CN101894316A - Method and system for monitoring indexes of international market prosperity conditions - Google Patents

Abstract

The invention discloses a dynamic management system and a dynamic management method for monitoring and forecasting indexes of international market conditions. The method comprises the following steps of: summarizing and feeding back operational information recorded by distributed database servers in time; calculating and deducing the operational information to form the monitoring indexes of the international market prosperity conditions; and periodically releasing dynamic monitoring reports on the international market conditions, so that medium and small-sized enterprises and administrative departments of governments at all levels can know the current conditions and development tendency of the international market and assist the international market in making synthetic judgments on the development of the international market. In a system for monitoring the indexes of the international market prosperity conditions, a core part consists of a service processing subsystem, a data integration subsystem and a monitoring support subsystem. The normalized and standardized system for monitoring the indexes of the international market prosperity conditions is established to be used as a weatherglass for the medium and small-sized enterprises and the related government administrative departments to make judgments on the requirement of the international market on Chinese products and macro economy tendency.

Description

The monitoring index method and the system of the prosperous situation in a kind of international market

Technical field:

The present invention relates to ecommerce B2B transaction, be particularly related to a kind of index method that is used for the prosperous condition monitoring in international market, it is characterized in that the running log information of B2B platform is discerned, classified and gathers, and regularly, regularly be pooled to the present invention, and then calculate, deducing becomes all kinds of international markets consumer confidence index, the present invention provides the method and system of a supervision and prediction foreign trade for marcoeconomic management department, B2B platform management organization and the dealer that participates in international trade.

Background technology:

Current Small and Medium Enterprises in China has surpassed more than 4,000 ten thousand families, account for more than 99% of national enterprise sum, the final products that medium-sized and small enterprises are created and the value of service have accounted for nearly 6 one-tenth of gross domestic product (GDP), the cities and towns job that provides has accounted for 75%, has become China and has reached following most active economic sector at present.Under such economic situation, promote the market development capability and the viability of medium-sized and small enterprises effectively, become medium-sized and small enterprises self and even related management department of government the common problem of paying close attention to.

Enquiry data according to " the ecommerce world " shows, in investigated 1488 tame enterprises based on medium-sized and small enterprises, nearly enterprise more than 70% has carried out polytype ecommerce, wherein, carry out in the enterprise of ecommerce, 67% enterprise thinks that the affiliate increases, and professional channel is wideer; 58% corporate client increases, sales volume increases.Obviously, utilization ecommerce means can be created more commercial opportunity for medium-sized and small enterprises.In addition, in the investigation enterprise, 36.29% enterprise negotiates business by network, and 25.81% enterprise searches market conditions on the net, illustrates by investigation enterprise very vigorous to the demand of trade information.

Made in China net (Made-in-China.com) is a B2B E-commerce platform, this platform has compiled a large amount of visits of external buyer and domestic medium-sized and small enterprises seller and the product information data of " made in China ", and present whole foreground system has nearly 1,000 ten thousand pages visit every day; By on Dec 31st, 2009, have nearly 400 myriabits of medium-sized and small enterprises sum of registration; Medium-sized and small enterprises in effective product information of made in China net e-commerce platform (Made-in-China.com) issue and market conditions above 1,500 ten thousand; Made in China net e-commerce platform (Made-in-China.com) has been arranged in by the end of December, 2009 from the user capture that surpasses 220 countries and regions, the visitor surpasses 20,0,000,000 person-times, the discovery of carrying out commercial opportunity every day from the businessman of more than 200 countries and regions, the world by the made in China net with exchange, the total amount of the orders that struck a bargain on made in China net platform in 2009 is above 20,000,000,000 dollars.

Based on large-scale user capture and actual B2B transaction, the operation data storehouse of made in China net platform has accumulated the operation data of magnanimity, a lot of data recording and reflected that domestic and international enterprise customer is engaged in the trading activity of international trade by e-commerce platform, has reflected the variation of international market demand real-time.For example, the visit capacity of each phase of platform and inquiry amount all in time and have significantly reflected the variation characteristic of international market demand.Grasp the demand information of international market real-time and accurately, carry out analysis of market conditions and commercial opportunity discovery has important practical significance for the Small and Medium Enterprises in China administrative authority relevant with government.Because daily operation data is distributed in a plurality of different databases, and data volume is very big, how daily operation data these distributions, magnanimity is put in order, discerns, analyzed and excavates, need the integration and the reconstruct of system and method, could obtain real-time international market market conditions change dynamics information like this.

For a long time, all require to have a kind of method that can address the above problem, the prosperous situation in international market accurately is provided in real time in medium-sized and small enterprises international trade and marcoeconomic management field always.

The three phases that made in China net (Made-in-China.com) follows Small and Medium Enterprises in China externally to do business by e-commerce platform.This three phases is respectively: lead-in stage (1996-1999), difficulty stage (2000-2003), developing stage (2004 so far).In this surplus ten years domestic Development of E-business processes, there is bigger unbalancedness in B2B E-commerce between the development each province and city of China; There is bigger difference in the various countries buyer to the manufacturing product demand of Small and Medium Enterprises in China and actual China export abroad total value; The e-commerce development level of every profession and trade also occurs than large deviation.Rely on the experience of direct feel and successful project, be not enough to address the above problem.We need a kind of monitoring index number system and the system that can explore the prosperous situation in international market, for scientific system ground the state of development of international market and trend are judged to provide support.

Because made in China net e-commerce platform is in first camp of China's B2B E-commerce, have certain guiding for economic behaviours such as analyzing international market buying to China based on the prosperous condition monitoring system in the international market of mass data and system, because it is relatively large that the coastland of ecommerce prosperity accounts for the proportion of China's economic total volume, be in the upstream of whole economy industrial chain, therefore by realizing the monitoring of the prosperous situation in international market, can indicate following variation in the whole nation or other areas.And the branch province can be associated with relevant commodity production and sale with industry-specific more refinement monitoring, under certain condition, can predict the variation of more commodity field or other economic fields.

Set up the monitoring index number system and the system of the prosperous situation in standardization and standardized international market, make it to become medium-sized and small enterprises and related management department of government barometer international market product demand to China and the judgement of macroeconomy tendency.

Summary of the invention:

The objective of the invention is to provide index monitoring and the prediction dynamic management system and the method for a kind of international market market conditions, the running information that when it is conceived substantially the database server that distributes is write down in time gathers, feeds back, and become the prosperous condition monitoring index in international market by calculating, deducing, regularly publish international market market conditions dynamic monitoring report, thereby medium-sized and small enterprises and administrative authority of governments at all levels are understood international market the present situation, development trend, and assisted its development to make comprehensive judgement the international market.

The technical solution used in the present invention is: the index monitoring of a kind of international market market conditions and prediction dynamic management system, and it comprises service process subsystem, data integration subsystem, monitoring support subsystem, it is characterized in that:

Described service process subsystem, set up the data acquisition system (DAS) relevant with operation flow, comprised promising external buyer and domestic medium-sized and small enterprises the website operation link of service is provided, the user use and with the process of user interactions in produce a large amount of business datums, collect raw data with storage in this course and be the basis of the integrated and index analysis of late time data; In source data collection and the storage, require all relevant application systems all to gather various data, store into then in certain medium according to specific form, frequency, granularity, interrelational form; These data comprise that the personal user visits behavior, the client uploads data, client's Email;

Described data integration subsystem, realize data integration, arrangement and conversion, calculating and analysis requirement according to analysis of market conditions software, raw data is passed through to derive, cleaning and filtering, the numerical value conversion, crosscheck, the dimension table upgrades, steps such as history lists filing are classified and are integrated, by carrying out data integration arrangement and conversion ETL operation, with the data conversion in raw data source is the multidimensional analysis data of complete unanimity, form client's behavioral data fairground, customer profile data fairground and product information Data Mart, for ease of follow-up data mining analysis, realize immediate inquiring, the report form showing function;

Subsystem is supported in described monitoring, set up the evaluation index and the appraisement system of science, make up data mining and business intelligence tool platform, realize the dynamic calculation of the prosperous condition monitoring index in international market, regularly finish sme development and international market monitoring report by the report personnel; Finish and from each Data Mart, read required data, form the required base values of Index for Calculation, finish the calculating and the analytical work of consumer confidence index.

Monitoring of the index of above-mentioned international market market conditions and prediction dynamic management system, it is further characterized in that: described data integration subsystem comprises the identification index and the search unit of the collection of mass data and centralized stores unit, mass data; The identification index and the search unit of described mass data, adopt Distributed Calculation application technology, the mass data of magnanimity unstructured data on-line analytical processing OLAP, give the analytic type BI supporting technology of semantic identification, realize identification, index and the search subscriber inquiry and the report capability of daily record data, finish every basic behavioral indicator, category index and the search carried out according to the time at every profession and trade, domestic each province seller and global various countries buyer; The index of setting customizes as required.

The invention also discloses the monitoring index method of the prosperous situation in a kind of international market, it is characterized in that:

Set up the data acquisition system (DAS) relevant with operation flow, comprised promising external buyer and domestic medium-sized and small enterprises the website operation link of service is provided, the user use and with the process of user interactions in produce a large amount of business datums, collect raw data with storage in this course and be the basis of the integrated and index analysis of late time data; In source data collection and the storage, require all relevant application systems all to gather various data, store into then in certain medium according to specific form, frequency, granularity, interrelational form; These data comprise that the personal user visits behavior, the client uploads data, client's Email;

Realize data integration, arrangement and conversion, calculating and analysis requirement according to analysis of market conditions software, raw data classified by steps such as derivation, cleaning and filtering, numerical value conversion, crosscheck, the renewal of dimension table, history lists filings and integrated, by carrying out data integration arrangement and conversion ETL operation, with the data conversion in raw data source is the multidimensional analysis data of complete unanimity, form client's behavioral data fairground, customer profile data fairground and product information Data Mart, for ease of follow-up data mining analysis, realize immediate inquiring, report form showing function;

Set up the evaluation index and the appraisement system of science, make up data mining and business intelligence tool platform, realize the dynamic calculation of the prosperous condition monitoring index in international market, regularly finish sme development and international market monitoring report by the report personnel; Finish and from each Data Mart, read required data, form the required base values of Index for Calculation, finish the calculating and the analytical work of consumer confidence index.

The monitoring index method of the prosperous situation in above-mentioned international market, it is further characterized in that: described index comprises buyer's active index BAI, seller's active index SAI, international market consumer confidence index IMPI, industry active index, region active index; External any active ues login times year of described buyer's active index BAI=abroad increases newly apart from the external any active ues of speed of development * 25%+ year apart from speed of development * 35%+ and registered user's year sends out inquiry total amount year apart from speed of development * 15% apart from speed of development * 25%+; Domestic any active ues login times year of described seller's active index SAI=is received inquiry amount year apart from speed of development * 20% apart from speed of development * 15%+ domestic newly-increased registered user's year apart from newly-increased product information amount year of speed of development * 15%+ apart from the domestic any active ues of speed of development * 30%+ year apart from speed of development * 20%+; International market consumer confidence index concentrated expression both parties participate in the active degree of international trade practice, and the geometric mean of getting buyer's active index and seller's active index obtains, and is described

The present invention has following beneficial effect: the present invention has set up the monitoring index number system and the system of the prosperous situation of standardization and standardized international market, makes it to become medium-sized and small enterprises and related management department of the government barometer to international market product demand to China and the judgement of macroeconomy tendency.

The present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.The drawings and specific embodiments do not limit the scope of protection of present invention.

Description of drawings

Fig. 1 is the integrated stand composition of the prosperous detection system of the present invention;

Fig. 2 is data acquisition of the present invention, arrangement, identification and flow path switch figure;

Fig. 3 is a consumer confidence index system assumption diagram of the present invention;

Fig. 4 is that international market of the present invention consumer confidence index (IMPI), industrial added value and total export year are apart from the speed of development time series.

Embodiment

Implementation method of the present invention:

1. prosperous monitoring system overall architecture.The monitoring index number system and the system of the prosperous situation in international market abbreviate prosperous monitoring system as, and the core supports subsystem to form (the integrated stand composition of prosperous monitoring system as shown in Figure 1) by service process subsystem, data integration subsystem, monitoring.

1) sets up the data acquisition system (DAS) relevant with operation flow.Comprised promising external buyer and domestic medium-sized and small enterprises the website operation link of service is provided, the user use and with the process of user interactions in produce a large amount of business datums, collect raw data with storage in this course and be the basis of the integrated and index analysis of late time data.In source data collection and the storage, require all relevant application systems all to gather various data, store into then in (as database, daily record text and other destructuring carriers) in certain medium according to specific form, frequency, granularity, interrelational form; These data comprise that the personal user visits behavior, the client uploads data, client's Email etc.;

2) realize data integration, arrangement and conversion, main calculating and analysis requirement according to analysis of market conditions software, raw data is passed through to derive, cleaning and filtering, the numerical value conversion, crosscheck, the dimension table upgrades, steps such as history lists filing are classified and are integrated, by carrying out ETL operations such as data integration arrangement and conversion, with the data conversion in raw data source is the multidimensional analysis data of complete unanimity, form client's behavioral data fairground, customer profile data fairground and product information Data Mart, for ease of follow-up data mining analysis, realize immediate inquiring, functions such as report form showing.

3) set up the evaluation index and the appraisement system of science, make up tool platforms such as data mining and business intelligence, realize the dynamic calculation of the prosperous condition monitoring index in international market, regularly finish sme development and the international market monitoring is reported by the report personnel.Mainly need finish and read required data from each Data Mart, form the required base values of Index for Calculation, finish the calculating and the analytical work of consumer confidence index, the consumer confidence index architecture as shown in Figure 2.And then by special data analyst be responsible for regularly generating statistical report form and put in order written, for medium-sized and small enterprises and government department provide the economic situation prediction.

2. data acquisition, arrangement, identification and conversion

1) collection of mass data and centralized stores, the present invention will be from other platform and the system of made in China net and focus science and technology operation, and collects mass data from the internet related web site, also the data drawing-in system that will collect from alternate manner and channel.The existing way of various data, storage format difference, system realize automatically these data being carried out collection regularly, storage, and processing.Idiographic flow is seen Fig. 2, data acquisition of the present invention, arrangement, identification and flow path switch figure.

2) identification of mass data, index and search, the present invention adopts the Distributed Calculation application technology of magnanimity unstructured data, the on-line analytical processing of mass data (On-Line Analytical Processing, OLAP), the analytic type BI that gives semantic identification supports the gordian technique of (Analytic Application) etc., realized the user inquiring and form (the Query ﹠amp such as identification, index and search of daily record data; Reports) function can be finished the every basic behavioral indicator at every profession and trade, domestic each province seller and global various countries buyer, category index and the search carried out according to the time.

3) index of She Dinging can customize as required.For example domestic active users index, we set every month once member of visit is the active membership, and creates dimension in view of the above and add up, and can understand in the distribution situation of time, area, industry the active membership.Fig. 2 data acquisition, arrangement, identification and flow path switch figure.

Below be a specific embodiment with external buyer's the inquiry information of sending out, collection, identification, arrangement and storage, conversion and the displaying flow process of raw information among the present invention is described.

(1) information is initiated: for example a certain external buyer, open the MIC website, browse product after, find that some product meets oneself needs, can send the inquiry mail by the website, inquire relevant information.

(2) information Recognition, index and storage: Web server receives " sending inquiry " this user behavior operation information synchronously, can divide field store in Service Database the operation key feature, as fields such as inquiry content, an inquiry people, receipts inquiry people, industry, locations.The BI system then can enter data warehouse with the Various types of data collection in the Service Database by the ETL process every day.

(3) information translation: data cube is added up and set up to user behavior operation informations such as inquiry with fields such as inquiry content, an inquiry people, receipts inquiry people, industry, locations as required by analysis engine after entering data warehouse.

(4) information exhibition: generate daily paper at last on demand, weekly, all kinds of forms such as monthly magazine store on the preceding station server.The user asks to conduct interviews and browse by the web server is sent.

3. consumer confidence index system

1) the whole thinking of consumer confidence index System Design.With reference to business condition index, entrepreneur's confidence index and Merchandising Manager's index, above-mentioned index is based on the diffusion index of sample survey and subjective assessment, therefore yet the final embodiment of customer transaction confidence is still implemented in the trading activity of reality, adopts data by made in China net platform recording user trading activity aspect as the base values of reflection buyer's active degree and seller's active degree.

At first dealer's behavior related data of platform record has been carried out combing, obtained the base values of five aspects, comprised 1. any active ues login times: the login times that is the domestic and international active membership user of made in China net platform in the report period; 2. active users: be meant that in the report period made in China net platform counts the home and overseas number of users that access platform within a certain period of time surpasses certain number of times; 3. domestic newly-increased registered user: be the report period demestic user recruitment of determining according to user's registration information; 4. newly-increased product information amount: the newly-increased dosage of the product information of user's issue in the report period, wherein rejected the product information amount of issuing or deleting of being limited; 5. receive the inquiry amount: domestic seller collects the actual effectively total amount of inquiry in the report period; 6. send out the inquiry amount: external user sends the actual effectively total amount of inquiry in the report period, because buyer's an envelope inquiry is issued a plurality of sellers, the mass-sending inquiry is only calculated once among the present invention.

Secondly, analyzed the variation characteristic of These parameters trend feature, cyclical variation feature and rate of growth respectively, and adopted the industrial added value and the total export of China to contrast index, carried out each index time lag correlation analysis, periodicity analysis respectively as benchmark.According to the concrete feature of dealer's behavior that is embodied at the analysis result of each index and each index, determine to calculate the base values of buyer's active index and seller's active index and calculated weight, and adopt year to remove the seasonal factor of base values, and then the computing method of international market consumer confidence index have been designed apart from the relative number method.Wherein, the year of each base values apart from relative number (being called year again apart from speed of development) computing formula is:

2) consumer confidence index architecture (consumer confidence index system assumption diagram as shown in Figure 3)

Each index implication and computing method related among Fig. 3 are respectively:

1. (Buyer Activity Index BAI), reflects the degree of prosperity of external buyer's (purchaser) active degree and international market buying to China to buyer's active index, also is the index that the external buyer of reflection purchases confidence; Main according to external visit people next year apart from speed of development, abroad newly-increased registered user's several years apart from speed of development, send out and calculate in inquiry amount year apart from speed of development.This index reflects the compared with the same period of last year relative active degree of current buyer's participatory economy activity, is that a year is apart from relative number.

Computing formula is:

Buyer's active index=external any active ues login times year abroad increases newly apart from the external any active ues of speed of development * 25%+ year apart from speed of development * 35%+ and registered user's year sends out inquiry total amount year apart from speed of development * 15% apart from speed of development * 25%+

2. (Supplier Activity Index SAI), reflects domestic seller (domestic medium-sized and small enterprises) active degree to seller's active index, also is the index of the domestic medium-sized and small enterprises outlet of the reflection degree of prosperity; Mainly calculate in inquiry total amount year in next year apart from speed of development, domestic newly-increased registered user's several years apart from speed of development, newly-increased product information amount year apart from speed of development apart from speed of development and receipts according to domestic visit people.SAI also is year apart from relative number, and computing formula is:

Seller's active index=domestic any active ues login times year is received inquiry amount year apart from speed of development * 20% apart from speed of development * 15%+ domestic newly-increased registered user's year apart from newly-increased product information amount year of speed of development * 15%+ apart from the domestic any active ues of speed of development * 30%+ year apart from speed of development * 20%+

3. international market consumer confidence index (International Market Prosperity Index, IMPI), concentrated expression both parties participate in the active degree of international trade practice, the geometric mean of getting buyer's active index and seller's active index obtains, and is the quantification index of the concentrated expression international market degree of prosperity.IMPI is year apart from relative number, and computing formula is:

4. the industry active index reflects every profession and trade domestic seller active degree and the International traders active degree of State Statistics Bureau to the manufacturing industry classification.39 big class industries according to State Statistics Bureau are integrated coupling to the related data of platform, and according to the industry analysis that statistics bureau defines, obtain the comprehensive plate index of buyer's active index, seller's active index and industry of main industries.Because the products catalogue of made in China net covers more than 1000 the big products catalogue that the Chinese foreign outlet is relatively more active, and is relevant in corresponding industrial trade by mapping relations.Its computing method are respectively with reference to BAI and SAI.

5. the region active index reflects buyer's active degree of each main trading nation and buyer's active degree of domestic each province.Wherein: one, each main trading nation buyer active index be primarily aimed at the country that in China's Foreign Trade market, occupies critical role and in platform the country of brisk trade analyze, this index can provide the data support for the national changes in distribution that medium-sized and small enterprises and related management department of government in time grasp China main exit market; Two, the domestic each province seller active index medium-sized and small enterprises seller that is primarily aimed at each administrative division of China active degree that participates in trade activity is analyzed, and can allow medium-sized and small enterprises and related management department of government in time understand the distribution and the variation in the active area of China's Foreign Trade.

3) the index weight of BAI and SAI calculating is selected

In BAI (buyer's active index) and SAI (seller's active index) computing formula, each base values Method of Weight Determination has adopted principal component analysis (PCA) and two kinds of methods of step analysis, wherein principal component analysis (PCA) is the tax power of index being carried out according to the variability of achievement data, and analytical hierarchy process is to carry out the result that the characteristic root analysis obtains according to the judgment matrix that the importance that the expert carries out between any two to index relatively obtains, what principal component analysis (PCA) reflected is the feature that data itself are had, what analytical hierarchy process reflected is the feature that the index intension is had, the present invention makes up two kinds of methods, obtain the calculating weight of seller's active index and buyer's active index, make the weight of being calculated to reflect the make a variation feature of two aspects of index intension and numerical value preferably.Simultaneously according to the result of total quantity index year apart from the analysis of the incidence relation between speed of development and the macroeconomy base values, practical significance in conjunction with the website operation user behavior active degree that experience and index characterized, the preliminary weight of determining is adjusted, thereby obtain existing calculating weight, as shown in table 1:

Table 1 active index base values is composed power table as a result

(1) computation process of principal component analysis (PCA):

The first, suppose that we have observed n object, remember that the observed reading of an i object of observation p index is respectively:

x _I1, x _I2, L, x _Ip, then the observed reading of all a n object p index can be expressed as following matrix form:

$X=\left[\begin{array}{cccc}{x}_{11}& x_{12}& L& x_{1p}\\ x_{21}& x_{22}& L& x_{2p}\\ M& M& M& M\\ x_{n1}& x_{n2}& L& x_{\mathrm{np}}\end{array}\right]$

Wherein, n is an object of observation number, and p is index or variable number.

The second, raw data is carried out standardization, disposal route is:

${x_{\mathrm{ik}}}^{\′}=(x_{\mathrm{ik}}-{\stackrel{\‾}{x}}_k)/s_k,$ i＝1，2L?n，k＝1，2L?p

In the formula ${\stackrel{\‾}{x}}_k=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{x}_{\mathrm{ik}}/n,$ $s_k^2=\frac{1}{n-1}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{(x_{\mathrm{ik}}-{\stackrel{\‾}{x}}_k)}^2$

After the standardization, the variance of variable or index is 1, and average is 0.

The 3rd, the correlation matrix of establishing the observed value formation is

$R=\left[\begin{array}{cccc}{r}_{11}& r_{12}& L& r_{1p}\\ r_{21}& r_{22}& L& r_{2p}\\ M& M& M& M\\ r_{n1}& r_{n2}& L& r_{\mathrm{np}}\end{array}\right]$

The related coefficient of the data after standardization is:

$r_{\mathrm{ij}}=\underset{k=1}{\overset{n}{\mathrm{\Σ}}}{x_{\mathrm{ki}}}^{\′}{x_{\mathrm{kj}}}^{\′}/(n-1),$ i，j＝1，2L，p

The 4th, corresponding to correlation matrix R, ask secular equation | the p of R-λ I|=0 non-negative eigenwert

λ ₁，λ ₂，L，λ _p。Corresponding to eigenvalue _iProper vector be:

C _i＝(c _1i，c _2i?L，c _pi)′，i＝1，2，Lp

The 5th, ask major component.P the major component of being made up of proper vector is:

F _i＝c _1iX ₁+c _2iX ₂+L+c _piX _p

Major component F ₁, F ₂, L, F _pBetween be independent of each other, and their variance is successively decreased.

The 6th, select m (the individual principal component of m＜p).If front m major component F ₁, F ₂, L, F _mThe variance sum ratio that accounts for whole population variances approach 1 (in general) as long as reach 82%, we just get preceding m major component F ₁, F ₂, L, F _mThis m principal component variance sum accounts for more than 85% of whole population variances, promptly kept original index or variable X substantially ₁, X ₂, L, X _pInformation.

(2) the analytical hierarchy process concrete steps are as follows:

The first, the structure judgment matrix.Index i and j to same level carry out comparing in twos of corresponding importance, judge and quantification according to certain comparison scale, form relatively judgment matrix

The second, carry out the single preface of level.Make judgment matrix R=(r _Ij) _{N * n}, r _IjThe importance that expression factor i compares with respect to the inferior attribute of last layer than factor j, n is the order of matrix number.At first calculate the product of each row element of judgment matrix

Calculate root then n time

At last, right

Carry out normalization

W then _i(i=1,2, L n) has shown the relative priority degree of n factor, is the element of forming the judgment matrix characteristic vector W, i.e. W=(W ₁W ₂W _n) ^T, the sequencing weight of the corresponding factor relative importance of same level just.

The 3rd, consistency check.At first calculate the eigenvalue of maximum of judgment matrix

In the formula

I element for vectorial AW; Calculate the judgment matrix coincident indicator then

Calculate the Consistency Check in Judgement Matrix coefficient at last

Wherein RIRI is the mean random coincident indicator, and it is relevant with judgment matrix exponent number n, and when CR＜0.10, Consistency Check in Judgement Matrix is passed through, otherwise, just can't pass, need make judgement again.

3. predicted portions: be divided into two parts of macroeconomic forecasting and consumer confidence index prediction

1) macro-performance indicator prediction

1. prediction science, be the macro-performance indicator of checking based on international market consumer confidence index (IMPI)---the science of industrial added value and total export prediction, adopt the method for correlation analysis and Granger cause and effect check that existing cause-effect relationship between IMPI and the macro-performance indicator is verified respectively.Checking obtains two main results, and the one, there is correlativity highly in international market consumer confidence index (IMPI) with macro-performance indicator; The 2nd, the cause-effect relationship that exists between international market consumer confidence index (IMPI) and the macroeconomic undulation has the better prediction ability for macro-performance indicator.

Embodiment: in September, 2009, from correlation analysis, the correlativity of IMPI index and macro-performance indicator fluctuation has roughly experienced the process that grows from weak to strong, and sees Fig. 4 and table 2.This is with the opening of China economy and to make the historical progress of net platform development relevant, makes the net platform service 06,07 year and belongs to quick expansion period, and every data are higher, and is not obvious with macroeconomic association moving property; In year July in January, 08 to 09, platform development becomes surely, moves property with macroeconomic association and progressively embodies.Fig. 4 international market consumer confidence index (IMPI), industrial added value and total export year are apart from the speed of development time series.

The related coefficient of table 2 international market consumer confidence index (IMPI) and macroeconomy base values

By table as can be seen, IMPI and industrial added value year are higher than 0.5 apart from the speed of development related coefficient, remarkable with macroeconomic correlationship, buyer and seller's the behavior of participating in business is subjected to the influence that macroeconomy changes significantly in the platform, the international market consumer confidence index is high more, macroeconomy is flourishing more, and vice versa.

Result from the check of Granger cause and effect, the international economy consumer confidence index is to calculate gained by the associated transaction data of made in China net, the transaction data of a large amount of users at home and abroad has to a certain degree reflected the fluctuation of home and abroad economy, check the cause-effect relationship of observing international consumer confidence index and Chinese macro-indicators with the Granger cause and effect, shown in table 1 and table 2.

The Granger cause and effect assay of table 1 international market consumer confidence index (IMPI) and industrial added value speed of development

The IMPI index is not that the probability of the Granger reason of Chinese industrial added value speed of development growth is 1.28%, show under 98.72% confidence level, the IMPI index is that to influence the Granger reason industrial added value speed of development that the industrial added value speed of development increases not remarkable to the Granger check of international market consumer confidence index, can not form the Granger reason.

The Granger cause and effect assay of table 2 international market consumer confidence index (IMPI) and total export speed of development

The IMPI index is not that the probability of the Granger reason of Chinese gross foreign export value speed of development growth is 4.52%, show under 95.48% confidence level, the IMPI index is to influence the reason that the gross foreign export value speed of development increases, and gross foreign export value is not remarkable to the Granger check of international market consumer confidence index, can not form the Granger reason.

The above analysis illustrates that the international market consumer confidence index of calculating according to the transaction data that is reflected of making net (IMPI) has better predicting function to China's Macroscopic economy.In fact, in first and second report phase according to the IMPI index to July and industry in August added value predict the outcome with actual error less than 2%, illustrated directly that also the IMPI index is for macroeconomic predictive ability.Therefore, from the design and the calculating of index, the index that is obtained can characterize macroeconomic variation preferably.

2. Forecasting Methodology, the method that has mainly adopted exponential smoothing and trend modeling to combine.

Exponential smoothing:

Exponential smoothing is to carry out forecast method on the basis of single time series statistical model.Exponential smoothing is divided into exponential smoothing and exponential smoothing repeatedly.Generally speaking, using maximum is the double smoothing method, and its formula is:

$F_t^{\left(1\right)}=\mathrm{\α}{Y}_t^{\left(1\right)}+(1-\mathrm{\α})F_{t-1}^{\left(1\right)},F_t^{\left(2\right)}=\mathrm{\α}{F}_t^{\left(1\right)}+(1-\mathrm{\α})F_{t-1}^{\left(2\right)}$

Y _t+m＝α _t+b _tm

$a_t=2F_t^{\left(1\right)}-F_t^{\left(2\right)}$

$b_t=\frac{\mathrm{\α}}{1-\mathrm{\α}}(F_t^{\left(1\right)}-F_t^{\left(2\right)})$

Wherein

It is an exponential smoothing value in t cycle;

The double smoothing that is the t cycle is on duty for the day; It is the double smoothing value in t-1 cycle; α is a smoothing factor.

The exponential smoothing forecast model will be used smoothing factor α, and α value both representative model has determined the ability of forecast model smooth error again to the reaction velocity that time series data changes.The α value is big more, and then the proportion that accounts for of new data is just big more, and the latest observed value influence is big more, and prediction just depends on recent information more.The key of selecting is, makes the error minimum between predicted value and the actual value, and general α value is taken between 0 to 1.

The equation expression formula of multiple regression:

$\hat{y}=b_0+b_1{x}_1+b_2{x}_2+b_3{x}_3+K+b_k{x}_k$

Wherein, b ₀Being the intercept of regression plane on the y axle, is a constant.b ₁, b ₂, b ₃, b _kBe the coefficient of regression plane, claim partial regression coefficient again.

Embodiment:

Industrial added value is speed of development=88.3289+0.2515 * IMPI on year-on-year basis

T test value (31.52) (8.675)

This equation model goodness is 0.774, and the F test value is 75.25, P＜0.001.

In January, 2010, IMPI (international market consumer confidence index) is 149.6, and the industrial added value that obtains in Dec, 2010 according to this predictive equation speed of development predicted value on year-on-year basis is 125.95, and the predicted value of industrial added value is 14179.6 hundred million yuan.3. explanation predicts the outcome, as can be seen from Table 5, utilization IMPI is to the trend modeling and the Forecasting Methodology of industrial added value, predicted 2009 the 7 prediction numbers to domestic industry added value in November in " sme development and the international market monitoring report " of several phases before, actual number error of its prediction number and the announcement of the national statistics inning moon is all less than 2%.

Table 5 in July, 2009～November industry added value and each phase address prediction number contrast

2) consumer confidence index prediction

1. prediction science is the timing variations feature of analysis and excavation international market consumer confidence index (IMPI), and the present invention adopts spectrum analysis method that IMPI index cyclic swing is measured.Wherein, the spectral analysis method ultimate principle is as follows:

To a stationary time series { X _i, if its auto-covariance function R (k) satisfies

Then its spectral density function h (f) must exist and with R (k) following fourier transform relation be arranged:

$h\left(f\right)=\underset{k=-\∞}{\overset{\∞}{\mathrm{\Σ}}}R\left(k\right)e^{-i2\mathrm{\πfk}}$ |f|≤1/2

The equal sign both sides then can obtain normalized spectral density function p (f), promptly divided by variance R (0) in following formula

$p\left(f\right)=\frac{h\left(f\right)}{R\left(0\right)}=\underset{k=-\∞}{\overset{\∞}{\mathrm{\Σ}}}\frac{R\left(k\right)}{R\left(0\right)}{e}^{-i2\mathrm{\πfk}}=\underset{k=-\∞}{\overset{\∞}{\mathrm{\Σ}}}r\left(k\right)e^{-i2\mathrm{\πfk}}$ |f|≤1/2

Wherein, r (k) is the sequence autocorrelation function.

Right

With

The estimated value of standard spectrum density function of { X (t) } is as follows to utilize the truncation fourier transform of Tu Ji-Hanning window to obtain:

$\hat{p}\left(f_i\right)=\frac{\hat{h}\left(f_i\right)}{\hat{R}\left(0\right)}=1+2\underset{k=1}{\overset{M}{\mathrm{\Σ}}}w\left(k\right)\hat{r}\left(k\right)\cos 2\mathrm{\π}{f}_{i}k$ 0≤f _i≤1/2

Wherein

$w\left(k\right)=\left\{\begin{array}{cc}\frac{1}{2}(1+\cos \frac{\mathrm{\πk}}{M})& \left|k\right|\≤M\\ 0& \left|k\right|>M\end{array}\right.$

Be Tu Ji-Han Ning lag window.

From the DATA DISTRIBUTION of IMPI as can be seen, data present clocklike deferred in time variation tendency, and the year that the base values of IMPI Index for Calculation is based on each total quantity index such as inquiry amount apart from speed of development, removed the influence of seasonal factor, and passed through unit root test.Thereby, having steadily the time series of process immediately for one, common available fourier series comes match, and promptly spectrum analysis method is analyzed the Cycle Length of IMPI.

Table 6 IMPI seasonal effect in time series spectral density value and pairing Cycle Length

As seen from Table 6, the spectral density 9359.2 when i=1 is a spectrum peak, and this moment, corresponding Cycle Length was 46 months.Because whether IMPI seasonal effect in time series length was 46 phases, exist the longer cycle also to be difficult to determine, but from existing spectral density value and IMPI seasonal effect in time series distribution situation, the IMPI cycle of expection was believable between 46～50 months.

2. the time series modeling of IMPI index and prediction, the method that has mainly adopted exponential smoothing and trend modeling to combine

Embodiment: can see from the analysis of IMPI time series and cyclic swing in October, 2009, be in IMPI index upward period in the recent period, intercept the data in January, 2007～2009 year October, adopt the method for trend prediction and exponential smoothing respectively, obtain predicting the outcome shown in the table 7:

The IMPI exponential forecasting data in table 72009 year January in year November to 2010

The EXSMOOTH prediction only is fit to do the latest prediction, so exponential smoothing is only predicted the IMPI index in November, 2009, adopt 0.1 respectively, 0.2, ..., 0.9 as smoothing factor, chooses MSE predicted value hour, at this moment smoothing factor α=0.9 illustrates that also recent IMPI index has stronger trend feature.And then adopt the CUBIC model as the trend estimation basic model, data are carried out match, goodness of fit R ²Reach 0.897, F test value not 87.06, every coefficient of model is all remarkable, and it is as shown in table 6 to predict the outcome.

Predicting the outcome of two kinds of methods of contrast can see that in general, the international market prosperity degree of the fourth quater in 2009 can improve slightly, estimates that the fourth quater in 2009 domestic medium-sized and small enterprises economic activity meeting strengthens.

Claims (9)

1. the index of international market market conditions is monitored and the prediction dynamic management system, and it comprises service process subsystem, data integration subsystem, monitoring support subsystem, it is characterized in that:

Described service process subsystem, set up the data acquisition system (DAS) relevant with operation flow, comprised promising external buyer and domestic medium-sized and small enterprises the website operation link of service is provided, the user use and with the process of user interactions in produce a large amount of business datums, collect raw data with storage in this course and be the basis of the integrated and index analysis of late time data; In source data collection and the storage, require all relevant application systems all to gather various data, store into then in certain medium according to specific form, frequency, granularity, interrelational form; These data comprise that the personal user visits behavior, the client uploads data, client's Email;

Described data integration subsystem, realize data integration, arrangement and conversion, calculating and analysis requirement according to analysis of market conditions software, raw data is passed through to derive, cleaning and filtering, the numerical value conversion, crosscheck, the dimension table upgrades, steps such as history lists filing are classified and are integrated, by carrying out data integration arrangement and conversion ETL operation, with the data conversion in raw data source is the multidimensional analysis data of complete unanimity, form client's behavioral data fairground, customer profile data fairground and product information Data Mart, for ease of follow-up data mining analysis, realize immediate inquiring, the report form showing function;

Subsystem is supported in described monitoring, set up the evaluation index and the appraisement system of science, make up data mining and business intelligence tool platform, realize the dynamic calculation of the prosperous condition monitoring index in international market, regularly finish sme development and international market monitoring report by the report personnel; Finish and from each Data Mart, read required data, form the required base values of Index for Calculation, finish the calculating and the analytical work of consumer confidence index.

2. monitoring of the index of international market according to claim 1 market conditions and prediction dynamic management system is characterized in that: described data integration subsystem comprises the identification index and the search unit of the collection of mass data and centralized stores unit, mass data.

3. the index monitoring of international market according to claim 2 market conditions and prediction dynamic management system, it is characterized in that: the identification index and the search unit of described mass data, adopt the Distributed Calculation application technology of magnanimity unstructured data, the on-line analytical processing OLAP of mass data, give the analytic type BI supporting technology of semantic identification, realize the identification of daily record data, index and search subscriber inquiry and report capability, finish at every profession and trade, domestic each province seller and global various countries buyer's every basic behavioral indicator, category index and the search carried out according to the time.

4. the index monitoring of international market according to claim 3 market conditions and prediction dynamic management system, it is characterized in that: the index of setting customizes as required.

5. the monitoring index method of the prosperous situation in an international market is characterized in that:

Set up the data acquisition system (DAS) relevant with operation flow, comprised promising external buyer and domestic medium-sized and small enterprises the website operation link of service is provided, the user use and with the process of user interactions in produce a large amount of business datums, collect raw data with storage in this course and be the basis of the integrated and index analysis of late time data; In source data collection and the storage, require all relevant application systems all to gather various data, store into then in certain medium according to specific form, frequency, granularity, interrelational form; These data comprise that the personal user visits behavior, the client uploads data, client's Email;

Realize data integration, arrangement and conversion, calculating and analysis requirement according to analysis of market conditions software, raw data classified by steps such as derivation, cleaning and filtering, numerical value conversion, crosscheck, the renewal of dimension table, history lists filings and integrated, by carrying out data integration arrangement and conversion ETL operation, with the data conversion in raw data source is the multidimensional analysis data of complete unanimity, form client's behavioral data fairground, customer profile data fairground and product information Data Mart, for ease of follow-up data mining analysis, realize immediate inquiring, report form showing function;

Set up the evaluation index and the appraisement system of science, make up data mining and business intelligence tool platform, realize the dynamic calculation of the prosperous condition monitoring index in international market, regularly finish sme development and international market monitoring report by the report personnel; Finish and from each Data Mart, read required data, form the required base values of Index for Calculation, finish the calculating and the analytical work of consumer confidence index.

6. the monitoring index method of the prosperous situation in international market according to claim 5, it is characterized in that: described index comprises buyer's active index BAI, seller's active index SAI, international market consumer confidence index IMPI, industry active index, region active index.

7. the monitoring index method of the prosperous situation in international market according to claim 6 is characterized in that: external any active ues login times year of described buyer's active index BAI=abroad increases newly apart from the external any active ues of speed of development * 25%+ year apart from speed of development * 35%+ and registered user's year sends out inquiry total amount year apart from speed of development * 15% apart from speed of development * 25%+.

8. the monitoring index method of the prosperous situation in international market according to claim 6 is characterized in that: domestic any active ues login times year of described seller's active index SAI=is received inquiry amount year apart from speed of development * 20% apart from speed of development * 15%+ domestic newly-increased registered user's year apart from newly-increased product information amount year of speed of development * 15%+ apart from the domestic any active ues of speed of development * 30%+ year apart from speed of development * 20%+.

9. the monitoring index method of the prosperous situation in international market according to claim 6 is characterized in that: described

Images