CN107798571A - Identifying system, the method and device of malice address/malice order - Google Patents
Identifying system, the method and device of malice address/malice order Download PDFInfo
- Publication number
- CN107798571A CN107798571A CN201610797563.7A CN201610797563A CN107798571A CN 107798571 A CN107798571 A CN 107798571A CN 201610797563 A CN201610797563 A CN 201610797563A CN 107798571 A CN107798571 A CN 107798571A
- Authority
- CN
- China
- Prior art keywords
- address
- identified
- order
- probability
- malice
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0609—Buyer or seller confidence or verification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
- G06Q30/0637—Approvals
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Computer And Data Communications (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention discloses a kind of identifying system, the method and device of malice address/malice order, is related to Internet technical field, can solve the problem that and identifies the problem of malice address/malice order accuracy rate is relatively low in the prior art.The method of the present invention mainly includes:Receive the address to be identified that subscription client is sent;The hierarchical processing in address is carried out to the address to be identified, obtains each address level of the address to be identified;Probability distribution is redirected using the address level for analyzing to obtain by history normal address, calculate the probability that redirects that each address level in the address to be identified jumps to adjacent next address level, what the address level redirected that probability distribution includes that any one address level jumps to another address level redirects probability;Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address to be identified.
Description
Technical field
The present invention relates to Internet technical field, more particularly to a kind of malice address/identifying system of malice order, side
Method and device.
Background technology
With the development of Internet technology, people can not only realize viewing video by network, browse webpage, chat etc.
Operation, can also be done shopping, and realize that the operating process of shopping is also very convenient.
However, in actual applications, but it occur frequently that some buyers are by deliberately filling in incomplete ship-to, mistake
The malicious act such as ship-to commodity is not sent to, thus bring economic loss, the phenomenon of credibility loss to businessman, because
This, how to identify malice address is extremely important to businessman.The mode of existing identification malice address mainly has three kinds:(1) lead to
Cross and matched address to be identified with default malice keyword, to determine whether address to be identified is malice address;(2) pass through
Address to be identified is matched respectively with the address in black and white lists, to determine whether address to be identified is malice address;(3)
By carrying out hierarchical structuring division to address to be identified, then matched with preset address hierarchical structure, to determine to wait to know
Whether other address is malice address.
Although above-mentioned three kinds of modes can identify part malice address to a certain extent, None- identified goes out one
Hiding malice address a bit, or normal address may be mistaken for malice address.For example, for same keyword,
It may be malice keyword in an address, but may be normal keyword in another address, if therefore by the key
Word is identified as default malice address, then is likely to occur the phenomenon that normal address is mistaken for malice address.And for example, due to
Black and white lists are the lists for the manual maintenance that the actual feedback after being delivered according to businessman is carried out, so being known using black and white lists
Not only need to consume manpower otherwise, new malice address can't be identified in time.For another example, for some address level knots
Structure is complete but the address that is not present in actual life, if be identified using preset address hierarchical structure, can be judged by accident
For normal address.Therefore, identify that the accuracy rate of malice address is relatively low in the prior art, so that identification malice order is accurate
Rate is relatively low.
The content of the invention
In view of this, the present invention provides a kind of identifying system, the method and device of malice address/malice order, can solve
The problem of malice address/malice order accuracy rate is relatively low is certainly identified in the prior art.
In a first aspect, the invention provides a kind of identifying system of malice address, the system includes subscription client, clothes
Business device and merchant client;Wherein,
The subscription client is used for the address to be identified for receiving input, and the address to be identified is sent into the clothes
Business device;
The server is used to receiving the address to be identified that the subscription client is sent, and to described to be identified
Location carries out the hierarchical processing in address, obtains each address level of the address to be identified;Analyzed using by history normal address
To address level redirect probability distribution, calculate each address level in the address to be identified and jump to adjacent next address
Level redirects probability, and the address level redirects probability distribution and jumps to another address layer including any one address level
Level redirects probability;Multiplication processing is carried out to each probability that redirects of acquisition, the normal address for obtaining the address to be identified is general
Rate, and the recognition result that malice Address Recognition is carried out based on the normal address probability is sent to the merchant client;
The merchant client is used to receiving and exporting the recognition result that the server is sent.
Second aspect, the invention provides a kind of recognition methods of malice address, methods described includes:
Receive the address to be identified that subscription client is sent;
The hierarchical processing in address is carried out to the address to be identified, obtains each address level of the address to be identified;
Probability distribution is redirected using the address level for analyzing to obtain by history normal address, is calculated in the address to be identified
Each address level jumps to the probability that redirects of adjacent next address level, and the address level, which redirects probability distribution, to be included appointing
What one address level of meaning jumped to another address level redirects probability;
Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address to be identified.
The third aspect, the invention provides a kind of identification device of malice address, described device includes:
Receiving unit, for receiving the address to be identified of subscription client transmission;
First processing units, for carrying out the hierarchical processing in address to the address to be identified, acquisition is described to be identifiedly
Each address level of location;
Computing unit, for redirecting probability distribution using the address level for analyzing to obtain by history normal address, calculate institute
The probability that redirects that each address level in address to be identified jumps to adjacent next address level is stated, the address level redirects
What probability distribution included that any one address level jumps to another address level redirects probability;
Second processing unit, each probability that redirects for being obtained to the computing unit carry out multiplication processing, obtain institute
State the normal address probability of address to be identified.
Fourth aspect, the invention provides a kind of identifying system of malice order, the system includes subscription client, clothes
Business device and merchant client;Wherein,
The subscription client is used for the order to be identified for receiving input, and the order to be identified is sent into the clothes
Business device;
The server is used to receive the order to be identified that the subscription client is sent, and based on normal by history
The address level that adress analysis obtains redirects probability distribution, calculates each address level in the address of the order to be identified and redirects
To the probability that redirects of adjacent next address level, the address level, which redirects probability distribution, includes any one address level jump
Go to another address level redirects probability;Multiplication processing is carried out to each probability that redirects of acquisition, obtains the address
Normal address probability;Whether it is malice order according to order to be identified described in the normal address probabilistic determination, and will determine that knot
Fruit is sent to the merchant client;
The merchant client is used for the judged result for receiving and showing that the server is sent.
5th aspect, the invention provides a kind of recognition methods of malice order, methods described includes:
Receive the order to be identified that subscription client is sent;
Probability distribution is redirected based on the address level for analyzing to obtain by history normal address, calculates the order to be identified
Each address level jumps to the probability that redirects of adjacent next address level in address, and the address level redirects probability distribution
Probability is redirected including what any one address level jumped to another address level;
Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address;
Whether it is malice order according to order to be identified described in the normal address probabilistic determination.
6th aspect, the invention provides a kind of identification device of malice order, described device includes:
Receiving unit, for receiving the order to be identified of subscription client transmission;
Computing unit, for redirecting probability distribution based on the address level for analyzing to obtain by history normal address, calculate institute
State the probability that redirects that each address level in the address of order to be identified jumps to adjacent next address level, the address layer
What level redirected that probability distribution includes that any one address level jumps to another address level redirects probability;
Processing unit, for carrying out multiplication processing to each probability that redirects of acquisition, obtain the normal address of the address
Probability;
Judging unit, whether it is malice order for the order to be identified according to the normal address probabilistic determination.
By above-mentioned technical proposal, identifying system, the method and device of malice address/malice order provided by the invention,
Can after the address level that server obtains address to be identified and is analyzed to obtain by history normal address redirects probability distribution,
The hierarchical processing in address first is carried out to the address to be identified, obtains each address level of the address to be identified, then utilizes acquisition
Address level redirect probability distribution, calculate each address level in the address to be identified and jump to adjacent next address level
Redirect probability, and multiplication processing is carried out to each probability that redirects, obtains the probability that the address to be identified belongs to normal address, with
Just whether it is malice address according to probabilistic determination address to be identified, or the address to be identified is included according to the probabilistic determination
Whether order is malice order.It follows that with coarse filtration in the prior art pass through malice keyword, black and white lists or address
Hierarchical structure judges whether address to be identified is that malice address is compared, and the present invention is by each address layer in history normal address
Correlation is counted and analyzed between level, and judges that redirecting for each address level in address to be identified is general using analysis result
Rate, then the probability for being obtained whole address to be identified by redirecting probability and being belonged to normal address, so as to obtain comprising malice
The normal address probability of the address of keyword, the normal address probability of the address included in black and white lists and address level knot
The normal address probability of structure sufficient address, additionally it is possible to obtain the normal address probability, no of the address not comprising malice keyword
The normal address probability of address included in black and white lists and the normal address of the incomplete address of address hierarchical structure are general
Rate, and can determine whether address to be identified is malice address according to the normal address probability, so as to according to whether being malice
Address determines whether order to be identified is malice order, and then improves the accuracy rate of malice address/malice order identification.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention,
And can be practiced according to the content of specification, and in order to allow above and other objects of the present invention, feature and advantage can
Become apparent, below especially exemplified by the embodiment of the present invention.
Brief description of the drawings
By reading the detailed description of hereafter preferred embodiment, it is various other the advantages of and benefit it is common for this area
Technical staff will be clear understanding.Accompanying drawing is only used for showing the purpose of preferred embodiment, and is not considered as to the present invention
Limitation.And in whole accompanying drawing, identical part is denoted by the same reference numerals.In the accompanying drawings:
Fig. 1 shows a kind of identifying system schematic diagram of malice address provided in an embodiment of the present invention;
Fig. 2 shows a kind of merchant client side selection interface schematic diagram provided in an embodiment of the present invention;
Fig. 3 shows a kind of flow chart of the recognition methods of malice address provided in an embodiment of the present invention;
Fig. 4 shows the flow chart of the recognition methods of another malice address provided in an embodiment of the present invention;
Fig. 5 shows server and the interaction figure of client in malice address learning process provided in an embodiment of the present invention;
Fig. 6 shows a kind of composition frame chart of the identification device of malice address provided in an embodiment of the present invention;
Fig. 7 shows the composition frame chart of the identification device of another malice address provided in an embodiment of the present invention;
Fig. 8 shows a kind of flow chart of the recognition methods of malice order provided in an embodiment of the present invention;
Fig. 9 shows a kind of composition frame chart of the identification device of malice order provided in an embodiment of the present invention;
Figure 10 shows the composition frame chart of the identification device of another malice order provided in an embodiment of the present invention.
Embodiment
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although the disclosure is shown in accompanying drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
Limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
Completely it is communicated to those skilled in the art.
In order to improve the accuracy rate of identification malice address, the embodiments of the invention provide a kind of identification system of malice address
System, as shown in figure 1, system includes subscription client 11, server 12 and merchant client 13;Wherein,
Subscription client 11 is used for the address to be identified for receiving input, and address to be identified is sent into server 12;
Server 12 is used for the address to be identified for receiving the transmission of subscription client 11, and carries out address layer to address to be identified
Levelization processing, obtain each address level of address to be identified;Redirected using the address level for analyzing to obtain by history normal address
Probability distribution, calculate that each address level in address to be identified jumps to adjacent next address level redirects probability, address
What level redirected that probability distribution includes that any one address level jumps to another address level redirects probability;To each of acquisition
The individual probability that redirects carries out multiplication processing, obtains the normal address probability of address to be identified, and will be based on normal address probability and carry out
The recognition result of malice Address Recognition is sent to merchant client 13;
Merchant client 13 is used for the recognition result received and export server 12 is sent.
The identifying system of malice address provided in an embodiment of the present invention, subscription client can be received in server and sent
Address to be identified after, the hierarchical processing in address first is carried out to the address to be identified, obtains each address layer of the address to be identified
Level, then redirect probability distribution using address level, calculate each address level in the address to be identified jump to it is adjacent under
One address level redirects probability, and carries out multiplication processing to each probability that redirects, and obtains the address to be identified and belongs to normally
The probability of location, whether to be malice address according to probabilistic determination address to be identified.It follows that with coarse filtration in the prior art
Ground judges whether address to be identified is that malice address is compared by malice keyword, black and white lists or address hierarchical structure,
The present invention by the way that correlation between each address level in history normal address is counted and analyzed, and using analysis result come
Judge the probability that redirects of each address level in address to be identified, then belong to normal address by redirecting the whole address to be identified of probability acquisition
Probability, so that the normal address probability of the address comprising malice keyword can not only be obtained, included in black and white lists
The normal address probability of address and the normal address probability of address hierarchical structure sufficient address, additionally it is possible to obtain not including and dislike
Normal address probability, the normal address probability and address layer of the address being not included in black and white lists of the address of meaning keyword
The normal address probability of the incomplete address of level structure, and whether address to be identified can be determined according to the normal address probability
For malice address, so as to improve the accuracy rate of malice Address Recognition.
Further, server 12 is used for when it is malice address that recognition result, which is address to be identified, to merchant client
13 send early warning information;
Merchant client 13 is used for the early warning information received and export server 12 is sent.
Further, merchant client 13 is used for after early warning information is received, and exports for selecting to be identified
The selection interface for the recognition result that address is recognized, and receive the identification of based on selection interface input, secondary identification
As a result, the recognition result of secondary identification is returned into server 12.
Exemplary, as shown in Fig. 2 after merchant client receives early warning information, not only can be in interface display
The early warning information, a selection interface that secondary recognition result is selected for businessman can be also shown, as can in the selection interface
To have a content of text " buyer please be contact and reaffirm whether address * * * are malice address ", and two select buttons
"Yes" and "No", selected for user.
It should be noted that early warning information can be located at selection interface, another interface can also be located at.
Further, merchant client 13 is used in the case where not receiving early warning information, exports for selecting
The selection interface for the recognition result being recognized to address to be identified, and receive based on selection interface input, for retouching
The recognition result that address to be identified is malice address is stated, and the address to be identified for carrying malice mark is returned into server 12.
Further, a kind of malice address is additionally provided according to said system embodiment, an alternative embodiment of the invention
Recognition methods, as shown in figure 3, this method mainly includes:
201st, the address to be identified that subscription client is sent is received.
After user places an order successfully, order can be uploaded to server by subscription client (i.e. buyer client), service
After device receives the order, malice Address Recognition operation can be carried out to the order, and by order and the recognition result of order
Merchant client is sent to, so that businessman carries out respective handling according to recognition result to the order.Received due to server
Some nonsensical data in order to be identified often be present, so in order to prevent these data from disturbing address to be identified
Identification, after order to be identified is obtained, server needs first to pre-process the order to be identified, then again after pretreatment
Order to be identified in extract address to be identified.
Therefore, obtaining the specific implementation process of address to be identified can be:Obtain order to be identified;Order to be identified is entered
The processing of row redundancy and formatting processing;Address to be identified is obtained from the order to be identified after processing.
Wherein, redundancy processing is carried out to order to be identified and formatting processing specifically includes:
(1) to meeting that the word of default filter condition filters in the address to be identified of order to be identified.
Because user may fill in some emoticons, insignificant English alphabet and some other nothing in the address
The data of meaning, it is possible to detect in address to be identified whether contain these information, carried out these information if containing
Filter.
(2) dirty data in order to be identified is filtered.
Because server is when preserving order to be identified, some may be preserved and include HTML (HyperText Markup
Language, HTML) abnormal information such as text, JSON (JavaScript Object Notation) character string
Dirty data, so server can be filtered these dirty datas.
(3) according to predetermined formatting processing rule, processing is formatted to the order to be identified after filtering.
Because user is in information such as fill address, phones, may add space, using the complex form of Chinese characters, using phonetic etc.,
So for the ease of can subsequently accurately identify address to be identified, after being filtered to order to be identified, it is also necessary to gone
Except space, the conversion of full-shape half-angle, either traditional and simplified characters conversion, phonetic turn the format manipulations such as Chinese character, so that the address obtained has
Unified form.
It should be noted that when analyzing history normal address and history malice address, be equally also required into
The above-mentioned pretreatment operation of row.
202nd, the hierarchical processing in address is carried out to address to be identified, obtains each address level of address to be identified.
Because each level of address is only relevant with neighbouring last layer level, and unrelated with other levels, so address layer
Level structure meets Markov property, so as to carry out the hierarchical processing in address using conditional random field models.Wherein, knowledge is treated
The specific implementation that other address carries out the hierarchical processing in address is:After address to be identified is obtained, server can pass through bar
Part random field models are segmented to address to be identified, address level mark, so as to obtain each address layer of address to be identified
Level.For example, address to be identified is Sichuan Province Chengdu Qingyang District Su Po streets * * homes the 5th unit of building 1, then each address level is divided
It is not:" save:Sichuan Province, city:Chengdu, district:Qingyang District, road:Su Po streets, cell:* homes, Lou Hao:No. 5 building, list
Member number:Unit 1 ".
203rd, probability distribution is redirected using the address level for analyzing to obtain by history normal address, calculated in address to be identified
What each address level jumped to adjacent next address level redirects probability.
Wherein, address level, which redirects probability distribution, includes the jump that any one address level jumps to another address level
Turn probability.Because history normal address is the successful address of businessman's delivery, so server is obtaining substantial amounts of history normally
Behind location, situation can be redirected to the address level of history normal address and be counted and analyzed, therefrom obtained address level and redirect
Probability distribution, probability distribution is redirected to will pass through address layer level and determines to redirect feelings between each address level of address to be identified
Condition.
After each address level of address to be identified is obtained, server can redirect probability distribution calculating using address level
Go out neighbor address level in address to be identified redirects probability, i.e. n-th layer level jumps to the probability of N+1 levels.For example, obtaining
Each address level for obtaining address to be identified " saves:Sichuan Province, city:Chengdu, district:Qingyang District, road:It is Su Po streets, small
Area:* homes, Lou Hao:No. 5 building, unit number:After Unit 1 ", probability distribution can be redirected using address level, obtains " Sichuan Province
Jump to the probability in Chengdu, Chengdu jumps to the probability of Qingyang District, Qingyang District jumps to the probability in Su Po streets, Su Po streets
Road jumps to the probability at * * homes, * * homes jump to the probability in No. 5 building and No. 5 building jumps to the probability of Unit 1 ".
204th, multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of address to be identified.
When the address level to history normal address redirects probability and is trained, can utilize in the whole country in large quantities
Location is trained, and obtain each address level in address to be identified redirect probability after, these can be redirected probability multiplication, obtained
Obtain the probability that address to be identified belongs to normal address.
In actual applications, some malice addresses are probably to be slapped together by multiple different places orders inside the province, for example,
Shenzhen * * * Co., Ltds in the business parking of Zhabei District in Shanghai Longgang Township town Long Gang streets, wherein, " Shanghai City
Zhabei District " belongs to Shanghai City address, " Shenzhen * * * Co., Ltds in the business parking of Longgang Township town Long Gang streets " category
In Guangdong Province address.Therefore, when being trained using nationwide a large amount of normal addresses, only Zhabei District to Longgang District this
Redirecting between individual address level is abnormal, and other are all normally to redirect, so as to which the whole address obtained belongs to normal address
Probability it is larger, and then be mistaken for normal address;And if instructed using only magnanimity history normal address in Shanghai City
Practice, then whole address only have Shanghai City to Zhabei District redirect be it is normal, and redirecting between other address levels be all extremely
, so as to obtain whole address belong to malice address probability it is larger, and then be defined as malice address.Therefore, increasing
After this variable of province, the accuracy rate of malice Address Recognition is improved.
In actual applications, after this variable of province is increased, the probability meter that address to be identified belongs to normal address is calculated
Calculating formula can be:
Wherein, S represents address to be identified,
wiThe i-th address level in address to be identified is represented, C represents the affiliated province in address to be identified.
The recognition methods of malice address provided in an embodiment of the present invention, first this can be treated after address to be identified is obtained
Identify that address carries out the hierarchical processing in address, obtain each address level of the address to be identified, then redirected using address level
Probability distribution, the probability that redirects that each address level in the address to be identified jumps to adjacent next address level is calculated, and
Multiplication processing is carried out to each probability that redirects, the probability that the address to be identified belongs to normal address is obtained, so as to according to the probability
Judge whether address to be identified is malice address.It follows that with coarse filtration in the prior art pass through malice keyword, black and white name
List or address hierarchical structure judge whether address to be identified is that malice address is compared, and the present invention is by history normal address
In between each address level correlation counted and analyzed, and each address level in address to be identified is judged using analysis result
Redirect probability, then the probability for being obtained whole address to be identified by redirecting probability and being belonged to normal address, so as to obtain
The normal address probability of address comprising malice keyword, the normal address probability and ground of the address included in black and white lists
The normal address probability of location hierarchical structure sufficient address, additionally it is possible to obtain the normal address of the address not comprising malice keyword
Probability, the normal address probability for the address being not included in black and white lists and the incomplete address of address hierarchical structure it is normal
Address probability, and can determine whether address to be identified is malice address according to the normal address probability, disliked so as to improve
The accuracy rate for Address Recognition of anticipating.
Further, after obtaining address to be identified and belonging to the probability of normal address, can according to default recognition rule with
And the normal address probability of address to be identified, judge whether address to be identified is malice address.
Specifically, after the normal address probability of address to be identified is obtained, normal address probabilistic determination can be directly utilized
Whether address to be identified is malice address, and other features corresponding to address to be identified can also be analyzed, then according to just
Normal address probability and other characteristic synthetics judge whether address to be identified is malice address (such as the institute of following step 305 to 307
State).Wherein, whether it is that the specific implementation of malice address is directly using normal address probabilistic determination address to be identified:Sentence
Whether the normal address probability of disconnected address to be identified is more than predetermined probabilities threshold value;If the normal address probability of address to be identified is more than
Predetermined probabilities threshold value, it is determined that address to be identified is normal address;If the normal address probability of address to be identified is less than or equal to
Predetermined probabilities threshold value, it is determined that address to be identified is malice address.
In addition, after server obtains recognition result, the recognition result can be sent to merchant client, so as to businessman
Client receives and shows recognition result, determines whether to deliver according to recognition result for businessman.
Further, a kind of knowledge of malice address is additionally provided according to above-described embodiment, an alternative embodiment of the invention
Other method, as shown in figure 4, this method mainly includes:
301st, the address to be identified that subscription client is sent is received..
302nd, the hierarchical processing in address is carried out to address to be identified, obtains each address level of address to be identified.
303rd, probability distribution is redirected using the address level for analyzing to obtain by history normal address, calculated in address to be identified
What each address level jumped to adjacent next address level redirects probability.
304th, multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of address to be identified.
305th, extract and use from History Order corresponding to order to be identified corresponding to address to be identified and/or order to be identified
In identify address to be identified whether be malice address default identification feature.
Specifically, default identification feature includes following any one or any several combination:Address text message is special
Sign, history Shopping Behaviors feature, order feature and cross feature.
Accordingly, this step can specifically be refined as following step a-d:
(a) the address text message feature corresponding to extraction from address to be identified.
Wherein, text message feature in address includes:Whether include the numeral of preset length, whether include default sensitive word with
And whether including advertising message etc..Wherein, preset length includes cell-phone number length, home Tel length, QQ number length etc..
Due to user may in order to abuse businessman or for oneself, the commodity of oneself advertise and fill in one in the address
Information, cell-phone number, advertising message etc. are abused, and the user for filling in these contents may fill in a malice address, institute
So that address text message feature can be extracted from address to be identified, whether to be malice from dimensional analysis address to be identified
Address.
(b) history Shopping Behaviors feature is extracted from History Order corresponding to order to be identified.
Because the history Shopping Behaviors of user can reflect whether it may fill in a malice address, for example, often with
Businessman occur dispute, often without reason reimbursement, the relatively low user of success rate of merchandising fill in malice address possibility it is larger, and never
With businessman occur dispute, never moved back money, the higher user of transaction success rate fill in malice address possibility it is smaller, so can
To extract history Shopping Behaviors feature from History Order corresponding to order to be identified, and this feature is to be identified as judging
Location whether be malice address a dimension.
In addition, in actual applications, history Shopping Behaviors feature mainly includes:It is pay invoice number in preset time period, pre-
If total payment, the reimbursement initiation total amount in preset time period, transaction success rate, preset time in preset time period in the period
Dispute businessman number in section, initiation rate is complained in preset time period, reimbursement dispute accounting etc. in preset time period.Wherein, Ge Geli
The preset time period of history Shopping Behaviors feature can be with identical, can also be different.
(c) the order feature corresponding to extraction from order to be identified.
Specifically, order feature includes:The use time of whether normal, the to be identified address of telephone number in order to be identified
Whether number is more than default using the correlation behavior in shop corresponding to threshold value, order to be identified and order to be identified to corresponding business
The correlation behavior of product.Wherein, the correlation behavior in shop includes:The opening time in shop, the ripple that shop is scored in the nearest period
Dynamic, shop is by number of malicious attack etc.;The correlation behavior of commodity includes:Whether the sales volume of commodity, the price of commodity, commodity are hot
Door etc..
Because user is in fill address, the telephone number of mistake may be deliberately filled in, or fills in and never used
New address, and malicious act is often focused on big businessman or much-sought-after item, so server can be ordered to be identified
These order features are singly extracted, and analyze whether address to be identified is malice address by this dimension of order feature.
(d) according to the normal of address text message feature, history Shopping Behaviors feature, order feature and address to be identified
The combination of at least two in the probability of address, obtains cross feature corresponding to address to be identified.
In actual applications, by address text message feature, history Shopping Behaviors feature, order feature and to be identified
The normal address probability of location these essential characteristics carry out combined crosswise, can produce more abstract feature description, such as by address
Text message feature carries out combined crosswise with order feature, can obtain in address to be identified not only not comprising insignificant text
This description (does not have the information such as portable phone number, QQ number, default sensitive word, advertising message) in address, and the address is use
The conventional address at family.Therefore, can be using cross feature corresponding to address to be identified as another dimension for identifying malice address.
306th, the default identification model trained by History Order is obtained.
Specifically, the implementation of the default identification model of server training can be:First obtain History Order;Then basis
Address level redirects probability distribution, obtains the normal address probability of the historical address carried in History Order;Again from History Order
The middle default identification feature of extraction;Normal address probability and corresponding default identification feature instruction finally by each historical address
Practice default identification model.
Wherein, History Order includes the normal order of history and history malice order of preset ratio, and when history is normal
The ratio of order and history malice order is about 4:When 1, the accuracy rate of malice Address Recognition is of a relatively high.
It should be noted that in actual applications, the default identification model that this step needs to train can be GBDT
(Gradient Boosting Decision Tree, gradient lifting decision tree) model, or other models, such as SVM
(Support Vector Machine, SVMs) model, LR (Logistic Regression, logistic regression) model,
Neural network model etc..
307th, according to the normal address probability of address to be identified, default identification feature and default identification model, judge to treat
Identify whether address is malice address.
After the normal address probability of address to be identified and default identification feature is obtained, server can be special by these
Sign, which is input in default identification model, to be identified, and can be carried out comprehensive analysis to these features to preset identification model, be obtained
Address to be identified finally belongs to the probability of normal address or the probability of malice address, and according to default normal probability threshold value or
Person presets malice address probability threshold value to determine whether the address to be identified is malice address.
If the 308, judging address to be identified for malice address, early warning information is sent to merchant client, so as to business
Family's client receives and exports early warning information.
After server judges address to be identified for malice address, in order to avoid businessman because malice address and caused by it is economical,
Prestige etc. is lost, and server can be sent to be used to indicate address while order to be identified is sent to merchant client
May be the early warning information of malice address, can be according to the phone in order after businessman receives the early warning information
Contacted with buyer, so as to judge the address whether really for malice address;If businessman determines that the address is malice address,
It can refuse to deliver, if it is normal address that businessman, which determines the address, rather than malice address, then it can trust delivery.
In addition, if server judges that the address to be identified for normal address, can only send to merchant client and wait to know
Other order, without sending early warning information;, can be direct when the order that businessman has found to receive does not have early warning information
Delivered address in order.However, malice address may be mistaken for normal address by server, therefore, work as business
When family finds that address can not be sent in actual releases, businessman can select address as malice address in merchant client
Button, so that the address to be identified for carrying malice mark is sent to server by merchant client, and received in server
Behind the address to be identified for carrying malice mark, history normal address storehouse, history malice address base are updated, and to default
Identification model carries out re -training.
309th, knowledge that merchant client is sent, being recognized based on early warning information to address to be identified is received
Other result.
In actual applications, when it is malice address that businessman, which determines address to be identified, can early warning instrument the page (or
The selection interface referred in person's said system embodiment) in select for indicating to be defined as the button of malice address, so as to businessman
The address to be identified for carrying malice mark is sent to server by client;When businessman determine address to be identified be normal address and
During non-malicious address, it can select to be defined as the button of normal address for instruction in the page of early warning instrument, so as to businessman
The address to be identified that carrying normally identifies is sent to server by client.
If the 310th, recognition result be address to be identified be normal address, more new historical normal address storehouse, history maliciously
Location storehouse and default identification model.
When the recognition result for the secondary identification that merchant client is sent is that address to be identified is normal address, then server is true
Its fixed misjudgment, and history normal address storehouse, history malice address base are updated immediately, probability point is then redirected to address level
Cloth is reanalysed, and re -training is carried out to default identification model.
In addition, by taking GBDT models as an example, interaction in the embodiment of the present invention between server and client can be as
Shown in Fig. 5, and by above-described embodiment, it is preliminary that the embodiment of the present invention can not only redirect probability distribution based on address level
Obtain the probability that address to be identified belongs to normal address, additionally it is possible to obtain address text from History Order and order to be identified
Other default identification features such as information characteristics, history Shopping Behaviors feature, order feature and cross feature, and by be identified
It is comprehensive that the normal address probability of location and these default identification features input into GBDT models (or other identification models) progress
Analysis is closed, judges whether address to be identified is malice address, so as to further increase the accuracy rate of malice Address Recognition.This
Outside, when it is malice address that server, which finally determines the address to be identified, additionally it is possible to send early warning letter to merchant client
Breath, so that businessman can examine whether the address is malice address by being contacted with buyer, to decide whether to deliver, and then
Avoid producing loss.Further, after businessman determines according to actual conditions whether address is malice address, additionally it is possible in businessman
Confirming button corresponding to selection in client, so that merchant client determines that result feeds back to server by actual, so as to service
Device can determine if to judge by accident according to the feedback of merchant client, can be in time to GBDT moulds if judging by accident
Type carries out re -training, makes GBDT models more perfect, and which thereby enhance the accuracy rate of follow-up malice Address Recognition.
Further, a kind of malice address is additionally provided according to above method embodiment, an alternative embodiment of the invention
Identification device, as shown in fig. 6, the device mainly includes:Receiving unit 41, first processing units 42, computing unit 43, second
Processing unit 44.Wherein,
Receiving unit 41, for receiving the address to be identified of subscription client transmission;
First processing units 42, for carrying out the hierarchical processing in address to address to be identified, obtain each of address to be identified
Address level;
Computing unit 43, for redirecting probability distribution using the address level for analyzing to obtain by history normal address, calculate
Each address level jumps to the probability that redirects of adjacent next address level in address to be identified, and address level redirects probability point
What cloth included that any one address level jumps to another address level redirects probability;
Second processing unit 44, each probability that redirects for being obtained to computing unit 43 carry out multiplication processing, treated
Identify the normal address probability of address.
The identification device of malice address provided in an embodiment of the present invention, first this can be treated after address to be identified is obtained
Identify that address carries out the hierarchical processing in address, obtain each address level of the address to be identified, then redirected using address level
Probability distribution, the probability that redirects that each address level in the address to be identified jumps to adjacent next address level is calculated, and
Multiplication processing is carried out to each probability that redirects, the probability that the address to be identified belongs to normal address is obtained, so as to according to the probability
Judge whether address to be identified is malice address.It follows that with coarse filtration in the prior art pass through malice keyword, black and white name
List or address hierarchical structure judge whether address to be identified is that malice address is compared, and the present invention is by history normal address
In between each address level correlation counted and analyzed, and each address level in address to be identified is judged using analysis result
Redirect probability, then the probability for being obtained whole address to be identified by redirecting probability and being belonged to normal address, so as to obtain
The normal address probability of address comprising malice keyword, the normal address probability and ground of the address included in black and white lists
The normal address probability of location hierarchical structure sufficient address, additionally it is possible to obtain the normal address of the address not comprising malice keyword
Probability, the normal address probability for the address being not included in black and white lists and the incomplete address of address hierarchical structure it is normal
Address probability, and can determine whether address to be identified is malice address according to the normal address probability, disliked so as to improve
The accuracy rate for Address Recognition of anticipating.
Further, as shown in fig. 7, the device also includes:
Judging unit 45, for after the normal address probability of the address to be identified is obtained, being advised according to default identification
Then and address to be identified normal address probability, judge whether address to be identified is malice address.
Further, as shown in fig. 7, judging unit 45 includes:
Extraction module 451, for from history corresponding to order to be identified corresponding to address to be identified and/or order to be identified
Extracted in order for identify address to be identified whether be malice address default identification feature;
Acquisition module 452, for obtaining the default identification model trained by History Order;
First judge module 453, for the normal address probability according to address to be identified, default identification feature and preset
Identification model, judge whether address to be identified is malice address.
Further, as shown in fig. 7, extraction module 451 includes:
First extracting sub-module 4511, for address text message feature corresponding to the extraction from address to be identified;
Second extracting sub-module 4512, it is special for extracting history Shopping Behaviors from History Order corresponding to order to be identified
Sign;
3rd extracting sub-module 4513, for order feature corresponding to the extraction from order to be identified.
Further, as shown in fig. 7, extraction module 451 also includes:
Acquisition submodule 4514, for according to address text message feature, history Shopping Behaviors feature, order feature and
The combination of at least two, obtains cross feature corresponding to address to be identified in the normal address probability of address to be identified.
Further, the address text message feature of the first extracting sub-module 4511 extraction includes:Whether default length is included
The numeral of degree, whether include default sensitive word and whether include advertising message;
The order feature of 3rd extracting sub-module 4513 extraction includes:Whether telephone number in order to be identified normal,
Whether the access times of address to be identified, which are more than, is preset using the correlation behavior in shop corresponding to threshold value, order to be identified and treats
Identify the correlation behavior of order commodity corresponding to.
Further, acquisition module 452 is additionally operable to obtain History Order, and History Order is including the history of preset ratio just
Normal order and history malice order;
Acquisition module 452 is additionally operable to redirect probability distribution according to conditional random field models and address level, obtains history
The normal address probability of the historical address carried in order;
Extraction module 451 is additionally operable to extract default identification feature from History Order;
As shown in fig. 7, judging unit 45 also includes:
Training module 454, for the normal address probability by each historical address and corresponding default identification feature
The default identification model of training.
Further, as shown in fig. 7, judging unit 45 includes:
Second judge module 455, for judging whether the normal address probability of address to be identified is more than predetermined probabilities threshold value;
Determining module 456 is big for the normal address probability of address to be identified for the judged result when the second judge module
When predetermined probabilities threshold value, it is normal address to determine address to be identified, when the judged result of the second judge module is to be identified
When the normal address probability of location is less than or equal to predetermined probabilities threshold value, it is malice address to determine address to be identified.
Further, as shown in fig. 7, the device also includes:
First transmitting element 46, for will determine that whether address to be identified is that the recognition result of malice address is sent to businessman
Client, so that merchant client receives and exports recognition result.
Further, as shown in fig. 7, the device also includes:
Second transmitting element 47, for when judging unit 45 judges address to be identified for malice address, to businessman client
End sends early warning information, so that the merchant client receives and exports the early warning information;
Receiving unit 41, for receive merchant client transmission, based on early warning information to address to be identified carry out
The recognition result of secondary identification;
First updating block 48, the recognition result for being received when the first receiving unit 48 are addresses to be identified for normally
During location, more new historical normal address storehouse, history malice address base and default identification model.
Further, receiving unit 41, the address to be identified of the carrying malice mark for receiving merchant client transmission;
As shown in fig. 7, the device also includes:
Second updating block 49, for more new historical normal address storehouse, history malice address base and default identification model.
Further, address to be identified is to carry out redundancy processing and form to order to be identified in first processing units 42
The address obtained after change processing.
Further, as shown in fig. 7, first processing units 42 include:
Filtering module 421, for meeting that the word of default filter condition is carried out in the address to be identified to order to be identified
Filtering;
Filtering module 421 is additionally operable to filter the dirty data in order to be identified;
Processing module 422, for according to predetermined formatting processing rule, to be identified ordering after being filtered to filtering module 421
Singly it is formatted processing.
The identification device of malice address provided in an embodiment of the present invention, probability distribution can not only be redirected based on address level
Tentatively obtain the probability that address to be identified belongs to normal address, additionally it is possible to obtain address from History Order and order to be identified
Other default identification features such as text message feature, history Shopping Behaviors feature, order feature and cross feature, and will wait to know
The normal address probability of other address and these default identification features input and comprehensive analysis are carried out into default identification model, judge
Whether address to be identified is malice address, so as to further increase the accuracy rate of malice Address Recognition.In addition, work as server most
When to determine the address to be identified eventually be malice address, additionally it is possible to early warning information is sent to merchant client, so that business
Family can examine whether the address is malice address by being contacted with buyer, to decide whether to deliver, and then avoid producing loss.
Further, after businessman determines according to actual conditions whether address is malice address, additionally it is possible to selected in merchant client
Corresponding confirming button, so that merchant client determines that result feeds back to server by actual, so as to which server can be according to business
The feedback of family's client is determined if to judge by accident, if judging by accident, default identification model can be carried out again in time
New training, make default identification model more perfect, and which thereby enhance the accuracy rate of follow-up malice Address Recognition.
Further, in order to improve the accuracy rate of identification malice order, an alternative embodiment of the invention provides one kind
The identifying system of malice order, the system include subscription client, server and merchant client;Wherein,
Subscription client is used for the order to be identified for receiving input, and order to be identified is sent into server;
Server is used for the order to be identified for receiving subscription client transmission, and is based on analyzing to obtain by history normal address
Address level redirect probability distribution, calculate each address level in the address of order to be identified and jump to adjacent next address
Level redirects probability, and address level redirects probability distribution and jumps to another address level including any one address level
Redirect probability;Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of address;According to normal address
Whether probabilistic determination order to be identified is malice order, and will determine that result is sent to merchant client;
Merchant client is used for the judged result received and display server is sent.
The identifying system of malice order provided in an embodiment of the present invention, treating for subscription client transmission is received in server
After identifying order, probability distribution is redirected first with address level, the address calculated in the order to be identified belongs to normal address
Probability, then recycles whether the probabilistic determination order to be identified is malice order.It follows that with coarse filtration in the prior art
Ground judges whether address to be identified is that malice address is compared by malice keyword, black and white lists or address hierarchical structure,
The present invention by the way that correlation between each address level in history normal address is counted and analyzed, and using analysis result come
Judge the probability that redirects of each address level in address to be identified, then belong to normal address by redirecting the whole address to be identified of probability acquisition
Probability, so that the normal address probability of the address comprising malice keyword can not only be obtained, included in black and white lists
The normal address probability of address and the normal address probability of address hierarchical structure sufficient address, additionally it is possible to obtain not including and dislike
Normal address probability, the normal address probability and address layer of the address being not included in black and white lists of the address of meaning keyword
The normal address probability of the incomplete address of level structure, and whether be malice address according to normal address determine the probability address, from
And the accuracy rate of identification malice address is improved, and then improve the accuracy rate of identification malice order.
Further, the identifying system according to the malice order referred in above-described embodiment, another implementation of the invention
Example provides a kind of recognition methods of malice order, as shown in figure 8, this method mainly includes:
501st, the order to be identified that subscription client is sent is received.
After user places an order successfully, order can be uploaded to server by subscription client, and server receives the order
Afterwards, malice Address Recognition operation can be carried out to the order.
502nd, probability distribution is redirected based on the address level for analyzing to obtain by history normal address, calculates order to be identified
What each address level jumped to adjacent next address level in address redirects probability.
Wherein, address level, which redirects probability distribution, includes the jump that any one address level jumps to another address level
Turn probability.
Specifically, server first can carry out the hierarchical processing in address to the address of order to be identified, each of address is obtained
Address level (refers to above-mentioned steps 202);It is then based on address level and redirects probability distribution, calculates each address level and jump to
Adjacent next address level redirects probability (referring to above-mentioned steps 203).
503rd, multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of address.
The specific implementation of this step is identical with above-mentioned steps 204, will not be repeated here.
504th, whether it is malice order according to normal address probabilistic determination order to be identified.
Specifically, whether server can be first maliciously according to the address of normal address probabilistic determination order to be identified
Location;If the address of order to be identified is malice address, it is determined that order to be identified is malice order;If the address of order to be identified
For normal address, it is determined that order to be identified is normal order.
Wherein, according to the address of normal address probabilistic determination order to be identified whether be malice address specific implementation
It is identical with the specific implementation in the embodiment of above-mentioned " recognition methods of malice address ", it will not be repeated here.
Further, due in actual applications, malicious user except by way of crossing addition malice address to businessman
Bring puzzlement outer, perplex businessman by other means toward contact, such as fill in be engaged in telephone number so that businessman can not be with it
Contacted, so when normal address is in the address for judging order to be identified, it is also necessary to judge the electricity in order to be identified again
Whether normal talk about number.If telephone number is abnormal, it is determined that order to be identified is malice order;If telephone number is normal, really
Fixed order to be identified is normal order.
Wherein, judge whether abnormal method can be telephone number:A normal telephone number storehouse is built, will be to be identified
Telephone number is matched with normal telephone number storehouse, if it fails to match, it is determined that telephone number to be identified is abnormal, if matching into
Work(, it is determined that telephone number to be identified is normal.
The recognition methods of malice order provided in an embodiment of the present invention, treating for subscription client transmission is received in server
After identifying order, probability distribution is redirected first with address level, the address calculated in the order to be identified belongs to normal address
Probability, then recycles whether the probabilistic determination order to be identified is malice order.It follows that with coarse filtration in the prior art
Ground judges whether address to be identified is that malice address is compared by malice keyword, black and white lists or address hierarchical structure,
The present invention by the way that correlation between each address level in history normal address is counted and analyzed, and using analysis result come
Judge the probability that redirects of each address level in address to be identified, then belong to normal address by redirecting the whole address to be identified of probability acquisition
Probability, so that the normal address probability of the address comprising malice keyword can not only be obtained, included in black and white lists
The normal address probability of address and the normal address probability of address hierarchical structure sufficient address, additionally it is possible to obtain not including and dislike
Normal address probability, the normal address probability and address layer of the address being not included in black and white lists of the address of meaning keyword
The normal address probability of the incomplete address of level structure, and whether be malice address according to normal address determine the probability address, from
And the accuracy rate of identification malice address is improved, and then improve the accuracy rate of identification malice order.
Further, a kind of knowledge of malice order is provided according to the method shown in Fig. 8, an alternative embodiment of the invention
Other device, as shown in figure 9, the device mainly includes:
Receiving unit 61, for receiving the order to be identified of subscription client transmission;
Computing unit 62, for redirecting probability distribution based on the address level for analyzing to obtain by history normal address, calculate
Each address level jumps to the probability that redirects of adjacent next address level in the address of order to be identified, and address level redirects
What probability distribution included that any one address level jumps to another address level redirects probability;
Processing unit 63, for carrying out multiplication processing to each probability that redirects of acquisition, the normal address for obtaining address is general
Rate;
Judging unit 64, for whether being malice order according to normal address probabilistic determination order to be identified.
Further, as shown in Figure 10, judging unit 64 includes:
Judge module 641, whether it is malice address for the address according to normal address probabilistic determination order to be identified;
Determining module 642, for when the address of order to be identified is malice address, determining that order to be identified is ordered for malice
It is single.
Further, judge module 641 is additionally operable to, when the address of order to be identified is normal address, judge to be identified order
Whether the telephone number in list is normal;
Determining module 642 is additionally operable to when telephone number exception, and it is malice order to determine order to be identified.
Further, as shown in Figure 10, computing unit 62 includes:
Processing module 621, for carrying out the hierarchical processing in address to the address of order to be identified, obtain each address of address
Level;
Computing module 622, for redirecting probability distribution based on address level, calculate each address level jump to it is adjacent
Next address level redirects probability.
The identification device of malice order provided in an embodiment of the present invention, treating for subscription client transmission is received in server
After identifying order, probability distribution is redirected first with address level, the address calculated in the order to be identified belongs to normal address
Probability, then recycles whether the probabilistic determination order to be identified is malice order.It follows that with coarse filtration in the prior art
Ground judges whether address to be identified is that malice address is compared by malice keyword, black and white lists or address hierarchical structure,
The present invention by the way that correlation between each address level in history normal address is counted and analyzed, and using analysis result come
Judge the probability that redirects of each address level in address to be identified, then belong to normal address by redirecting the whole address to be identified of probability acquisition
Probability, so that the normal address probability of the address comprising malice keyword can not only be obtained, included in black and white lists
The normal address probability of address and the normal address probability of address hierarchical structure sufficient address, additionally it is possible to obtain not including and dislike
Normal address probability, the normal address probability and address layer of the address being not included in black and white lists of the address of meaning keyword
The normal address probability of the incomplete address of level structure, and whether be malice address according to normal address determine the probability address, from
And the accuracy rate of identification malice address is improved, and then improve the accuracy rate of identification malice order.
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and does not have the portion being described in detail in some embodiment
Point, it may refer to the associated description of other embodiment.
It is understood that the correlated characteristic in the above method, apparatus and system can be referred to mutually.In addition, above-mentioned reality
It is to be used to distinguish each embodiment to apply " first " in example, " second " etc., and does not represent the quality of each embodiment.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, the corresponding process in preceding method embodiment is may be referred to, will not be repeated here.
Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein.
Various general-purpose systems can also be used together with teaching based on this.As described above, required by constructing this kind of system
Structure be obvious.In addition, the present invention is not also directed to any certain programmed language.It should be understood that it can utilize various
Programming language realizes the content of invention described herein, and the description done above to language-specific is to disclose this hair
Bright preferred forms.
In the specification that this place provides, numerous specific details are set forth.It is to be appreciated, however, that the implementation of the present invention
Example can be put into practice in the case of these no details.In some instances, known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help to understand one or more of each inventive aspect,
Above in the description to the exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention:I.e. required guarantor
The application claims of shield features more more than the feature being expressly recited in each claim.It is more precisely, such as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following embodiment are expressly incorporated in the embodiment, wherein each claim is in itself
Separate embodiments all as the present invention.
Those skilled in the art, which are appreciated that, to be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more equipment different from the embodiment.Can be the module or list in embodiment
Member or component be combined into a module or unit or component, and can be divided into addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit exclude each other, it can use any
Combination is disclosed to all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so to appoint
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (including adjoint power
Profit requires, summary and accompanying drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation
Replace.
In addition, it will be appreciated by those of skill in the art that although some embodiments described herein include other embodiments
In included some features rather than further feature, but the combination of the feature of different embodiments means in of the invention
Within the scope of and form different embodiments.For example, in the following claims, embodiment claimed is appointed
One of meaning mode can use in any combination.
The all parts embodiment of the present invention can be realized with hardware, or to be run on one or more processor
Software module realize, or realized with combinations thereof.It will be understood by those of skill in the art that it can use in practice
Microprocessor or digital signal processor (DSP) realize the knowledge of malice address/malice order according to embodiments of the present invention
The some or all functions of some or all parts in other system, method and device.The present invention is also implemented as using
In some or all equipment or program of device for performing method as described herein (for example, computer program and meter
Calculation machine program product).Such program for realizing the present invention can store on a computer-readable medium, or can have one
The form of individual or multiple signals.Such signal can be downloaded from internet website and obtained, or above be carried in carrier signal
For, or provided in the form of any other.
It should be noted that the present invention will be described rather than limits the invention for above-described embodiment, and ability
Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of some different elements and being come by means of properly programmed computer real
It is existing.In if the unit claim of equipment for drying is listed, several in these devices can be by same hardware branch
To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame
Claim.
Claims (25)
1. a kind of identifying system of malice address, it is characterised in that the system includes subscription client, server and businessman visitor
Family end;Wherein,
The subscription client is used for the address to be identified for receiving input, and the address to be identified is sent into the service
Device;
The server is used to receive the address to be identified that the subscription client is sent, and the address to be identified is entered
The hierarchical processing of row address, obtain each address level of the address to be identified;Analyze what is obtained using by history normal address
Address level redirects probability distribution, calculates each address level in the address to be identified and jumps to adjacent next address level
Redirect probability, the address level redirects probability distribution and jumps to another address level including any one address level
Redirect probability;Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address to be identified, and
The recognition result that malice Address Recognition is carried out based on the normal address probability is sent to the merchant client;
The merchant client is used to receiving and exporting the recognition result that the server is sent.
2. system according to claim 1, it is characterised in that the server is used for when the recognition result is described treats
When identification address is malice address, early warning information is sent to the merchant client;
The merchant client is used to receiving and exporting the early warning information that the server is sent.
3. system according to claim 2, it is characterised in that the merchant client is used to carry receiving the early warning
After showing information, the selection interface of the recognition result for selecting to be recognized the address to be identified is exported, and is received
Based on the recognition result of selection interface input, secondary identification, the recognition result of the secondary identification is returned to described
Server.
4. the system according to Claims 2 or 3, it is characterised in that the merchant client be used for do not receive it is described
In the case of early warning information, the selection of the recognition result for selecting to be recognized the address to be identified is exported
Interface, and receive based on the selection interface input, for describe the address to be identified be malice address recognition result,
And the address to be identified for carrying malice mark is returned into the server.
5. a kind of recognition methods of malice address, it is characterised in that methods described includes:
Receive the address to be identified that subscription client is sent;
The hierarchical processing in address is carried out to the address to be identified, obtains each address level of the address to be identified;
Probability distribution is redirected using the address level for analyzing to obtain by history normal address, is calculated each in the address to be identified
Address level jumps to the probability that redirects of adjacent next address level, and the address level redirects probability distribution including any one
What individual address level jumped to another address level redirects probability;
Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address to be identified.
6. according to the method for claim 5, it is characterised in that obtain the address to be identified normal address probability it
Afterwards, methods described also includes:
According to default recognition rule and the normal address probability of the address to be identified, judge the address to be identified whether be
Malice address.
7. according to the method for claim 6, it is characterised in that according to default recognition rule and the address to be identified
Normal address probability, judge whether the address to be identified is that malice address includes:
Extract and use from History Order corresponding to order to be identified corresponding to the address to be identified and/or the order to be identified
In identify the address to be identified whether be malice address default identification feature;
Obtain the default identification model trained by History Order;
According to the normal address probability of the address to be identified, the default identification feature and the default identification model, sentence
Whether the disconnected address to be identified is malice address.
8. according to the method for claim 7, it is characterised in that from order to be identified corresponding to the address to be identified and/
Or extracted in History Order corresponding to the order to be identified for identifying whether the address to be identified is the pre- of malice address
If identification feature includes:
The address text message feature corresponding to extraction from the address to be identified;
And/or history Shopping Behaviors feature is extracted from History Order corresponding to the order to be identified;
And/or the order feature corresponding to extraction from the order to be identified.
9. according to the method for claim 8, it is characterised in that from order to be identified corresponding to the address to be identified and/
Or extracted in History Order corresponding to the order to be identified for identifying whether the address to be identified is the pre- of malice address
If identification feature also includes:
According to the address text message feature, the history Shopping Behaviors feature, the order feature and described to be identified
The combination of at least two in the normal address probability of address, obtain cross feature corresponding to the address to be identified.
10. according to the method for claim 8, it is characterised in that the address text message feature includes:Whether include pre-
If the numeral of length, whether include default sensitive word and whether include advertising message;
The order feature includes:The use of whether normal, the described address to be identified of telephone number in the order to be identified
Whether number is more than default using the correlation behavior in shop and the order to be identified corresponding to threshold value, the order to be identified
To the correlation behavior of corresponding commodity.
11. according to the method for claim 7, it is characterised in that obtaining the default identification mould by History Order training
Before type, methods described also includes:
History Order is obtained, the History Order includes the normal order of history and history malice order of preset ratio;
Probability distribution is redirected according to the address level, the normal address for obtaining the historical address carried in the History Order is general
Rate;
Default identification feature is extracted from the History Order;
The default identification model is trained by the normal address probability and corresponding default identification feature of each historical address.
12. according to the method for claim 6, it is characterised in that according to default recognition rule and the address to be identified
Normal address probability, judge whether the address to be identified is that malice address includes:
Judge whether the normal address probability of the address to be identified is more than predetermined probabilities threshold value;
If the normal address probability of the address to be identified is more than the predetermined probabilities threshold value, it is determined that the address to be identified is
Normal address;
If the normal address probability of the address to be identified is less than or equal to the predetermined probabilities threshold value, it is determined that described to be identified
Address is malice address.
13. according to the method for claim 6, it is characterised in that methods described also includes:
It will determine that whether the address to be identified is that the recognition result of malice address is sent to merchant client, so as to the businessman
Client receives and exports the recognition result.
14. according to the method for claim 6, it is characterised in that methods described also includes:
If judging the address to be identified for malice address, early warning information is sent to merchant client, so as to the business
Family's client receives and exports the early warning information;
Receive it is that the merchant client is sent, the address to be identified is recognized based on the early warning information
Recognition result;
If the recognition result is the address to be identified be normal address, more new historical normal address storehouse, history maliciously
Location storehouse and the default identification model.
15. according to the method for claim 6, it is characterised in that methods described also includes:
Receive the address to be identified for the carrying malice mark that the merchant client is sent;
More new historical normal address storehouse, history malice address base and the default identification model.
16. according to the method for claim 5, it is characterised in that the address to be identified is to be carried out to order to be identified
The address obtained after redundancy processing and formatting processing.
17. according to the method for claim 16, it is characterised in that redundancy processing and formatting are carried out to order to be identified
Processing includes:
To meeting that the word of default filter condition filters in the address to be identified of the order to be identified;
Dirty data in the order to be identified is filtered;
According to predetermined formatting processing rule, processing is formatted to the order to be identified after filtering.
18. the method according to any one of claim 5 to 17, it is characterised in that carry out ground to the address to be identified
The hierarchical processing in location includes:
Based on conditional random field models, the hierarchical processing in address is carried out to the address to be identified.
19. a kind of identification device of malice address, it is characterised in that described device includes:
Receiving unit, for receiving the address to be identified of subscription client transmission;
First processing units, for carrying out the hierarchical processing in address to the address to be identified, obtain the address to be identified
Each address level;
Computing unit, for redirecting probability distribution using the address level for analyzing to obtain by history normal address, treated described in calculating
Each address level jumps to the probability that redirects of adjacent next address level in identification address, and the address level redirects probability
What distribution included that any one address level jumps to another address level redirects probability;
Second processing unit, each probability that redirects for being obtained to the computing unit carry out multiplication processing, treated described in acquisition
Identify the normal address probability of address.
20. a kind of identifying system of malice order, it is characterised in that the system includes subscription client, server and businessman
Client;Wherein,
The subscription client is used for the order to be identified for receiving input, and the order to be identified is sent into the service
Device;
The server is used to receive the order to be identified that the subscription client is sent, and based on by history normal address
Analyze obtained address level and redirect probability distribution, calculate each address level in the address of the order to be identified and jump to phase
Adjacent next address level redirects probability, and the address level redirects probability distribution and jumped to including any one address level
Another address level redirects probability;Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal of the address
Address probability;Whether it is malice order according to order to be identified described in the normal address probabilistic determination, and will determine that result is sent out
Give the merchant client;
The merchant client is used for the judged result for receiving and showing that the server is sent.
21. a kind of recognition methods of malice order, it is characterised in that methods described includes:
Receive the order to be identified that subscription client is sent;
Probability distribution is redirected based on the address level for analyzing to obtain by history normal address, calculates the address of the order to be identified
In each address level jump to the probability that redirects of adjacent next address level, the address level, which redirects probability distribution, to be included
What any one address level jumped to another address level redirects probability;
Multiplication processing is carried out to each probability that redirects of acquisition, obtains the normal address probability of the address;
Whether it is malice order according to order to be identified described in the normal address probabilistic determination.
22. according to the method for claim 21, it is characterised in that to be identified according to the normal address probabilistic determination
Whether order is that malice order includes:
Whether it is malice address according to the address of order to be identified described in the normal address probabilistic determination;
If the address of the order to be identified is malice address, it is determined that the order to be identified is malice order.
23. according to the method for claim 22, it is characterised in that if the address of the order to be identified is normal address,
Then methods described also includes:
Judge whether the telephone number in the order to be identified is normal;
If the telephone number is abnormal, it is determined that the order to be identified is malice order.
24. the method according to any one of claim 21 to 23, it is characterised in that based on being analyzed by history normal address
Obtained address level redirects probability distribution, calculate each address level in the address of the order to be identified jump to it is adjacent
The probability that redirects of next address level includes:
The hierarchical processing in address is carried out to the address of the order to be identified, obtains each address level of the address;
Probability distribution is redirected based on the address level, calculates the jump that each address level jumps to adjacent next address level
Turn probability.
25. a kind of identification device of malice order, it is characterised in that described device includes:
Receiving unit, for receiving the order to be identified of subscription client transmission;
Computing unit, for redirecting probability distribution based on the address level for analyzing to obtain by history normal address, treated described in calculating
Identify that each address level jumps to the probability that redirects of adjacent next address level in the address of order, the address level is jumped
Turn that probability distribution includes that any one address level jumps to another address level redirects probability;
Processing unit, for carrying out multiplication processing to each probability that redirects of acquisition, obtain the normal address probability of the address;
Judging unit, whether it is malice order for the order to be identified according to the normal address probabilistic determination.
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610797563.7A CN107798571B (en) | 2016-08-31 | 2016-08-31 | Malice address/malice order identifying system, method and device |
TW106119860A TW201812689A (en) | 2016-08-31 | 2017-06-14 | System, method, and device for identifying malicious address/malicious purchase order |
PCT/CN2017/097953 WO2018040944A1 (en) | 2016-08-31 | 2017-08-18 | System, method, and device for identifying malicious address/malicious purchase order |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610797563.7A CN107798571B (en) | 2016-08-31 | 2016-08-31 | Malice address/malice order identifying system, method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107798571A true CN107798571A (en) | 2018-03-13 |
CN107798571B CN107798571B (en) | 2019-08-30 |
Family
ID=61301279
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610797563.7A Active CN107798571B (en) | 2016-08-31 | 2016-08-31 | Malice address/malice order identifying system, method and device |
Country Status (3)
Country | Link |
---|---|
CN (1) | CN107798571B (en) |
TW (1) | TW201812689A (en) |
WO (1) | WO2018040944A1 (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108683749A (en) * | 2018-05-18 | 2018-10-19 | 携程旅游信息技术(上海)有限公司 | A kind of judgment method, equipment and the medium of random email address |
CN108876545A (en) * | 2018-06-22 | 2018-11-23 | 北京小米移动软件有限公司 | Order recognition methods, device and readable storage medium storing program for executing |
CN109345332A (en) * | 2018-08-27 | 2019-02-15 | 中国民航信息网络股份有限公司 | A kind of intelligent detecting method of Airline reservation malicious act |
CN109407504A (en) * | 2018-11-30 | 2019-03-01 | 华南理工大学 | A kind of personal safety detection system and method based on smartwatch |
CN109947564A (en) * | 2019-03-07 | 2019-06-28 | 阿里巴巴集团控股有限公司 | Method for processing business, device, equipment and storage medium |
CN110335115A (en) * | 2019-07-01 | 2019-10-15 | 阿里巴巴集团控股有限公司 | A kind of service order processing method and processing device |
CN110503517A (en) * | 2019-08-13 | 2019-11-26 | 蚌埠聚本电子商务产业园有限公司 | A kind of fallacious message detection and method of disposal for e-commerce |
CN110807685A (en) * | 2019-10-22 | 2020-02-18 | 上海钧正网络科技有限公司 | Information processing method, device, terminal and readable storage medium |
CN110874778A (en) * | 2018-08-31 | 2020-03-10 | 阿里巴巴集团控股有限公司 | Abnormal order detection method and device |
CN111461815A (en) * | 2020-03-17 | 2020-07-28 | 上海携程国际旅行社有限公司 | Order recognition model generation method, recognition method, system, device and medium |
CN111859956A (en) * | 2020-07-09 | 2020-10-30 | 睿智合创(北京)科技有限公司 | Address word segmentation method for financial industry |
CN112686732A (en) * | 2021-01-06 | 2021-04-20 | 中国联合网络通信集团有限公司 | Abnormal address data identification method, device, equipment, medium and product |
CN112950298A (en) * | 2019-11-26 | 2021-06-11 | 北京沃东天骏信息技术有限公司 | Malicious order identification method and device and storage medium |
CN113076752A (en) * | 2021-03-26 | 2021-07-06 | 中国联合网络通信集团有限公司 | Method and device for identifying address |
CN113240480A (en) * | 2021-01-25 | 2021-08-10 | 天津五八到家货运服务有限公司 | Order processing method and device, electronic terminal and storage medium |
CN116934418A (en) * | 2023-06-15 | 2023-10-24 | 广州淘通科技股份有限公司 | Abnormal order detection and early warning method, system, equipment and storage medium |
Families Citing this family (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110852080A (en) * | 2018-08-01 | 2020-02-28 | 北京京东尚科信息技术有限公司 | Order address identification method, system, equipment and storage medium |
CN109587248B (en) * | 2018-12-06 | 2023-08-29 | 腾讯科技(深圳)有限公司 | User identification method, device, server and storage medium |
CN111132144B (en) * | 2019-12-25 | 2022-09-13 | 中国联合网络通信集团有限公司 | Abnormal number identification method and equipment |
CN111935646B (en) * | 2020-07-22 | 2022-09-20 | 北京明略昭辉科技有限公司 | Method and system for estimating common address of mobile equipment user |
CN111915256B (en) * | 2020-07-31 | 2023-09-26 | 上海寻梦信息技术有限公司 | Method for constructing dispatch fence, off-site signing and identifying method and related equipment |
CN112101993B (en) * | 2020-09-11 | 2022-12-23 | 厦门美图之家科技有限公司 | Offline anti-cheating method and device, electronic equipment and readable storage medium |
CN112446425A (en) * | 2020-11-20 | 2021-03-05 | 北京思特奇信息技术股份有限公司 | Method and device for automatically acquiring suspected card-maintaining channel |
CN112491863B (en) * | 2020-11-23 | 2022-07-29 | 中国联合网络通信集团有限公司 | IP address black and gray list analysis method, server, terminal and storage medium |
CN113449523B (en) * | 2021-06-29 | 2024-05-24 | 京东科技控股股份有限公司 | Method and device for determining abnormal address text, electronic equipment and storage medium |
CN117371893A (en) * | 2023-10-09 | 2024-01-09 | 杭州正马软件科技有限公司 | System and method for automatically changing e-commerce order address |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008038017A1 (en) * | 2006-09-29 | 2008-04-03 | British Telecommunications Public Limited Company | Information processing system and related method |
CN103095711A (en) * | 2013-01-18 | 2013-05-08 | 重庆邮电大学 | Application layer distributed denial of service (DDoS) attack detection method and defensive system aimed at website |
CN104462059A (en) * | 2014-12-01 | 2015-03-25 | 银联智惠信息服务(上海)有限公司 | Commercial tenant address information recognition method and device |
CN105389722A (en) * | 2015-11-20 | 2016-03-09 | 小米科技有限责任公司 | Malicious order identification method and device |
CN105468742A (en) * | 2015-11-25 | 2016-04-06 | 小米科技有限责任公司 | Malicious order recognition method and device |
-
2016
- 2016-08-31 CN CN201610797563.7A patent/CN107798571B/en active Active
-
2017
- 2017-06-14 TW TW106119860A patent/TW201812689A/en unknown
- 2017-08-18 WO PCT/CN2017/097953 patent/WO2018040944A1/en active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008038017A1 (en) * | 2006-09-29 | 2008-04-03 | British Telecommunications Public Limited Company | Information processing system and related method |
CN103095711A (en) * | 2013-01-18 | 2013-05-08 | 重庆邮电大学 | Application layer distributed denial of service (DDoS) attack detection method and defensive system aimed at website |
CN104462059A (en) * | 2014-12-01 | 2015-03-25 | 银联智惠信息服务(上海)有限公司 | Commercial tenant address information recognition method and device |
CN105389722A (en) * | 2015-11-20 | 2016-03-09 | 小米科技有限责任公司 | Malicious order identification method and device |
CN105468742A (en) * | 2015-11-25 | 2016-04-06 | 小米科技有限责任公司 | Malicious order recognition method and device |
CN105468742B (en) * | 2015-11-25 | 2018-11-20 | 小米科技有限责任公司 | The recognition methods of malice order and device |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108683749A (en) * | 2018-05-18 | 2018-10-19 | 携程旅游信息技术(上海)有限公司 | A kind of judgment method, equipment and the medium of random email address |
CN108876545A (en) * | 2018-06-22 | 2018-11-23 | 北京小米移动软件有限公司 | Order recognition methods, device and readable storage medium storing program for executing |
CN109345332A (en) * | 2018-08-27 | 2019-02-15 | 中国民航信息网络股份有限公司 | A kind of intelligent detecting method of Airline reservation malicious act |
CN110874778A (en) * | 2018-08-31 | 2020-03-10 | 阿里巴巴集团控股有限公司 | Abnormal order detection method and device |
CN110874778B (en) * | 2018-08-31 | 2023-04-25 | 阿里巴巴集团控股有限公司 | Abnormal order detection method and device |
CN109407504B (en) * | 2018-11-30 | 2021-05-14 | 华南理工大学 | Personal safety detection system and method based on smart watch |
CN109407504A (en) * | 2018-11-30 | 2019-03-01 | 华南理工大学 | A kind of personal safety detection system and method based on smartwatch |
CN109947564A (en) * | 2019-03-07 | 2019-06-28 | 阿里巴巴集团控股有限公司 | Method for processing business, device, equipment and storage medium |
CN109947564B (en) * | 2019-03-07 | 2023-04-11 | 蚂蚁金服(杭州)网络技术有限公司 | Service processing method, device, equipment and storage medium |
CN110335115A (en) * | 2019-07-01 | 2019-10-15 | 阿里巴巴集团控股有限公司 | A kind of service order processing method and processing device |
CN110503517A (en) * | 2019-08-13 | 2019-11-26 | 蚌埠聚本电子商务产业园有限公司 | A kind of fallacious message detection and method of disposal for e-commerce |
CN110807685A (en) * | 2019-10-22 | 2020-02-18 | 上海钧正网络科技有限公司 | Information processing method, device, terminal and readable storage medium |
CN110807685B (en) * | 2019-10-22 | 2021-09-07 | 上海钧正网络科技有限公司 | Information processing method, device, terminal and readable storage medium |
CN112950298A (en) * | 2019-11-26 | 2021-06-11 | 北京沃东天骏信息技术有限公司 | Malicious order identification method and device and storage medium |
CN111461815A (en) * | 2020-03-17 | 2020-07-28 | 上海携程国际旅行社有限公司 | Order recognition model generation method, recognition method, system, device and medium |
CN111461815B (en) * | 2020-03-17 | 2023-04-28 | 上海携程国际旅行社有限公司 | Order recognition model generation method, recognition method, system, equipment and medium |
CN111859956A (en) * | 2020-07-09 | 2020-10-30 | 睿智合创(北京)科技有限公司 | Address word segmentation method for financial industry |
CN112686732A (en) * | 2021-01-06 | 2021-04-20 | 中国联合网络通信集团有限公司 | Abnormal address data identification method, device, equipment, medium and product |
CN112686732B (en) * | 2021-01-06 | 2023-07-11 | 中国联合网络通信集团有限公司 | Abnormal address data identification method, device, equipment and medium |
CN113240480A (en) * | 2021-01-25 | 2021-08-10 | 天津五八到家货运服务有限公司 | Order processing method and device, electronic terminal and storage medium |
CN113076752A (en) * | 2021-03-26 | 2021-07-06 | 中国联合网络通信集团有限公司 | Method and device for identifying address |
CN116934418A (en) * | 2023-06-15 | 2023-10-24 | 广州淘通科技股份有限公司 | Abnormal order detection and early warning method, system, equipment and storage medium |
CN116934418B (en) * | 2023-06-15 | 2024-03-19 | 广州淘通科技股份有限公司 | Abnormal order detection and early warning method, system, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
WO2018040944A1 (en) | 2018-03-08 |
TW201812689A (en) | 2018-04-01 |
CN107798571B (en) | 2019-08-30 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107798571B (en) | Malice address/malice order identifying system, method and device | |
US11403532B2 (en) | Method and system for finding a solution to a provided problem by selecting a winner in evolutionary optimization of a genetic algorithm | |
US9607273B2 (en) | Optimal time to post for maximum social engagement | |
CN107105031A (en) | Information-pushing method and device | |
CN111444334A (en) | Data processing method, text recognition device and computer equipment | |
CN103620588A (en) | Identifying matching applications based on browsing activity | |
CN111107048A (en) | Phishing website detection method and device and storage medium | |
CN108090568A (en) | Recognize robot credit parser | |
CN103177129A (en) | Internet real-time information recommendation and prediction system | |
CN112215658A (en) | Big data-based addressing method and device, computer equipment and storage medium | |
CN107077455A (en) | Flow mass is determined using the score traffic based on event | |
US20140082550A1 (en) | Systems and methods for integrated query and navigation of an information resource | |
CN106383857A (en) | Information processing method and electronic equipment | |
CN111523315A (en) | Data processing method, text recognition device and computer equipment | |
CN105528436A (en) | Data display method and device | |
CN113360788A (en) | Address recommendation method, device, equipment and storage medium | |
Xia et al. | Click-through rate prediction of online banners featuring multimodal analysis | |
CN107835305A (en) | Data inputting method and device for the terminal device with screen | |
CN116823410A (en) | Data processing method, object processing method, recommending method and computing device | |
CN111159399A (en) | Automobile vertical website water army discrimination method | |
CN112348661B (en) | Service policy distribution method and device based on user behavior track and electronic equipment | |
CN109978607A (en) | Advertisement recommended method, device and computer readable storage medium | |
CN115439156A (en) | Method and device for recommending advertisement space, computer equipment and storage medium | |
CN111026981B (en) | Visual display method, device and equipment for hot topics | |
CN114925275A (en) | Product recommendation method and device, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 1252396 Country of ref document: HK |
|
GR01 | Patent grant | ||
GR01 | Patent grant |