CN105809195A - Method and apparatus for determining whether merchant belongs to particular merchant class - Google Patents

Abstract

The invention relates to a data processing technology, in particular relates to a method for judging whether a merchant belongs to a particular merchant class and a device for realizing the method. The method for judging whether a merchant belongs to a particular merchant class includes the following steps that: trade volume data relative to time of a plurality of merchant samples forming one sample set are received, wherein the merchant samples are located in the same geographic region and have the same merchant class; a trade distribution sequence relative to time of the sample set is generated based on the received trade volume data, and a trade distribution sequence relative to time of a merchant to be identified is generated according to the trade volume data of the merchant to be identified; and the trade distribution sequence of the sample set is compared with the trade distribution sequence of the merchant to be identified, so that whether a class to which the merchant to be identified is matched with a class to which the merchant samples belong is determined.

Description

For differentiating the method and apparatus whether a trade company belongs to particular merchant classification

Technical field

The present invention relates to data processing technique, in particular to being used for differentiating whether a trade company belongs to particular merchant class method for distinguishing and realize the device of said method.

Background technology

The online single market of lower receipts, each trade company is assigned with corresponding trade company's classification code (MCC).MCC is based on trade division, and for different MCC classifications, the formality rate of swiping the card of trade company there are differences.Ordering about by economic interests, low button rate MCC often can apply mechanically to reach to reduce the purpose of fee expenditure in the trade company of some high fee button rates, and this brings huge economic loss and goodwill loss to card tissue.But for card tissue, limit by man power and material, be again infeasible by comprehensively this unlawful practice is investigated and prosecuted in checking on the spot to whole businessmans work.

For above-mentioned limitation, industry generally adopts two kinds of methods to detect MMC at present and applies mechanically behavior in violation of rules and regulations.One of which method is dependent on holder and reports and verify on the spot, and the shortcoming of this method is that comparison is passive, and place one's entire reliance upon report, and therefore for the scale of construction of whole market, identified amount is only small.Another kind of method is based on to the participle of trade company's name and semantic analysis, and it filters out according to the key word feature that every profession and trade trade company is named and is named the trade company not meeting industry characteristic.Although this kind of method is it appeared that more Tao Ma trade company, but rate of false alarm is relatively larger, some is usually named special regular trade company and is also classified as trade company in violation of rules and regulations, and the fuzzy violation trade company of some trade company's name identification gets away with.

For above-mentioned reasons, applying mechanically in violation of rules and regulations MCC becomes POS under line and receives the persistent ailment of single medium-term and long-term existence in market, therefore accurately identify out in the way of low cost the unmatched technology of trade company's classification that trade company claims with it be industry in the urgent need to.

Summary of the invention

The present invention is provided to differentiate the trade company method and apparatus that whether belongs to particular merchant classification, it has and is easy to implement and high accuracy for examination.

According to an aspect of the invention, it is provided one is used for differentiating whether a trade company belongs to Particular merchant class method for distinguishing, comprises the following steps:

Receiving the turnover data of the multiple trade companies sample constituting a sample set and the turnover data of trade company to be identified, described trade company sample is positioned at same geographic area and has identical trade company's classification；

The transaction distribution series relative to the time of sample set described in the turnover data genaration of sample set, and by the turnover data genaration of trade company to be identified its relative to the transaction distribution series of time；And

The distribution series of concluding the business of transaction distribution series and the trade company to be identified of sample set comparing to determine, whether classification belonging to this trade company to be identified mates with the classification belonging to trade company sample.

Preferably, in the above-mentioned methods, the step relative to the transaction distribution series of time generating described sample set comprises the following steps:

For each trade company sample in sample set, generate its turnover ordered series of numbers relative to the time；

The turnover ordered series of numbers relative to the time of each trade company sample is carried out standardization processing, and wherein by standardization processing, the span of each data item of turnover ordered series of numbers is between 0-1；

Corresponding data item in the turnover ordered series of numbers of each trade company sample after standardization processing is collected the turnover ordered series of numbers relative to the time obtaining sample set；And

The turnover ordered series of numbers of sample set is carried out standardization processing to obtain the transaction distribution series relative to the time of sample set, wherein at the sample set obtained by standardization processing relative in the transaction distribution series of time, the span of each data item is between 0-1.

Preferably, in the above-mentioned methods, described standardization carries out as follows:

$x^{\′}=\frac{x-min}{max-min}$

Wherein, x is the value of one of them data item of turnover ordered series of numbers, and x' is this data item value after standardization processing, maximum in max and min respectively turnover ordered series of numbers and minima.

Preferably, in the above-mentioned methods, it is determined that the step whether this classification belonging to trade company to be identified mates with the classification belonging to trade company sample comprises the following steps:

Determine the first confidence interval and second confidence interval of following statistic of test:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, then determine this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if described statistic of test is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

Preferably, in the above-mentioned methods, it is determined that the step whether this classification belonging to trade company to be identified mates with the classification belonging to trade company sample comprises the following steps:

Determining whether the title of trade company to be identified exists occurrence in set of keywords, wherein, described set of keywords is made up of the keyword that the word frequency in the title of the trade company's sample in described sample set is higher；

If there is occurrence, it is determined that this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, otherwise, it is determined that the first confidence interval of following statistic of test and the second confidence interval:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, it is determined that this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if described statistic of test It is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

Preferably, in the above-mentioned methods, the following step is farther included:

If described statistic of test is positioned at outside described first confidence interval and is positioned within described second confidence interval, then calculate the fluctuation dependency of the transaction distribution series of sample set and the transaction distribution series of trade company to be measured as follows:

$r=\frac{1}{n-1}\underset{i=1}{\overset{n}{\Σ}}\left(\frac{X_i-\stackrel{\‾}{X}}{S_X}\right)\left(\frac{Y_i-\stackrel{\‾}{Y}}{S_Y}\right)$

Wherein, r is described fluctuation dependency, X_iAnd Y_iThe respectively i-th data item in the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set,WithThe respectively average of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, S_XAnd S_YThe respectively standard deviation of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set；And

If the threshold value that described fluctuation dependency is default more than or equal to, it is determined that this classification belonging to trade company to be identified and categorical match belonging to trade company's sample, otherwise, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

According to another aspect of the present invention, it is provided that a kind of for differentiating whether a trade company belongs to the device of particular merchant classification, including:

I/O unit, it is configured to receive the turnover data of multiple trade companies sample constituting a sample set and the turnover data of trade company to be identified and export differentiation result, and described trade company sample is positioned at same geographic area and has identical trade company's classification；

The data processing unit coupled with described I/O unit, comprising:

Ordered series of numbers generation module, it is configured to the transaction distribution series relative to the time of sample set described in the turnover data genaration of sample set and by the turnover data genaration of trade company to be identified accordingly relative to the transaction distribution series of time；And

Comparison module, it is configured to compare the distribution series of concluding the business of transaction distribution series and the trade company to be identified of sample set to determine whether classification belonging to this trade company to be identified mates with the classification belonging to trade company sample.

Accompanying drawing explanation

Above-mentioned and/or the other side of the present invention and advantage are by by becoming more fully apparent below in conjunction with the description of the various aspects of accompanying drawing and be easier to understand, and in accompanying drawing, same or analogous unit is adopted and is indicated by the same numeral, and accompanying drawing includes:

Fig. 1 be according to one embodiment of the invention for differentiating whether a trade company belongs to the device of particular merchant classification.

Fig. 2 is the schematic diagram of trade company's sample turnover ordered series of numbers within a period of time in sample set, and wherein, abscissa is the time, by hour in units of, vertical coordinate is turnover.

Fig. 3 be according to another embodiment of the present invention for differentiating whether a trade company belongs to the flow chart of particular merchant class method for distinguishing.

Fig. 4 is the flow chart that can be used for the ordered series of numbers generating routine in method shown in Fig. 3.

Fig. 5 is the flow chart that can be used for the comparison routine in method shown in Fig. 3.

Fig. 6 is the flow chart that can be used for another comparison routine in method shown in Fig. 3.

Detailed description of the invention

The present invention is more fully illustrated referring to the accompanying drawing which illustrates illustrative examples of the present invention.But the present invention can realize by multi-form, and is not construed as being only limitted to each embodiment given herein.The various embodiments described above provided are intended to make disclosure herein comprehensively complete, so that protection scope of the present invention more fully to convey to those skilled in the art.

Such as " comprising " and the term of " including " etc represents except having the unit and step having in the specification and in the claims directly and clearly state, technical scheme is also not excluded for the situation with other unit and the step directly or clearly do not stated.

According to a feature of the present invention, the time response of merchant transaction pipelined data (such as turnover data) can portray the trade mode of trade company well, therefore the matching degree of the trade company's classification claimed when swiping the card of trade company to be identified is determined with it by analyzing the time response of turnover data, thus screening out the violation trade company applying mechanically MCC code.Particularly, the present inventor finds through further investigation, trade company for same industry type, although the volume of the flow of passengers of different regions, different cities is variant in absolute quantity, but there is similar time response, i.e. for the trade company of same industry type, the turnover of they every days fluctuation situation difference in time is little, and commuter rush hour distribution is basically identical.Based on above-mentioned discovery, another according to the present invention One feature, by the Trip distribution rule of the Trip distribution rule and trade company to be identified that calculate a certain trade company classification (such as low button rate MCC) and the difference degree comparing the two under confidence degree level, it is possible to determine whether this trade company belongs to this trade company's classification.

Describe below by accompanying drawing and realize embodiments of the invention.

Fig. 1 be according to one embodiment of the invention for differentiating whether a trade company belongs to the device of particular merchant classification.Device 100 shown in Fig. 1 includes I/O unit 110 and the data processing unit 120 coupled with I/O unit.

I/O unit 110 is configured to receive a sample set, and this sample set is made up of multiple turnover data belonging to same trade company classification and trade company's sample of being positioned at same geographic area.I/O unit 110 is additionally configured to receive the turnover data of trade company to be identified.

For each trade company sample and band identification trade company, its turnover data are time serieses, and in time series, the interval at consecutive number strong point can be equal, it is also possible to unequal, and the unit at interval is also arbitrary, for instance include but not limited to minute, hour and day etc..

The selection of trade company's sample can have various ways, an exemplary example given below.

First, for some low button rate MCC, calculate the monthly average turnover of all trade companies belonging to this MCC in a certain geographic area (such as city), then monthly average turnover according to order sequence from high to low and is therefrom filtered out front N (such as N=200) individual trade company to determine the list of trade company in sample set.

It is alternatively possible to remove to obtain revised sample set from above-mentioned sample set by being likely to trade company in violation of rules and regulations by the mode of artificial screening.

The differentiation result that I/O unit 110 is additionally configured to determine data processing unit 120 exports to external equipment (not shown).

Data processing unit 120 is configured to compare the Trip distribution rule of sample set of I/O unit reception and the Trip distribution rule of trade company to be identified to determine whether this trade company belongs to the trade company's classification corresponding to sample set.As it is shown in figure 1, data processing unit 120 includes ordered series of numbers generation module 121 and comparison module 122.Ordered series of numbers generation module 121 is configured to the transaction distribution series relative to the time of the turnover data genaration sample set received from I/O unit 110；On the other hand, it is additionally configured to from the turnover data genaration of trade company to be identified accordingly relative to the transaction distribution series of time.

An illustrative examples generating transaction distribution series given below.

For each trade company sample in sample set, each hour interior turnover of every day that calculate it within a period of time (such as one month)；Subsequently by same at not same date of this trade company's sample Within individual hour, interior turnover adds up, thus forming the turnover ordered series of numbers of this trade company's sample, as shown in Figure 2.Alternatively but not necessarily, for the All Activity record of some month of each trade company sample in sample set, it is possible to only retain working day transaction data for the generation process of above-mentioned turnover ordered series of numbers.

Preferably, it is possible to the turnover ordered series of numbers of each trade company sample is standardized and is mapped in the interval of [0,1] with the value by each data item in ordered series of numbers.Data normalization, without influence on the distribution situation of data, also retains original quantity of information in data simultaneously.Better, it is possible to adopt following formula to carry out standardization processing:

$x^{\′}=\frac{x-min}{max-min}---\left(1\right)$

Wherein, x is the value of one of them data item of turnover ordered series of numbers, and x' is this data item value after standardization processing, maximum in max and min respectively turnover ordered series of numbers and minima.

After generating normalized turnover ordered series of numbers, it is possible to data item corresponding in the turnover ordered series of numbers of all trade companies sample (here for identical hour interior normalized turnover) is collected the turnover ordered series of numbers relative to the time obtaining sample set.For the turnover ordered series of numbers of sample set, again implement standardization processing as above, thus obtaining the transaction distribution series of sample set.

The transaction distribution series relative to the time for trade company to be identified, it would however also be possible to employ aforesaid way is from the turnover data genaration of this trade company.Specifically, for trade company to be identified, each hour interior turnover of every day that calculate it within a period of time (such as one month)；Subsequently this trade company is added up in same hour interior turnover of not same date, thus forming the turnover ordered series of numbers of this trade company.Preferably, it is possible to the turnover ordered series of numbers of trade company to be identified is standardized (mode for example with above formula (1)) is mapped in the interval of [0,1] with the value by each data item in ordered series of numbers.

Comparison module 122 is configured to compare the distribution series of concluding the business of transaction distribution series and the trade company to be identified of sample set to determine whether classification belonging to this trade company to be identified mates with the classification belonging to trade company sample.

An illustrative examples determining coupling given below.

In the present embodiment, utilize the statistic of test of following form to determine the matching degree between the transaction distribution series of sample set and the transaction distribution series of trade company to be tested:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}---\left(2\right)$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, for instance can value be 24, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified.

In the present embodiment, above-mentioned statistic of test is considered to obey the χ that degree of freedom is (n-1)²Distribution, thus can calculate the confidence interval obtained under certain significance level and determine matching degree by investigating statistic of test location in confidence interval.Specifically, it is possible to calculate under two significance levels the confidence interval (would correspond to the confidence interval of 90% significance level below be called the first confidence interval and would correspond to the confidence interval of 70% significance level and be called the second confidence interval) of (such as significance level respectively 90% and 70%).

Concrete Confidence rule such as may is that

If statistic of test is positioned within the first confidence interval, then determine this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if this statistic of test is positioned at outside the second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

It is positioned at outside the first confidence interval for statistic of test and is positioned at the situation within the second confidence interval, it is preferable that, it is possible to utilize the fluctuation dependency between the transaction distribution series of sample set and the transaction distribution series of trade company to be measured to make further to differentiate.Specifically, it is possible to use following formula calculates fluctuation dependency:

$r=\frac{1}{n-1}\underset{i=1}{\overset{n}{\Σ}}\left(\frac{X_i-\stackrel{\‾}{X}}{S_X}\right)\left(\frac{Y_i-\stackrel{\‾}{Y}}{S_Y}\right)---\left(3\right)$

Wherein, r is fluctuation dependency, X_iAnd Y_iThe respectively i-th data item in the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set,WithThe respectively average of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, S_XAnd S_YThe respectively standard deviation of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set.

Concrete fluctuation dependency decision rule may is that

If the threshold value (such as 0.7) that fluctuation dependency is default more than or equal to, then really This classification belonging to trade company to be identified fixed and the categorical match belonging to trade company's sample, otherwise, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

Alternatively, before according to above-mentioned Confidence rule judgment matching degree, can first judge the title of trade company to be identified is a preset keyword concentrates whether there is occurrence, if there is occurrence, then no longer confidence decision rule and relative fluctuation decision rule.It is misjudged because of turnover data entry error that this processing mode can effectively prevent those substantially to belong to the trade company of sample set trade company classification, thus improving the accuracy rate of identification.

Set of keywords can utilize following manner to generate: the title of trade company in sample set carries out participle and word frequency statistics and processes, and is taken out by several (such as 20) keywords before the highest for word frequency and generates set of keywords.

Fig. 3 be according to another embodiment of the present invention for differentiating whether a trade company belongs to the flow chart of particular merchant class method for distinguishing.Exemplarily, the method for the present embodiment realizes by the device shown in Fig. 1.

Referring to Fig. 1, in step S310, input/output module 110 receives the turnover data of the multiple trade companies sample constituting a sample set and the turnover data with identification trade company, and these trade company's samples are positioned at same geographic area and have identical trade company's classification.

Subsequently, in step S320, the ordered series of numbers generation module 121 of data processing unit 120 performs ordered series of numbers generating routine, with the transaction distribution series relative to the time from the turnover data genaration sample set of sample set, and from the turnover data genaration of trade company to be identified its relative to the transaction distribution series of time.

Exemplarily, in the present embodiment, as shown in Figure 4, ordered series of numbers generating routine generates the transaction distribution series of sample set as follows:

Step S321: as described previously for each trade company sample in sample set, by by cumulative thus forming the turnover ordered series of numbers relative to the time of this trade company's sample according to corresponding hour for its turnover of every day within a period of time (such as one month).All Activity record similarly, for each trade company sample in sample set, it is possible to only reservation transaction data on working day is for the generation process of above-mentioned turnover ordered series of numbers.

Step S322: the turnover ordered series of numbers relative to the time of each trade company sample is carried out standardization processing (such as utilizing above formula (1)), wherein by standardization processing, the span of each data item of turnover ordered series of numbers is between 0-1.

Step S323: the corresponding data item in the turnover ordered series of numbers of each trade company sample after standardization processing is collected the turnover ordered series of numbers relative to the time obtaining sample set.

Step S324: the turnover ordered series of numbers of sample set is carried out standardization processing to obtain the transaction distribution series relative to the time of sample set, wherein at the sample set obtained by standardization processing relative in the transaction distribution series of time, the span of each data item is between 0-1.

Step S325: for trade company to be identified, generate the turnover ordered series of numbers relative to the time according to the mode same with step S321.

Step S326: the turnover ordered series of numbers of trade company to be identified is carried out standardization processing according to the mode same with step S322.

Then, in step S330, comparison module 122 performs comparison routine to determine whether classification belonging to this trade company to be identified mates with the classification belonging to trade company sample, and in step S340, I/O unit 110 differentiates result to external equipment output.

Exemplarily, in the present embodiment, as it is shown in figure 5, comparison routine carries out coupling differentiation as follows.

Step S331: determine statistic of testThe first confidence interval and the second confidence interval.

Here, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to the first confidence interval is higher than the significance level (such as respectively 90% and 70%) corresponding to the second confidence interval.

Step S332: if statistic of test is positioned within the first confidence interval, then enter step S333, generate the differentiation result of this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample and go to step S340, otherwise enters step S334.

Step S334: if statistic of test is positioned at outside the second confidence interval, then enter step S335, generates this classification belonging to trade company to be identified and the unmatched differentiation result of classification belonging to trade company's sample and goes to step S340, otherwise enters step S336.

Step S336: calculate the fluctuation dependency of the transaction distribution series of sample set and the transaction distribution series of trade company to be measured according to above formula (3).

Step S337: if the threshold value that fluctuation dependency is default more than or equal to, then enter Step S338, generate the differentiation result of this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample and go to step S340, otherwise, then enter step S339, generate this classification belonging to trade company to be identified and the unmatched differentiation result of classification belonging to trade company's sample and go to step S340.

In the present embodiment, it is also possible to perform comparison routine according to mode as shown in Figure 6.Shown in Fig. 6 and Fig. 5, routine is different in that, step S331A was first carried out before performing step S331, determine whether the title of trade company to be identified exists occurrence in set of keywords, if there is occurrence, then enter step S331B, generate the differentiation result of this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample and go to step S340, otherwise, then entering step S331.

In the above embodiment of the present invention, based on the difference between the consumption time pattern of different industries, identify and receive, by analyzing Trip distribution stability bandwidth and consumption period fluctuation dependency, the violation trade company applying mechanically MCC code in single market.The method provides the two-fold advantage high with accuracy rate that be easy to implement simultaneously.

Although having shown that and describe each exemplary embodiment, but those of ordinary skill in the art are it should be appreciated that the spirit and scope of the various present inventive concept being altered without departing from and being defined by the appended claims can be made in form and details these exemplary embodiments.

Claims (12)

1. one kind is used for differentiating whether a trade company belongs to particular merchant class method for distinguishing, it is characterised in that comprise the following steps:

Receiving the turnover data relative to the time of the multiple trade companies sample constituting a sample set, described trade company sample is positioned at same geographic area and has identical trade company's classification；

The transaction distribution series relative to the time of sample set described in the turnover data genaration of sample set, and by the turnover data genaration of trade company to be identified its relative to the transaction distribution series of time；And

The distribution series of concluding the business of transaction distribution series and the trade company to be identified of sample set comparing to determine, whether classification belonging to this trade company to be identified mates with the classification belonging to trade company sample.

2. the step relative to the transaction distribution series of time the method for claim 1, wherein generating described sample set comprises the following steps:

For each trade company sample in sample set, generate its turnover ordered series of numbers relative to the time；

The turnover ordered series of numbers relative to the time of each trade company sample is carried out standardization processing, and wherein by standardization processing, the span of each data item of turnover ordered series of numbers is between 0-1；

Corresponding data item in the turnover ordered series of numbers of each trade company sample after standardization processing is collected the turnover ordered series of numbers relative to the time obtaining sample set；And

The turnover ordered series of numbers of sample set is carried out standardization processing to obtain the transaction distribution series relative to the time of sample set, wherein at the sample set obtained by standardization processing relative in the transaction distribution series of time, the span of each data item is between 0-1.

3. method as claimed in claim 2, wherein, described standardization carries out as follows:

$x^{\′}=\frac{x-min}{max-min}$

Wherein, x is the value of one of them data item of turnover ordered series of numbers, and x' is this data item value after standardization processing, maximum in max and min respectively turnover ordered series of numbers and minima.

4. the method for claim 1, wherein determine this class belonging to trade company to be identified The step not whether do not mated with the classification belonging to trade company sample comprises the following steps:

Determine the first confidence interval and second confidence interval of following statistic of test:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, then determine this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if described statistic of test is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

5. the method for claim 1, wherein determine that the step whether this classification belonging to trade company to be identified mates with the classification belonging to trade company sample comprises the following steps:

Determining whether the title of trade company to be identified exists occurrence in set of keywords, wherein, described set of keywords is made up of the keyword that the word frequency in the title of the trade company's sample in described sample set is higher；

If there is occurrence, it is determined that this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, otherwise, it is determined that the first confidence interval of following statistic of test and the second confidence interval:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, it is determined that this waits to know Classification belonging to other trade company and categorical match belonging to trade company's sample, if described statistic of test is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

6. the method as described in claim 4 or 5, wherein, farther includes the following step:

If described statistic of test is positioned at outside described first confidence interval and is positioned within described second confidence interval, then calculate the fluctuation dependency of the transaction distribution series of sample set and the transaction distribution series of trade company to be measured as follows:

$r=\frac{1}{n-1}\underset{i=1}{\overset{n}{\Σ}}\left(\frac{X_i-\stackrel{\‾}{X}}{S_X}\right)\left(\frac{Y_i-\stackrel{\‾}{Y}}{S_Y}\right)$

Wherein, r is described fluctuation dependency, X_iAnd Y_iThe respectively i-th data item in the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set,WithThe respectively average of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, S_XAnd S_YThe respectively standard deviation of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set；And

If the threshold value that described fluctuation dependency is default more than or equal to, it is determined that this classification belonging to trade company to be identified and categorical match belonging to trade company's sample, otherwise, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

7. one kind is used for differentiating whether a trade company belongs to the device of particular merchant classification, it is characterised in that including:

I/O unit, its turnover data relative to the time being configured to receive the multiple trade companies sample constituting a sample set and output differentiate result, and described trade company sample is positioned at same geographic area and has identical trade company's classification；

The data processing unit coupled with described I/O unit, comprising:

Ordered series of numbers generation module, it is configured to the transaction distribution series relative to the time of sample set described in the turnover data genaration of sample set and by the turnover data genaration of trade company to be identified accordingly relative to the transaction distribution series of time；And

Comparison module, it is configured to compare to determine the classification belonging to this trade company to be identified and sample institute of trade company by the transaction distribution series of the transaction distribution series of sample set with trade company to be identified Whether the classification belonged to mates.

8. device as claimed in claim 7, wherein, data generation module is configured to generate as follows the transaction distribution series relative to the time of described sample set:

For each trade company sample in sample set, generate its first turnover ordered series of numbers relative to the time；

The turnover ordered series of numbers relative to the time of each trade company sample is carried out standardization processing, and wherein by standardization processing, the span of each data item of turnover ordered series of numbers is between 0-1；

Corresponding data item in the turnover ordered series of numbers of each trade company sample after standardization processing is collected the turnover ordered series of numbers relative to the time obtaining sample set；And

The turnover ordered series of numbers of sample set is carried out standardization processing to obtain the transaction distribution series relative to the time of sample set, wherein at the sample set obtained by standardization processing relative in the transaction distribution series of time, the span of each data item is between 0-1.

9. device as claimed in claim 8, wherein, described standardization carries out as follows:

$x^{\′}=\frac{x-min}{max-min}$

Wherein, x is the value of one of them data item of turnover ordered series of numbers, and x' is this data item value after standardization processing, maximum in max and min respectively turnover ordered series of numbers and minima.

10. device as claimed in claim 7, wherein, comparison module is configured to determine as follows whether classification belonging to trade company to be identified mates with the classification belonging to trade company sample:

Determine the first confidence interval and second confidence interval of following statistic of test:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, then determine this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if described statistic of test is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

11. device as claimed in claim 7, wherein, comparison module is configured to determine as follows whether classification belonging to trade company to be identified mates with the classification belonging to trade company sample:

Determining whether the title of trade company to be identified exists occurrence in set of keywords, wherein, described set of keywords is made up of the keyword that the word frequency in the title of the trade company's sample in described sample set is higher；

If there is occurrence, it is determined that this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, otherwise, it is determined that the first confidence interval of following statistic of test and the second confidence interval:

$\frac{(n-1)S_x^2}{{\mathrm{\σ}}_0^2}$

Wherein, n is the sample size of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, σ² ₀And S² _xThe respectively variance of the transaction distribution series of the transaction variance of distribution series of sample set and trade company to be identified, described statistic of test obeys the χ that degree of freedom is (n-1)²Distribution, the significance level corresponding to described first confidence interval is higher than the significance level corresponding to described second confidence interval；

If described statistic of test is positioned within described first confidence interval, then determine this classification belonging to trade company to be identified and the categorical match belonging to trade company's sample, if described statistic of test is positioned at outside described second confidence interval, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.

12. the device as described in claim 10 or 11, wherein, comparison module is further configured to:

If described statistic of test is positioned at outside described first confidence interval and is positioned within described second confidence interval, then calculate the fluctuation dependency of the transaction distribution series of sample set and the transaction distribution series of trade company to be measured as follows:

$r=\frac{1}{n-1}\underset{i=1}{\overset{n}{\Σ}}\left(\frac{X_i-\stackrel{\‾}{X}}{S_X}\right)\left(\frac{Y_i-\stackrel{\‾}{Y}}{S_Y}\right)$

Wherein, r is described fluctuation dependency, X_iAnd Y_iThe respectively i-th data item in the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set,WithThe respectively average of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set, S_XAnd S_YThe respectively standard deviation of the transaction distribution series of conclude the business distribution series and the trade company to be identified of sample set；And

If the threshold value that described fluctuation dependency is default more than or equal to, it is determined that this classification belonging to trade company to be identified and categorical match belonging to trade company's sample, otherwise, it is determined that this classification belonging to trade company to be identified is not mated with the classification belonging to trade company sample.