CN106327138A - Ship carrier aptitude evaluation method based on big data analysis - Google Patents

Abstract

The invention discloses a ship carrier aptitude evaluation method based on big data analysis. The method comprises the steps that credit indexes capable of comprehensively and objectively reflecting the carrying aptitudes of ship carriers are selected, and an aptitude evaluation index system is established; carrying services of ship carrier members are merged and comprehensively and systematically analyzed on the basis of index association and homogeneous factors to obtain a rank order of the carrying trading aptitudes of the ship carriers, and the ship carrier members are classified according to the carrying association degree of the ship carriers. By means of the aptitude evaluation index system, carrying data is better quantified into the carrying aptitudes of the ship carriers; by ordering and classifying the ship carrier members, the aptitude evaluation problem of cargo owners to the carriers can be effectively solved, a choice meeting the own carrying requirement is made, therefore, risk fluctuation of the waterway freight market is reduced, and the transportation quality is improved.

Description

A kind of boats and ships carrier's qualification evaluation methodology based on big data analysis

Technical field

The invention belongs to big data message analysis technical field, carry particularly to a kind of boats and ships based on big data analysis People's qualification evaluation methodology.

Background technology

Along with the development of information network technique, shipping collection is joined e-commerce platform and is gradually replaced traditional by service high-quality Transport power transaction form.Water-borne transport creates many as a kind of important transportation mode of shipping in transport power transaction change procedure Problem: sailing date delay, loading cost and leakage loss, Claims Resolution confusion etc., this is to join e-commerce platform due to shipping collection promoting transport power to hand over Boats and ships carrier cannot carry while easily expanding qualification effectively assess and cause.Existing logistics transportation ecommerce Platform qualification evaluation model many employings net purchase Website Evaluation mechanism to boats and ships carrier, it is considered to factor is less, easily produces and comments Valency distortion, the problem such as unreal.Therefore, shipping platform need a kind of boats and ships carrier's qualification based on big data analysis objective, Effectively evaluating, system of selection, thus improve shipping mass.

Summary of the invention

Goal of the invention: for problems of the prior art, the present invention provides a kind of for current transport power process of exchange Present in carrier carried the deficiency of quality assessment system, taking into full account that affecting boats and ships carrier carries the multinomial of qualification Key factor, sets up the objective measurement factor that can comprehensively reflect that boats and ships member carries qualification, it is provided that a kind of objective, base of quantization Boats and ships carrier's qualification evaluation methodology in big data analysis.

Technical scheme: for solving above-mentioned technical problem, the present invention provides a kind of boats and ships carrier based on big data analysis Qualification evaluation methodology, comprises the steps:

Step one: extract information from platform database, it is then determined that credit scoring model, finally build indicator evaluation system Layer of structure model；

From data base, selective extraction and carrier's credit appraisal relevant information, need with reference to industry this platform of selecting index Influence factor, according to the mutual relation between factor index and between each level index, that sets up assessment indicator system passs stratum Aggregated(particle) structure model；

Step 2: determine assessment system and characteristic parameter according to Criterion Attribute；

First, choose the boats and ships carrier of platform, determine assessment system S；

Then, choose the index system that can reflect that boats and ships carrier carries qualification, i.e. Life of Ship, boats and ships to insure Situation, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money, ship Oceangoing ship carrier meets with complaining number of times, boats and ships carrier freights on time, boats and ships carrier unloads on time, boats and ships carrier's loading cost and leakage loss ten Individual index is set to characteristic parameter, is X_ij(X_ijFor the jth index of boats and ships i, j=1,2,3 ... 10)；

Step 3: determine optimal sequence according to Criterion Attribute: in judge index boats and ships insure situation, boats and ships acknowledgement of consignment scale, Boats and ships carrier's transaction count, boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money, boats and ships carrier freight on time, Boats and ships carrier seven indexs of unloading on time are forward indexs, are the bigger the better, then the optimal sequence X of its correspondence_ojFor:

X'_oj=max (X '_ij), X_ojFor X_ijCorresponding optimal sequence, X'_ijFor X_ijIn forward index；

In judge index, Life of Ship, boats and ships carrier meet with complaining number of times, boats and ships carrier's loading cost and leakage loss three finger Mark is reverse index, the smaller the better, then the optimal sequence X of its correspondence_ojFor:

X”_oj=min (X_i”_j), X "_ijFor X_ijIn reverse index；

Step 4: data normalization processes: according to the attribute of each characteristic parameter, to participate in the initial data of analysis as the following formula Make Unified Measure to process, and make normalized simultaneously, each analytical data is compressed to [0,1] interval；

It is the index of forward to Criterion Attribute:

X'_ij=X_ij/maxX′_ij

It is reverse index to Criterion Attribute:

X”_ij=minX "_ij/X_ij；

Step 5: set up standardized analysis system after above-mentioned normalized, then sets up the similar square of Lycoperdon polymorphum Vitt Battle array, calculates coefficient of association, the degree of association, incidence matrix and Lycoperdon polymorphum Vitt similar matrix: calculate X_iTo X_pContact in the pass of K point Number ζ_i(k):

X_pK () is reference sequences, X_iK () is comparative sequences.

Calculate X_iTo X_pDegree of association γ_i:

Wherein n is boats and ships carrier's quantity of assessment；

Set up incidence matrix Γ:

γ_ijIt is with jth with the index series of i-th evaluation object for reference sequences The index series of evaluation object is the degree of association of comparative sequences, and m is evaluation object number and optimal sequence sum, i.e. m=n+1；

Set up Lycoperdon polymorphum Vitt similar matrix G；

Wherein g_ij=(γ_ij+γ_ji)/2, γ_ijAnd γ_jiFor in incidence matrix Γ Symmetrical item；

Step 6: inteerelated order arranges: in step 5, G matrix last column is i.e. with optimal sequence (X_oj) it is that reference sequences is asked The inteerelated order obtained, by g_m1, g_m2……g_mm(g_mm=1) sorting by size, the qualification that the order of its correspondence is assessment boats and ships is good and bad Sequentially；

Step 7: arranged the qualification obtaining assessing boats and ships by the inteerelated order in step 6 and sort, from Lycoperdon polymorphum Vitt similar matrix G The middle value taking out correspondence, the similarity formation sequence Z between the most adjacent sequence boats and ships:

Z=(Z₁,Z₂…Z_n-1), Z_iIt is the similarity between i-th and i+1 boats and ships.

Step 8: cluster analysis: set platform to ship classification boundary as λ (λ ∈ [0,1]), similar by step 7 Property sequence Z compare with λ, with λ for boundary, the acknowledgement of consignment qualification of assessment boats and ships is carried out good and bad level and divides.

Further, in step one, hierarchical structure model includes:

Destination layer: be used for setting decision objective and carry quality assessment as boats and ships carrier；

Rule layer: the quality assessment of boats and ships basic condition, the quality assessment of boats and ships carrier's trading situation and boats and ships acknowledgement of consignment The quality assessment of people's service quality；

Solution layer: choose Life of Ship, situation insured by boats and ships, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, Conclude the business tonnage, boats and ships carrier's dealing money, boats and ships carrier of boats and ships carrier meets with complaining number of times, boats and ships carrier to fill on time Goods, boats and ships carrier unload on time, ten evaluation indexes of boats and ships carrier's loading cost and leakage loss are as solution layer；

Boats and ships carrier in described destination layer carries quality assessment and is commented by the qualification of the boats and ships basic condition in rule layer Estimate, the quality assessment of the quality assessment of boats and ships carrier's trading situation and boats and ships carrier's service quality is constituted；

Boats and ships basic condition quality assessment in described rule layer is insured feelings by the Life of Ship in solution layer, boats and ships Condition and boats and ships acknowledgement of consignment scale are constituted；

The quality assessment of the boats and ships carrier's trading situation in described rule layer is concluded the business by the boats and ships carrier in solution layer Conclude the business tonnage, boats and ships carrier's dealing money and boats and ships carrier of number of times, boats and ships carrier meets with complaining number of times to constitute；

The quality assessment of the boats and ships carrier's service quality in described rule layer is punctual by the boats and ships carrier in solution layer Freight, boats and ships carrier unloads on time, boats and ships carrier's loading cost and leakage loss is constituted.

Compared with prior art, it is an advantage of the current invention that: the freight transport quality appraisal procedure of the present invention, can will carry number It is further used as the acknowledgement of consignment qualification of boats and ships carrier, the collating sort root of boats and ships carrier member according to being quantified as freight transport quality Can effectively solve owner of cargo's quality assessment problem to carrier according to the quality of shipping, and make the choosing meeting oneself acknowledgement of consignment demand Select, thus reduce the risk fluctuation of freight market, water route, improve shipping mass.

Accompanying drawing explanation

Fig. 1 is the method flow diagram of the present invention；

Fig. 2 is the hierarchical structure model that the present invention executes example.

Detailed description of the invention

Below in conjunction with the accompanying drawings and detailed description of the invention, it is further elucidated with the present invention.

A kind of boats and ships carrier's qualification evaluation methodology method based on big data analysis, selects based on principle comprehensive, objective Take foundation and can reflect that boats and ships carrier carries the credit appraisal system of qualification, it is adaptable to the ecommerce of logistics transport power transaction is put down Car and boat member is carried out comprehensively by platform, the selection of science.Specifically include following steps:

(1) extract platform database information, set up the recursive hierarchy structure of index

From data base, selective extraction and carrier's credit appraisal relevant information, need with reference to industry this platform of selecting index Influence factor, according to the mutual relation between factor index and between each level index, that sets up assessment indicator system passs stratum Aggregated(particle) structure.Destination layer: set decision objective and carry quality assessment as boats and ships carrier.Rule layer: the qualification of boats and ships basic condition Assessment, the quality assessment of boats and ships carrier's trading situation and the quality assessment of boats and ships carrier's service quality.Solution layer: choose ship Insure situation, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier of oceangoing ship service life, boats and ships concludes the business tonnage, ship Oceangoing ship carrier's dealing money, boats and ships carrier complained number of times, boats and ships carrier freights on time, boats and ships carrier unloads on time, Ten evaluation indexes of boats and ships carrier's loading cost and leakage loss are as solution layer.

(2) assessment system and characteristic parameter are determined according to Criterion Attribute

1. choose the boats and ships carrier of platform, determine assessment system S:

S=(S₁,S₂,S₃,…,S_n), wherein n is boats and ships carrier's numbers

2. choose the index system that can reflect that boats and ships carrier carries qualification, i.e. Life of Ship, boats and ships are insured feelings Condition, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money, boats and ships Carrier meets with complaining number of times, boats and ships carrier freights on time, boats and ships carrier unloads on time, boats and ships carrier's loading cost and leakage loss ten Index is set to characteristic parameter, is:

X_i=(X_i1,X_i2,X_i3,,X_i3,X_i4,X_i5,X_i6,X_i7,X_i8,X_i9,X_i10), X_iIndex set for boats and ships carrier i Close.

(3) determine optimal sequence according to Criterion Attribute and be normalized

1. in judge index boats and ships insure situation, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier hand over Easily tonnage, boats and ships carrier's dealing money, boats and ships carrier freight on time, boats and ships carrier seven indexs of unloading on time are forwards Index, is the bigger the better, then the optimal sequence X of its correspondence_ojFor:

X′_oj=max (X '_ij), X_ojFor X_ijCorresponding optimal sequence, X '_ijFor X_ijIn forward index.

2. in judge index, Life of Ship, boats and ships carrier meet with complaining number of times, boats and ships carrier's loading cost and leakage loss three Index is reverse index, the smaller the better, then the optimal sequence X of its correspondence_ojFor:

X″_oj=min (X "_ij), X "_ijFor X_ijIn reverse index.

3. according to the attribute of each characteristic parameter, with participate in the initial data of analysis make as the following formula Unified Measure process, and with Time make normalized, each analytical data is compressed to [0,1] interval.

It is the index of forward to Criterion Attribute:

X′_ij=X_ij/maxX′_ij

It is reverse index to Criterion Attribute:

X″_ij=minX "_ij/X_ij

Through above-mentioned normalized, it is established that standardized analysis system.

(4) Lycoperdon polymorphum Vitt similar matrix is set up

1. X is calculated_iTo X_oCoefficient of association ζ at K point_i(k):

X_pK () is reference sequences, X_iK () is comparative sequences.

2. X is calculated_iTo X_pDegree of association γ_i:

${\mathrm{\γ}}_i=\frac{1}{n}\underset{j=1}{\overset{n}{\Σ}}{\mathrm{\ξ}}_i\left(k\right)$

3. incidence matrix Γ is set up:

γ_ijIt is with the index series of i-th evaluation object as reference sequences, with jth The index series of individual evaluation object is the degree of association of comparative sequences, and m is evaluation object number and optimal sequence sum, i.e. m=n+ 1。

4. Lycoperdon polymorphum Vitt similar matrix G is set up.

Wherein g_ij=(γ_ij+γ_ji)/2, γ_ijAnd γ_jiFor incidence matrix Γ In symmetrical item.

G matrix last column is i.e. with optimal sequence (X_oj) it is the inteerelated order tried to achieve of reference sequences, by g_m1, g_m2……g_mm (g_mm=1) sorting by size, the order of its correspondence is the qualification order of quality of assessment boats and ships.

(5) cluster analysis

Inteerelated order sequence i.e. obtains assessing the qualification sequence of boats and ships, takes out the value of correspondence, i.e. phase from Lycoperdon polymorphum Vitt similar matrix G Similarity formation sequence Z between adjacent sequence boats and ships:

Z=(Z₁,Z₂…Z_n-1), Z_iIt is the similarity between i-th and i+1 boats and ships.

Set platform to ship classification boundary as λ (λ ∈ [0,1]), similarity sequence Z is compared with λ, with λ as boundary The acknowledgement of consignment qualification of assessment boats and ships is carried out good and bad level divide.

It is different from present most shipping collection to join e-commerce platform and use subjective assessment that carrier's qualification is estimated also And classification mode, this patent utilize big data analysis technique to carrier reality acknowledgement of consignment data carry out comprehensive, objectively divide Analyse, and carrier is classified by the characteristic shown from data itself, joins e-commerce platform for shipping collection and provides one Plant comprehensive, the determination methods of science.

According to a specific embodiment, step is as follows:

(1) extract platform database information, set up the recursive hierarchy structure of index

From data base, selective extraction and carrier's credit appraisal relevant information, need with reference to industry this platform of selecting index Influence factor, according to the mutual relation between factor index and between each level index, that sets up assessment indicator system passs stratum Aggregated(particle) structure, such as accompanying drawing 2.Destination layer: set decision objective and carry quality assessment as boats and ships carrier.Rule layer: the basic feelings of boats and ships The quality assessment of condition, the quality assessment of boats and ships carrier's trading situation and the quality assessment of boats and ships carrier's service quality.Scheme Layer: choose Life of Ship, situation insured by boats and ships, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier hand over Easily tonnage, boats and ships carrier's dealing money, boats and ships carrier meet with complaining number of times, boats and ships carrier to freight on time, boats and ships carrier Punctual unloading, ten evaluation indexes of boats and ships carrier's loading cost and leakage loss are as solution layer.

(2) assessment system and characteristic parameter are determined according to Criterion Attribute

1. according to originally executing example and choose 10 boats and ships carriers of platform, assessment system S is determined:

S=(S₁,S₂,S₃,…,S₁₀),

2. the quality assessment of boats and ships basic condition of evaluation object, the quality assessment of boats and ships carrier's trading situation and boats and ships The quality assessment data of carrier's service quality are as shown in table 1:

Table 1 is executed example boats and ships carrier and is carried quality assessment

Life of Ship, boats and ships insure situation, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier Transaction tonnage, boats and ships carrier's dealing money, boats and ships carrier meet with complaining number of times, boats and ships carrier freights on time, boats and ships are carried People unloads on time, ten indexs of boats and ships carrier's loading cost and leakage loss are set to characteristic parameter composing indexes set: X_i=(X_i1,X_i2, X_i3,,X_i3,X_i4,X_i5,X_i6,X_i7,X_i8,X_i9,X_i10), X_iIndex set for boats and ships carrier i.

(3) determine optimal sequence according to Criterion Attribute and be normalized

1. in judge index boats and ships insure situation, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier hand over Easily tonnage, boats and ships carrier's dealing money, boats and ships carrier freight on time, boats and ships carrier seven indexs of unloading on time are forwards Index, is the bigger the better, then the optimal sequence X of its correspondence_ojFor:

X′_oj=max (X '_ij), X_ojFor X_ijCorresponding optimal sequence, X '_ijFor X_ijIn forward index.

2. in judge index, Life of Ship, boats and ships carrier meet with complaining number of times, boats and ships carrier's loading cost and leakage loss three Index is reverse index, the smaller the better, then the optimal sequence X of its correspondence_ojFor:

X″_oj=min (X "_ij), X "_ijFor X_ijIn reverse index.

Then optimal sequence is:

X_o=(X_o1,X_o2,X_o3,,X_o3,X_o4,X_o5,X_o6,X_o7,X_o8,X_o9,X_o10)=(2,60,5000,12,170,700,2, 9,10,2)

3. according to the attribute of each characteristic parameter, with participate in the initial data of analysis make as the following formula Unified Measure process, and with Time make normalized, each analytical data is compressed to [0,1] interval.

It is the index of forward to Criterion Attribute:

X′_ij=X_ij/maxX′_ij

It is reverse index to Criterion Attribute:

X″_ij=minX "_ij/X_ij

Through above-mentioned normalized, it is established that standardized analysis system.

System data analyzed by table 2

	Xi1	Xi2	Xi3	Xi4	Xi5	Xi6	Xi7	Xi8	Xi9	Xi10
											X1	0.4000	0.8333	0.2000	0.6667	0.7500	0.5000	1.0000	0.6667	0.6923	1.0000
X2	0.3333	0.6667	0.3000	0.7333	0.8000	0.6000	0.5000	0.7500	0.6154	0.5000
											X3	0.2857	1.0000	0.3200	0.8000	0.9000	0.6500	0.5000	0.8333	0.7692	0.5000
X4	0.2500	0.7500	0.2800	0.5333	0.4500	0.4500	1.0000	0.5833	0.6154	1.0000
											X5	0.3333	0.5000	0.2600	0.4000	0.3500	0.4000	0.5000	0.3333	0.3077	1.0000
X6	0.4000	0.9167	0.4000	0.4667	0.7000	0.5000	0.3333	0.5000	0.5385	1.0000
											X7	0.5000	0.7500	0.5000	0.6667	0.7750	0.5500	1.0000	0.7500	0.6154	1.0000
X8	0.6667	0.8333	0.6000	1.0000	1.0000	1.0000	0.2500	1.0000	1.0000	0.3333
											X9	0.4000	0.6667	0.6400	0.8667	0.9000	0.8000	1.0000	0.8333	0.8462	0.5000
X10	1.0000	1.0000	1.0000	0.8000	0.8500	0.7000	0.5000	0.7500	0.7692	0.5000
											X0	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000	1.0000

(4) Lycoperdon polymorphum Vitt similar matrix is set up

The most respectively with X₁,X₂,X₃,X₄,X₅,X₆,X₇,X₈,X₉,X₁₀,X₀For reference sequences, with X₁,X₂,X₃,X₄,X₅,X₆,X₇, X₈,X₉,X₁₀,X₀For comparative sequences, calculate coefficient of association ζ_i(k):

With X₁For reference sequences, X₁、X₂、X₃……X₁₀、X₀During for comparative sequences,

Coefficient of association is:

$\mathrm{\ζ}=\left\{\begin{array}{cccccccccc}1.0000& 1.0000& 1.0000& 1.0000& 1.0000& 1.0000& 1.0000& 1.0000& 1.0000& 1.0000\\ 0.8571& 0.7059& 0.8000& 0.8571& 0.8889& 0.8000& 0.4444& 0.8276& 0.8387& 0.4444\\ 0.7778& 0.7059& 0.7692& 0.7500& 0.7273& 0.7273& 0.4444& 0.7059& 0.8387& 0.4444\\ 0.7273& 0.8276& 0.8333& 0.7500& 0.5714& 0.8889& 1.0000& 0.8276& 0.8387& 1.0000\\ 0.8571& 0.5455& 0.8696& 0.6000& 0.5000& 0.8000& 0.4444& 0.5455& 0.5098& 1.0000\\ 1.0000& 0.8276& 0.6667& 0.6667& 0.8889& 1.0000& 0.3750& 0.7059& 0.7222& 1.0000\\ 0.8000& 0.8276& 0.5714& 1.0000& 0.9412& 0.8889& 1.0000& 0.8276& 0.8387& 1.0000\\ 0.6000& 1.0000& 0.5000& 0.5455& 0.6154& 0.4444& 0.3478& 0.5455& 0.5652& 0.3750\\ 1.0000& 0.7059& 0.4762& 0.6667& 0.7273& 0.5714& 1.0000& 0.7059& 0.7222& 0.4444\\ 0.4000& 0.7059& 0.3333& 0.7500& 0.8000& 0.6667& 0.4444& 0.8276& 0.8387& 0.4444\\ 0.4000& 0.7059& 0.3333& 0.5455& 0.6154& 0.4444& 1.0000& 0.5455& 0.5652& 1.0000\end{array}\right\}$

The rest may be inferred by analogy.

2. degree of association γ is calculated_i:

${\mathrm{\γ}}_i=\frac{1}{n}\underset{j=1}{\overset{n}{\Σ}}{\mathrm{\ξ}}_i\left(k\right)$

3. incidence matrix Γ is set up:

$\mathrm{\Γ}=\left\{\begin{array}{ccccccccccc}1.0000& 0.7464& 0.6891& 0.8265& 0.6672& 0.7853& 0.8695& 0.5539& 0.7020& 0.6211& 0.6155\\ 0.7239& 1.0000& 0.8334& 0.6899& 0.6561& 0.6817& 0.7593& 0.5719& 0.7308& 0.7377& 0.4744\\ 0.6662& 0.8357& 1.0000& 0.6173& 0.5878& 0.6471& 0.6471& 0.6213& 0.7589& 0.8243& 0.5728\\ 0.8180& 0.7019& 0.6274& 1.0000& 0.7355& 0.7504& 0.7985& 0.4748& 0.6191& 0.5203& 0.5500\\ 0.6524& 0.6660& 0.5947& 0.7333& 1.0000& 0.7186& 0.6092& 0.4358& 0.5038& 0.4914& 0.4372\\ 0.7604& 0.6719& 0.6323& 0.7312& 0.6997& 1.0000& 0.7233& 0.5244& 0.5512& 0.5558& 0.4978\\ 0.8631& 0.7689& 0.6574& 0.7985& 0.6121& 0.7429& 1.0000& 0.5696& 0.7204& 0.6309& 0.6160\\ 0.5398& 0.5884& 0.6323& 0.4748& 0.4390& 0.5507& 0.5696& 1.0000& 0.6787& 0.6139& 0.7399\\ 0.6904& 0.7419& 0.7657& 0.6191& 0.5069& 0.5769& 0.7204& 0.6787& 1.0000& 0.7022& 0.6434\\ 0.6211& 0.7552& 0.8335& 0.5353& 0.5085& 0.5967& 0.6435& 0.6288& 0.7129& 1.0000& 0.7104\\ 0.6155& 0.5065& 0.5958& 0.5624& 0.4537& 0.5356& 0.6286& 0.7474& 0.6561& 0.7104& 1.0000\end{array}\right\}$

4. Lycoperdon polymorphum Vitt similar matrix G is set up.

$G=\left(\begin{array}{ccccccccccc}1.0000& & & & & & & & & & \\ 0.7352& 1.0000& & & & & & & & & \\ 0.6777& 0.8345& 1.0000& & & & & & & & \\ 0.8223& 0.6959& 0.6224& 1.0000& & & & & & & \\ 0.6598& 0.6611& 0.5913& 0.7344& 1.0000& & & & & & \\ 0.7728& 0.6768& 0.6397& 0.7408& 0.7091& 1.0000& & & & & \\ 0.8663& 0.7641& 0.6522& 0.7985& 0.6107& 0.7331& 1.0000& & & & \\ 0.5469& 0.5801& 0.6268& 0.4748& 0.4374& 0.5375& 0.5696& 1.0000& & & \\ 0.6962& 0.7363& 0.7623& 0.6191& 0.5054& 0.5640& 0.7204& 0.6787& 1.0000& & \\ 0.6211& 0.7465& 0.8289& 0.5278& 0.4999& 0.5762& 0.6372& 0.6213& 0.7075& 1.0000& \\ 0.6155& 0.4904& 0.5843& 0.5562& 0.4454& 0.5167& 0.6223& 0.7437& 0.6497& 0.7104& 1.0000\end{array}\right)$

, wherein g_ij=(γ_ij+γ_ji)/2。

G matrix last column is i.e. with optimal sequence (X_o) it is the inteerelated order tried to achieve of reference sequences, it is worth big I by it To obtain the good and bad sequence of evaluation object:

1.0000>0.7437>0.7104>0.6497>0.6223>0.6155>0.5843>0.5562>0.5167>0.4904 > 0.4454, obtain each assessment boats and ships quality sequence:

S8—S10—S9—S7—S1—S3—S4—S6—S2—S5

(5) cluster analysis

Inteerelated order sequence i.e. obtains assessing the qualification sequence of boats and ships, takes out the value of correspondence, i.e. phase from Lycoperdon polymorphum Vitt similar matrix G Similarity formation sequence Z between adjacent sequence boats and ships:

Z=(Z₁,Z₂…Z_n-1)=(0.6213,0.7075,0.7204,0.8663,0.6777,0.6224,0.7408, 0.6768,0.6611)

0.62130.70750.72040.86630.67770.62240.74080.67680.6611

S8——S10——S9——S7——S1———S3———S4———S6———S2——S5

λ=0.7000, compares similarity sequence Z with λ, and Z is divided into five classes: Z with λ for boundary₁=(0.6213)； (Z₂, Z₃, Z₄)=(0.7075,0.7204,0.8663)；(Z₅, Z₆)=(0.6777,0.6224)；Z₇=(0.7408)；(Z₈, Z₉) 10 boats and ships carriers can be divided into five grades by=(0.6768,0.6611): (S8, S10)；(S9、S7、S1)；(S3、 S4)；(S6)；(S2、S5).

The foregoing is only embodiments of the invention, be not limited to the present invention.All principles in the present invention Within, the equivalent made, should be included within the scope of the present invention.The content that the present invention is not elaborated belongs to In prior art known to this professional field technical staff.

Claims (2)

1. boats and ships carrier's qualification evaluation methodology based on big data analysis, it is characterised in that comprise the steps:

Step one: extract information from platform database, it is then determined that credit scoring model, finally build the knot of indicator evaluation system Structure hierarchical model；

Selective extraction and carrier's credit appraisal relevant information, the shadow needed with reference to industry this platform of selecting index from data base The factor of sound, according to the mutual relation between factor index and between each level index, sets up the Recurison order hierarchy knot of assessment indicator system Structure model；

Step 2: determine assessment system and characteristic parameter according to Criterion Attribute；

First, choose the boats and ships carrier of platform, determine assessment system S；

Then, choose the index system that can reflect that boats and ships carrier carries qualification, i.e. Life of Ship, boats and ships are insured feelings Condition, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money, boats and ships Carrier meets with complaining number of times, boats and ships carrier freights on time, boats and ships carrier unloads on time, boats and ships carrier's loading cost and leakage loss ten Index is set to characteristic parameter, is X_ij(X_ijFor the jth index of boats and ships i, j=1,2,3 ... 10)；

Step 3: determine optimal sequence according to Criterion Attribute: in judge index, boats and ships are insured situation, boats and ships acknowledgement of consignment scale, boats and ships Carrier's transaction count, boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money, boats and ships carrier freight on time, boats and ships Carrier's seven indexs of unloading on time are forward indexs, are the bigger the better, then the optimal sequence X of its correspondence_ojFor:

X'_oj=max (X '_ij), X_ojFor X_ijCorresponding optimal sequence, X'_ijFor X_ijIn forward index；

Life of Ship in judge index, boats and ships carrier complained number of times, three indexs of boats and ships carrier's loading cost and leakage loss are Reverse index, the smaller the better, then the optimal sequence X of its correspondence_ojFor:

X”_oj=min (X_i'_j'), X "_ijFor X_ijIn reverse index；

Step 4: data normalization processes: according to the attribute of each characteristic parameter, unites as the following formula participating in the initial data of analysis One estimates process, and makees normalized simultaneously, each analytical data is compressed to [0,1] interval；

It is the index of forward to Criterion Attribute:

X'_ij=X_ij/maxX′_ij

It is reverse index to Criterion Attribute:

X”_ij=minX "_ij/X_ij；

Step 5: set up standardized analysis system after above-mentioned normalized, then sets up Lycoperdon polymorphum Vitt similar matrix, right Coefficient of association, the degree of association, incidence matrix and Lycoperdon polymorphum Vitt similar matrix calculate: calculate X_iTo X_pCoefficient of association ζ at K point_i (k):

X_pK () is reference sequences, X_i(k) For comparative sequences.

Calculate X_iTo X_pDegree of association γ_i:

Wherein n is boats and ships carrier's quantity of assessment；

Set up incidence matrix Γ:

γ_ijIt is to assess with jth with the index series of i-th evaluation object for reference sequences The index series of object is the degree of association of comparative sequences, and m is evaluation object number and optimal sequence sum, i.e. m=n+1；

Set up Lycoperdon polymorphum Vitt similar matrix G；

Wherein g_ij=(γ_ij+γ_ji)/2, γ_ijAnd γ_jiFor the symmetry in incidence matrix Γ ?；

Step 6: inteerelated order arranges: in step 5, G matrix last column is i.e. with optimal sequence (X_oj) it is that reference sequences is tried to achieve Inteerelated order, by g_m1, g_m2……g_mm(g_mm=1) sorting by size, the qualification quality that the order of its correspondence is assessment boats and ships is suitable Sequence；

Step 7: arranged the qualification obtaining assessing boats and ships by the inteerelated order in step 6 and sort, take from Lycoperdon polymorphum Vitt similar matrix G Go out the value of correspondence, the similarity formation sequence Z between the most adjacent sequence boats and ships:

Z=(Z₁,Z₂…Z_n-1), Z_iIt is the similarity between i-th and i+1 boats and ships.

Step 8: cluster analysis: set platform to ship classification boundary as λ (λ ∈ [0,1]), by the similarity sequence in step 7 Row Z with λ compares, and with λ for boundary, the acknowledgement of consignment qualification of assessment boats and ships is carried out good and bad level and divides.

A kind of boats and ships carrier's qualification evaluation methodology based on big data analysis the most according to claim 1, its feature exists In, in described step one, hierarchical structure model includes:

Destination layer: be used for setting decision objective and carry quality assessment as boats and ships carrier；

Rule layer: the quality assessment of boats and ships basic condition, the quality assessment of boats and ships carrier's trading situation and boats and ships carrier clothes The quality assessment of business quality；

Solution layer: choose Life of Ship, situation insured by boats and ships, boats and ships acknowledgement of consignment scale, boats and ships carrier's transaction count, boats and ships Carrier conclude the business tonnage, boats and ships carrier's dealing money, boats and ships carrier meet with complaining number of times, boats and ships carrier to freight on time, ship Oceangoing ship carrier unloads on time, ten evaluation indexes of boats and ships carrier's loading cost and leakage loss are as solution layer；

Boats and ships carrier in described destination layer carries quality assessment by the quality assessment of boats and ships basic condition in rule layer, ship The quality assessment of oceangoing ship carrier's trading situation and the quality assessment of boats and ships carrier's service quality are constituted；

Boats and ships basic condition quality assessment in described rule layer by the Life of Ship in solution layer, boats and ships insure situation and Boats and ships acknowledgement of consignment scale is constituted；

The quality assessment of the boats and ships carrier's trading situation in described rule layer by the boats and ships carrier's transaction count in solution layer, Boats and ships carrier conclude the business tonnage, boats and ships carrier's dealing money and boats and ships carrier meet with complain number of times constitute；

The quality assessment of the boats and ships carrier's service quality in described rule layer freighted on time by the boats and ships carrier in solution layer, Boats and ships carrier unloads on time, boats and ships carrier's loading cost and leakage loss is constituted.