843B (en) * | 2016-12-15 | 2020-08-11 | 中国银联股份有限公司 | Method and device for analyzing merchant competitive relationship based on customer group characteristics |
CN106779843A (en) * | 2016-12-15 | 2017-05-31 | 中国银联股份有限公司 | A kind of competing method and apparatus for closing relationship analysis of trade company based on customer group's feature |
CN108664552A (en) * | 2018-04-02 | 2018-10-16 | 拉扎斯网络科技(上海)有限公司 | A kind of user preference method for digging and device |
CN112513898A (en) * | 2018-07-31 | 2021-03-16 | 株式会社彩 | Alcoholic drink information management system and management method |
CN110163686A (en) * | 2019-05-27 | 2019-08-23 | 成都魔方城科技有限公司 | Desired consumption portrait method and system based on consumer behaviour |
CN110874441A (en) * | 2020-01-19 | 2020-03-10 | 中国传媒大学 | User interest analysis method and system combining memory forgetting and memory enhancement |
CN110874441B (en) * | 2020-01-19 | 2020-05-19 | 中国传媒大学 | User interest analysis method and system combining memory forgetting and memory enhancement |
CN112990445A (en) * | 2021-05-13 | 2021-06-18 | 国网浙江省电力有限公司金华供电公司 | Intelligent analysis machine learning method for monitoring information of power distribution network |
CN112990445B (en) * | 2021-05-13 | 2021-07-16 | 国网浙江省电力有限公司金华供电公司 | Intelligent analysis machine learning method for monitoring information of power distribution network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102693502A (en) | Consumer consumption behavior oriented time-lapse data analysis model establishment method | |
Gajowniczek et al. | Short term electricity forecasting using individual smart meter data | |
Eickmeier et al. | How successful are dynamic factor models at forecasting output and inflation? A meta‐analytic approach | |
Guo et al. | Research on game model and complexity of retailer collecting and selling in closed-loop supply chain | |
Balcilar et al. | Long memory, economic policy uncertainty and forecasting US inflation: a Bayesian VARFIMA approach | |
Binder et al. | Dynamic models for managing durables using a stratified approach: the case of Tunja, Colombia | |
Kaufmann | Dating and forecasting turning points by Bayesian clustering with dynamic structure: A suggestion with an application to Austrian data | |
Kara | Does US monetary policy respond to oil and food prices? | |
Özmen et al. | Disaggregated evidence for exchange rate and import price pass-through in the light of identification issues, aggregation bias and heterogeneity | |
Corgel et al. | Hotel industry demand curves | |
Balcilar et al. | The role of economic policy uncertainty in forecasting US inflation using a VARFIMA model | |
Parr et al. | A method for socially evaluating the effects of long-run demographic paths on living standards | |
Augenbroe et al. | Decision model for energy performance improvements in existing buildings | |
Castillo-Manzano et al. | How much does water consumption drop when each household takes charge of its own consumption? The case of the city of Seville | |
Popeangă | Data mining smart energy time series | |
Skrodzka | Development of knowledge-based economy in European union in 2000-2014 | |
Brika et al. | Analysis of Algerian trade performance: From 1970 to 2014 | |
Maki | Introduction to estimating economic models | |
Sil et al. | The Effect of Entrepreneurship Competence, Entrepreneurship and Learning Orientation Toward Sustainable Competitive Advantages in Improving Managerial Performance on Medium Enterprises in Bengkulu Province | |
Šafr | Pilot Application of the Dynamic Input-Output Model. Case Study of the Czech Republic 2005-2013. | |
Suriya et al. | How to estimate the model of sustainable profit and corporate social responsibility | |
Pradhan et al. | Time series least square forecasting analysis and evaluation for natural gas consumption | |
Boasson et al. | Applying modern portfolio theory to municipal financial and capital budgeting decisions | |
Miskolczi et al. | Dependency between gross domestic product and unemployment in the Czech Republic | |
Mateia | Simple Moving Average vs Linear Regression Forecast |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120926 |
|
RJ01 | Rejection of invention patent application after publication |