CN106934498A - The recommendation method and system of hotel's house type in OTA websites - Google Patents
The recommendation method and system of hotel's house type in OTA websites Download PDFInfo
- Publication number
- CN106934498A CN106934498A CN201710150879.1A CN201710150879A CN106934498A CN 106934498 A CN106934498 A CN 106934498A CN 201710150879 A CN201710150879 A CN 201710150879A CN 106934498 A CN106934498 A CN 106934498A
- Authority
- CN
- China
- Prior art keywords
- house type
- hotel
- type
- user
- house
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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]
- G06Q30/0631—Item recommendations
-
- 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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/12—Hotels or restaurants
-
- 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
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/14—Travel agencies
Abstract
The invention discloses a kind of recommendation method and system of hotel's house type in OTA websites, the recommendation method includes：Judge whether user successfully ordered hotel in the OTA websites；According to the sales volume order from high to low of composite services type in the History Order data in all hotels, the house type that there is respective combination COS is searched whether successively in target hotel, if stopping searching in the presence of if；The History Order data of History Order information and all hotels according to the user, train house type recommended models；Information and the house type recommended models, the probability that each house type is ordered by the user in the calculating target hotel are browsed according to the user, and selects the corresponding house type of maximum probability；Recommendation information of the output on the house type.Compared with prior art, the way of recommendation of the invention can quick-pick go out to meet the house type of user's request, the efficiency in user's shopping on net hotel is improve, while also improving the conversion ratio of order.
Description
Technical field
The present invention relates to internet arena, more particularly to a kind of OTA (Online Travel Agent, online travel agency)
The recommendation method and system of hotel's house type in website.
Background technology
With continuing to develop for network and development of Mobile Internet technology, increasing consumer's selection by OTA websites and
Mobile terminal APP (Application, application program) checks hotel information and orders hotel room.And many OTA websites are in order to carry
Rise user to browse and select the efficiency of Hotel Products, employ recommendation and ordering techniques show to user and recommend personalized product
Product.
At present, recommended technology widely used in electric business, is substantially, selection heat outside at shop (hotel's details page)
Door or user search for similarity product higher and are recommended recently；And when user enters shop inside, the product of displaying is then
Non-personalized static display is carried out by shop owner or page system.However, during hotel reservation, when user enters shop
It is also part that user is extremely concerned about to the selection of house type, or even have influence on selection of the user finally to hotel when paving is internal.Cause
This, the recommendation of house type is conducive to user be quickly found out it is suitable oneself hotel and room, greatly promote what is booked rooms on user network
Efficiency and experience.
Because Hotel Products possess some particularity compared with other electric business products, for the recommendation problem of hotel room
In the presence of its complexity and particularity.
For firstly, for other electric business, into after a shop, we can uniquely determine one by trade name
Product.It is identical due to many house type titles or numbering occurring in hotel's house type but the sub- product in hotel is different, but because of service
Type is different, or house type supply channel is different, the house type of different prices or identical price is resulted in, therefore, it is impossible to pass through
House type title, house type numbering make a distinction.As rich choice of products increases, the house type number of same names and identical numbering is not
Disconnected lifting.
Secondly, for other electric business, from appearance between different commodity in shop.Nominally have conspicuousness poor
Different, user can quickly, clearly be selected according to the hobby of oneself and needs.But between hotel's difference house type, no matter
It is otherness all very littles from COS or in price etc., each cannot be distinguished from the information for showing or even substantially
Difference between house type, part randomness is there may be for the selection of user, i.e., the house type do not ordered is not necessarily really
Negative sample.
Finally, for the recommendation for occurring carrying out user personalization in electric business, it is most of be all recommend it is popular salable
Product, and for newly entering product, be difficult to be recommended in the case of without any sales volume.
Therefore, in view of the particularity of hotel's house type product, widely used proposed algorithm can not be very in other electric business
Well suitable for the recommendation to hotel's house type.
The content of the invention
The technical problem to be solved in the present invention be in order to overcome select in the prior art popular or user search similarity compared with
Product high carry out recommendation method exist recommended range it is narrow, recommend product it is unreasonable and easily ignore new product etc. lack
Fall into, there is provided the recommendation method and system of hotel's house type in a kind of OTA websites recommended in different ways old and new users.
The present invention is to solve above-mentioned technical problem by following technical proposals：
On the one hand, there is provided a kind of recommendation method of hotel's house type in OTA websites, its feature is to comprise the following steps：
S1, judge whether user successfully ordered hotel in the OTA websites, if so, step S3 is then performed, if it is not, then
Perform step S2；
S2, the sales volume order from high to low according to composite services type in the History Order data in all hotels, in mesh
The house type that there is respective combination COS is searched whether in mark hotel successively, if in the presence of stopping is searched, and performs step
S5；
The History Order data of S3, the History Order information according to the user and all hotels, training house type is recommended
Model；
S4, information and the house type recommended models are browsed according to the user, calculate each in the target hotel
The probability that house type is ordered by the user, and select the corresponding house type of maximum probability；
S5, recommendation information of the output on the house type.
If it is preferred that the house type found in step S2 is multiple, being exported in step S5 on price in multiple house types
The recommendation information of minimum house type；If the cheapest house type is not the house type of lowest price in the target hotel, walk
Composite services class in rapid S5 according to corresponding to the house type of lowest price in the cheapest house type and the target hotel
Type, recommendation information of the output on the optimal house type of service.
If it is preferred that the lowest price of the house type found in step S2 is higher than default price, continuing in target hotel
Search whether there is the house type of next composite services type；
If the lowest price of the house type found in step S2 is above default price, exported in step S5 on the mesh
The recommendation information of the house type of lowest price in mark hotel；If the house type of lowest price is multiple in the target hotel, according to multiple
Composite services type corresponding to house type, recommendation information of the output on the optimal house type of service.
It is preferred that in step S4, if the corresponding house type of maximum probability is multiple, being exported in step S5 on multiple house types
Middle position rest against before house type recommendation information.
It is preferred that in step S4, if the price of the corresponding house type of maximum probability orders house type higher than the user's history
Highest price and the product of preset multiple, then select time corresponding house type of high probability, until the house type of selection is inexpensive in institute
State product；
If the price of each house type of selection is above the product, any recommendation information is not exported.
It is preferred that the composite services type includes at least one combination of following COS：On one's own account/non-self-operation, contain
Early/without early, confirmations/non-immediate acknowledgment, freely cancellation/non-free cancellation, spot payment/prepayment, lowest price immediately.
On the other hand, there is provided the commending system of hotel's house type in a kind of OTA websites, its feature is, including judge module,
Searching modul, training module, selecting module and recommending module：
The judge module is used to judge whether user successfully ordered hotel in the OTA websites, and in the situation for being
Under call the training module, the searching modul is called in a case of no；
The searching modul be used for according to composite services type in the History Order data in all hotels sales volume from height to
Low order, searches whether the house type that there is respective combination COS successively in target hotel, if in the presence of stopping is looked into
Look for, and call the recommending module；
The training module is used for History Order information and the History Order data in all hotels according to the user,
Training house type recommended models；
The selecting module is used to browse information and the house type recommended models according to the user, calculates the mesh
The probability that each house type is ordered by the user in mark hotel, and the corresponding house type of maximum probability is selected, call the recommendation mould
Block；
The recommending module is used to export the recommendation information on the house type.
If it is preferred that the house type that the searching modul finds is multiple, the recommending module output is on multiple rooms
The recommendation information of cheapest house type in type；If the cheapest house type is not the room of lowest price in the target hotel
Type, then the recommending module is according to corresponding to the house type of lowest price in the cheapest house type and the target hotel
Composite services type, recommendation information of the output on the optimal house type of service.
It is preferred that the searching modul is additionally operable to, when the lowest price of the house type for finding is higher than default price, continue
Search whether there is the house type of next composite services type in target hotel；
If the lowest price of the house type that the searching modul finds is above default price, the recommending module output is closed
The recommendation information of the house type of lowest price in the target hotel；If the house type of lowest price is multiple in the target hotel,
Composite services type according to corresponding to multiple house types, recommendation information of the output on the optimal house type of service.
It is preferred that the selecting module is used to be gone through higher than the user in the price of the corresponding house type of maximum probability of selection
When history orders the product of highest price and the preset multiple of house type, the selection time corresponding house type of high probability, until the house type of selection
It is inexpensive in the product；
If the price of each house type of the selecting module selection is above the product, the recommending module is not exported
Any recommendation information.
On the basis of common sense in the field is met, above-mentioned each optimum condition can be combined, and obtain final product each preferable reality of the present invention
Example.
Positive effect of the invention is：Compared with prior art, the present invention is by the user to logging in OTA websites
The classification of new/old user is carried out, the information on hotel's house type is recommended to user using the different ways of recommendation, can quickly chosen
The house type for meeting user's request is selected, the function of reasonable recommendation is realized, the efficiency in user's shopping on net hotel is improve, while
Also the conversion ratio of order is improved.
Brief description of the drawings
Fig. 1 is the recommendation method flow diagram of hotel's house type in the OTA websites of the embodiment of the present invention.
Fig. 2 is the structured flowchart of the commending system of hotel's house type in the OTA websites of the embodiment of the present invention.
Specific embodiment
The present invention is further illustrated below by the mode of embodiment, but does not therefore limit the present invention to described reality
Apply among a scope.
The application scenarios of the embodiment of the present invention are：When the user for logging in OTA websites enters the internal details page in target hotel
When, output meets the recommendation information on hotel's house type of user's request as far as possible, to improve the effect in user's shopping on net hotel
Rate.Wherein, there is the information of very multiple hotels in the OTA websites, when the information in wherein one hotel of user's click, can enter
The internal details page in the hotel, the way of recommendation of the prior art is that the house type for recommending fast sale per family is used to each, or according to
The condition of user's search recommends similarity house type higher, and this way of recommendation has that recommended range is narrow, product that is recommending does not conform to
The shortcomings of managing and easily ignore new product.
The present embodiment provides a kind of recommendation method of hotel's house type in OTA websites, as shown in figure 1, comprising the following steps：
Step 101, judge whether user successfully ordered hotel in the OTA websites, if so, step 103 is then performed, if
It is no, then perform step 102.
Wherein, the user that hotel was successfully ordered in the OTA websites is old user, is not ordered successfully in the OTA websites
The user in Gou Guo hotels is new user, takes different modes to recommend new/old user.
Step 102, the sales volume order from high to low according to composite services type in the History Order data in all hotels,
The house type that there is respective combination COS is searched whether successively in target hotel, if in the presence of stopping is searched, and performs step
Rapid 105.
In a kind of optional implementation method, COS can include：On one's own account/non-self-operation, containing it is early/without early, true immediately
Recognize/non-immediate acknowledgment, free cancellation/non-free cancellation, spot payment/prepayment etc..When the species of COS is k, 2 are hadk
Plant composite services type.By ordering conventional user the analysis of house type price, it is lower that most of order is partial to ordering price
House type, when house type price is close, can just consider to select COS, therefore, for certain customers, house type lowest price
It is one of factor that user values the most.Therefore, by lowest price separately as a new COS.Therefore, for each
For hotel, 2 are hadk+ a kind of composite services type.For example, sales volume highest composite services type is：On one's own account+containing morning+
Non-immediate acknowledgment+free cancellation；Sales volume time composite services type high is：On one's own account+contain morning+lowest price.
The History Order data of step 103, the History Order information according to the user and all hotels, train house type
Recommended models.
Step 104, information and the house type recommended models are browsed according to the user, in calculating the target hotel
The probability that each house type is ordered by the user, and select the corresponding house type of maximum probability.
In a kind of optional implementation method, the History Order data in all hotels, user are obtained by burying a mechanism
History Order information and user's browses information.For example, for user browsing each time, clicking on OTA websites,
The behaviors such as screening, purchase, in being stored in database in the form of " snapshot ", the scene such as order, browse in order to user, and then
For the information such as subsequent analysis user behavior, preference provide data basis.
Step 105, recommendation information of the output on the house type.
In for a kind of optional implementation method of new user, if the house type found in step 102 is multiple, step
The recommendation information on cheapest house type in multiple house types is exported in 105.
In for a kind of optional implementation method of new user, if above-mentioned cheapest house type is not the target hotel
The house type of middle lowest price, then according to the room of lowest price in the cheapest house type and the target hotel in step 105
Composite services type corresponding to type, recommendation information of the output on the optimal house type of service.Specifically, when being looked into step 102
During the house type Bu Shi hotels lowest price house type for finding, then need whether checking has advantage compared with hotel's lowest price house type premium, that is, use
Whether family is ready is spent to order non-hotel's lowest price house type more；When the house type found in step 102 is hotel's lowest price house type
When, then it is the optimal house type of integrated service in multiple hotel's lowest price house types to need to verify whether.
Wherein, the Superiority Value of each house type is calculated by following house type value value formula, and value values are smaller to be shown to get over
It is advantageous：
Valuei=Pi–Pmin* [whether whether the accounting * I of+COS of COS 12 { take the accounting * I of COS 1
Service type 2 }+...+COS N accountings * whether COS N] * 1/ ([whether whether+the I of COS 1 { services I
Type 2 }+...+I whether COS N+0.000001]) ^ ∝
Wherein, PiIt is each house type price, PminIt is hotel's lowest price house type price；N represents service type category sum,
∝ is represented with COS growth, the speed that Superiority Value increases.The respective services value of each house type is substituted into formula, is passed through
Compare the value value sizes between house type, and then compare whether house type has advantage.If the house type value Zhi Bi hotels recommended are most
Any one house type value values are big at a low price, illustrate that the recommendation house type is not optimal house type, then find next fast-selling combination clothes
The house type of service type, is recommended until searching out and servicing the house type of optimal value values.
In for a kind of optional implementation method of new user, if the lowest price of the house type found in step 102 is higher than
Default price, then continue the house type for searching whether to have next composite services type in target hotel；If being looked into step 102
The lowest price of the house type for finding is above default price, then the room on lowest price in the target hotel is exported in step 105
The recommendation information of type；If the house type of lowest price is multiple in the target hotel, the combination clothes according to corresponding to multiple house types
Service type, recommendation information of the output on the optimal house type of service.
For example, default price can be (meet can order, the condition such as non-hourly paid hotel room) in recommended range in target hotel
The minimum of house type return rear valency * S times, default price can also be that the minimum of house type that presently, there are in target hotel return rear valency * T
Times.Wherein, the value of S and T can be analyzed by simple data, so that (user orders house type price/hotel and can order non-clock in history
Point room lowest price house type) p% quantiles (such as p=97.5) be dividing value.
In for a kind of optional implementation method of old user, History Order information, user with reference to user browse letter
Breath and the History Order data in all hotels, can construct and order three related class dimensions of house type to user：The related dimension of user
Degree, hotel's relevant dimension and current click Hotel Products information.
Wherein, user's relevant dimension can be largely classified into user's history ordering information descriptive statistic index and user it is clear
Look at, click on, the desired value of filter information.First, mainly being included for the descriptive statistic index of user's history ordering information：With
Family history is ordered order price, price fluctuation coefficient, the price coefficient of variation, point city order order price, point star and is ordered
Univalent lattice, averagely order star, in advance scheduled duration, the various service preferences values of user, user's portrait respective labels etc..Second,
Browsed for user, clicked on, the desired value of filter information is mainly included：User T+1 weeks, T+0, with the browsing of session, click on, sieve
Select Hotel Star, hotel's price, house type ID, bed-type, service；User T+1 weeks, T+0, with session with currently browse hotel Zhong Ge rooms
Similarity of type etc..
Hotel's relevant dimension can be largely classified into hotel's yield, the statistical indicator of profit and scoring class Static State Index.First,
Mainly included for hotel's yield, the statistical indicator of profit：Nearly one week of main house type, nearly three days order volumes, a night amount, gross profit are accounted for
According to the ratio in whole hotel；The various trip classification order accountings in hotel；The main various trip classification order accountings of house type.Second, right
Mainly included in scoring class Static State Index：Hotel's overall score, hotel position, facility, health, service scoring；Each main house type is commented
Whole hotel's comment number accounting etc. is accounted for by number.
Current Hotel Products information of clicking on can be largely classified into hotel information, institute owner house type information, target house type information.
First, hotel information mainly includes Hotel Star, hotel ID, lowest price, average price, highest price.Second, institute owner house type information master
To include comment number, area, ovary type number, lowest price, average price, highest price can be ordered.Third, target house type information is mainly included：
Breakfast, self-operation, confirm immediately, cancel type, type of payment, guarantee type, whether completely ten send one, whether super return, price, return
Present price, bed-type, position, satisfaction is subscribed, confirmed in a few houres, constructed various COS lowest prices (for example, if be similar
Whether house type contains early lowest price, is that similar house type confirms lowest price etc. immediately), construct each ovary type to basic house type (hotel)
Whether lowest price premium, premium are beyond user's history premium value etc..
In order at utmost obtain the product structure that user ordered hotel at that time, although setting, there is " snapshot " characteristic to bury a little
Data go back original subscriber's scene at that time, but are often not fee from the situation that related data missing occurs.Equally, for instruction
The selected dimension of house type recommended models practised also occurs different types of missing.Accordingly, it would be desirable to the data for getting
Pre-processed (filling, screening etc.).Mainly data processing is carried out by following several aspects：
Firstth, occur lacking for static statistics information such as user's dimension, hotel's dimensions, will be clustered by k-means etc.
Method, is filled according to missing user, the user of hotel's same type, the average of hotel information, the index such as median；For
Current hotel's house type product information is (such as：Subscribe satisfaction, ovary type does not show on line such as confirmation itself in a few houres) according to exhibition
Show that the indexs such as quantile, the average of ovary type value are filled.
Secondth, there is missing for the product information (loss is buried in displaying on line itself) that active user enters hotel,
Place order is deleted in selection；For the extreme user data for occurring the types such as brush list, reptile, ox in training data, carry out
Reject.
3rd, it is unbalance for positive and negative sample, down-sampled mode can be taken so that positive and negative sample ratio is down to 1：m.
4th, it transform new dimension as each house type price：To each house type real price and hotel's lowest price
Ratio is according to jumping degree for gap carries out a point bucket.
Because training data order lowest price order accounting is more, causes model learning premium very little but contain more advantage
This kind of advantage house type of service ability it is weaker.In consideration of it, can be modified according to below equation to house type real price：
Wherein, Price_last represents the price of each house type；Hotel_bookable_minprice represented and can currently order
The lowest price of house type；Δ=I { whether meeting COS 1 }+I { whether meeting COS 2 }+... whether+I { meets service
Type N }+0.000001；ServiceiRepresent preference of the user for COS i；--- represent user couple
When the preference value of COS i is too small, the value being filled；Represent user for
The preference value of COS iWhen, useIt is filled；Whether I { meets service
Type i } when representing that the house type is belonging to COS i, value is 1, is otherwise 0.N represents COS sum；β represent with
COS increases, the speed of price indentation.
Realize that house type is recommended in order to efficient, quick, various disaggregated models can be applied.By observing dependent variable and independent variable number
According to the feature for presenting, in combination with model prediction accuracy, selection is a kind of efficiently, the degree of accuracy is high, renewal iteration speed is fast divides
Class model, specifically can be fast using renewal speed, the XGBoost models of high precision.Personalization is carried out with the model for training
Recommend, before old user clicks to enter hotel's details page, the static dimension needed for obtaining model from production environment storehouse, and adjust
With development interface on line, real-time online treatment is carried out for the hotel's house type information on line.Most all dimensions substitute into training at last
Good model, show that each can order the order probability of non-hourly paid hotel room.Selection is calculated orders probability highest house type for we are final
The ovary type to be recommended.
In for a kind of optional implementation method of old user, in step 104, if the corresponding house type of maximum probability is many
It is individual, then the recommendation information of the house type before being rested against on multiple house type middle positions is exported in step 105.Probability is calculated for each house type
According to equation below amendment：
changeprobi=probi-γ1*rank-γ2*master_seq。
Wherein, changeprobiRepresent the revised probability of each house type；probiRepresent that each house type is recommended by house type
The original probability value that model is calculated；Rank represents that each house type is located at the position of institute's owner's house type；γ1Represent rank positions
Influence power；Master_seq represents the position that main house type is located in hotel；γ2Represent the influence power of master_seq positions.
In for a kind of optional implementation method of old user, the order probable value for being recommended house type can not be too small, works as quilt
Recommend the order probable value of house type too small (such as：Less than the house type 0.1), is then illustrated with the matching degree needed for user not enough, to this
User will not recommend in this hotel.
In for a kind of optional implementation method of old user, in step 104, if the price of the corresponding house type of maximum probability
Higher than the highest price that the user's history orders house type and the product of preset multiple, then time corresponding house type of high probability is selected, directly
Extremely the house type of selection is inexpensive in the product；If the price of each house type of selection is above the product, not defeated
Go out any recommendation information.
Wherein, P points that ovary type price/user orders the maximum of order price in the past is re-ordered with old user's order
Digit is (such as：P=98%) as the value of preset multiple.
The present embodiment also provides a kind of commending system 20 of hotel's house type in OTA websites, as shown in Fig. 2 including judging mould
Block 21, searching modul 22, training module 23, selecting module 24 and recommending module 25.Separately below to the function of modules
Describe in detail.
The judge module is used to judge whether user successfully ordered hotel in the OTA websites, and in the situation for being
Under call the training module, the searching modul is called in a case of no.
The searching modul be used for according to composite services type in the History Order data in all hotels sales volume from height to
Low order, searches whether the house type that there is respective combination COS successively in target hotel, if in the presence of stopping is looked into
Look for, and call the recommending module.
The training module is used for History Order information and the History Order data in all hotels according to the user,
Training house type recommended models.
The selecting module is used to browse information and the house type recommended models according to the user, calculates the mesh
The probability that each house type is ordered by the user in mark hotel, and the corresponding house type of maximum probability is selected, call the recommendation mould
Block.
The recommending module is used to export the recommendation information on the house type.
In for a kind of optional implementation method of new user, if the house type that the searching modul finds is multiple,
The recommending module exports the recommendation information on cheapest house type in multiple house types；If the cheapest house type is not
It is the house type of lowest price in the target hotel, then the recommending module is according to the cheapest house type and the target
Composite services type in hotel corresponding to the house type of lowest price, recommendation information of the output on the optimal house type of service.
In for a kind of optional implementation method of new user, the searching modul is additionally operable in the house type for finding most
When at a low price higher than default price, continuation searches whether the house type that there is next composite services type in target hotel；
If the lowest price of the house type that the searching modul finds is above default price, the recommending module output is closed
The recommendation information of the house type of lowest price in the target hotel；If the house type of lowest price is multiple in the target hotel,
Composite services type of the recommending module according to corresponding to multiple house types, recommendation of the output on the optimal house type of service
Breath.
In for a kind of optional implementation method of old user, the selecting module is used for the maximum probability correspondence in selection
House type price highest price and preset multiple that house type is ordered higher than the user's history product when, selection time high probability pair
The house type answered, until the house type of selection is inexpensive in the product；
If the price of each house type of the selecting module selection is above the product, the recommending module is not exported
Any recommendation information.
In the present embodiment, by carrying out the classification of newly/old user to the user for logging in OTA websites, using different recommendations
Mode to user recommend the information on hotel's house type, can quick-pick go out to meet the house type of user's request, realize rationally
The function of recommendation, improves the efficiency in user's shopping on net hotel, while also improving the conversion ratio of order.It is specific at one
In application example, the order conversion ratio of whole OTA websites can be lifted 1%-2% or so.
Although the foregoing describing specific embodiment of the invention, it will be appreciated by those of skill in the art that these
It is merely illustrative of, protection scope of the present invention is defined by the appended claims.Those skilled in the art is not carrying on the back
On the premise of principle of the invention and essence, various changes or modifications can be made to these implementation methods, but these are changed
Protection scope of the present invention is each fallen within modification.
Claims (10)
1. in a kind of OTA websites hotel's house type recommendation method, it is characterised in that comprise the following steps：
S1, judge whether user successfully ordered hotel in the OTA websites, if so, step S3 is then performed, if it is not, then performing
Step S2；
S2, the sales volume order from high to low according to composite services type in the History Order data in all hotels, in target wine
The house type that there is respective combination COS is searched whether in shop successively, if in the presence of stopping is searched, and performs step S5；
The History Order data of S3, the History Order information according to the user and all hotels, train house type recommended models；
S4, information and the house type recommended models are browsed according to the user, calculate each house type in the target hotel
The probability ordered by the user, and select the corresponding house type of maximum probability；
S5, recommendation information of the output on the house type.
2. it is as claimed in claim 1 to recommend method, it is characterised in that if the house type found in step S2 is multiple, to walk
The recommendation information on cheapest house type in multiple house types is exported in rapid S5；If the cheapest house type is not described
The house type of lowest price in target hotel, then according to minimum in the cheapest house type and the target hotel in step S5
Composite services type corresponding to the house type of valency, recommendation information of the output on the optimal house type of service.
3. it is as claimed in claim 1 to recommend method, it is characterised in that if the lowest price of the house type found in step S2 is higher than
Default price, then continue the house type for searching whether to have next composite services type in target hotel；
If the lowest price of the house type found in step S2 is above default price, exported in step S5 on the target wine
The recommendation information of the house type of lowest price in shop；If the house type of lowest price is multiple in the target hotel, according to multiple house types
Corresponding composite services type, recommendation information of the output on the optimal house type of service.
4. it is as claimed in claim 1 to recommend method, it is characterised in that in step S4, if the corresponding house type of maximum probability is many
It is individual, then the recommendation information of the house type before being rested against on multiple house type middle positions is exported in step S5.
5. it is as claimed in claim 1 to recommend method, it is characterised in that in step S4, if the valency of the corresponding house type of maximum probability
Lattice order the highest price of house type and the product of preset multiple higher than the user's history, then select time corresponding house type of high probability,
Until the house type of selection is inexpensive in the product；
If the price of each house type of selection is above the product, any recommendation information is not exported.
6. it is as claimed in claim 1 to recommend method, it is characterised in that the composite services type includes following COS
At least one combination：On one's own account/non-self-operation, containing it is early/without early, confirmations/non-immediate acknowledgment, free cancellation/non-free cancellation, existing immediately
Pay/prepayment, lowest price.
7. in a kind of OTA websites hotel's house type commending system, it is characterised in that including judge module, searching modul, training mould
Block, selecting module and recommending module：
The judge module is used to judge that whether user successfully ordered hotel, and adjust in a case of yes in the OTA websites
The training module is used, the searching modul is called in a case of no；
The searching modul be used for according to composite services type in the History Order data in all hotels sales volume from high to low
Sequentially, the house type that there is respective combination COS is searched whether successively in target hotel, if in the presence of, stop searching, and
Call the recommending module；
The training module is used for History Order information and the History Order data in all hotels according to the user, training
House type recommended models；
The selecting module is used to browse information and the house type recommended models according to the user, calculates the target wine
The probability that each house type is ordered by the user in shop, and the corresponding house type of maximum probability is selected, call the recommending module；
The recommending module is used to export the recommendation information on the house type.
8. commending system as claimed in claim 7, it is characterised in that if the house type that the searching modul finds is multiple,
Then the recommending module exports the recommendation information on cheapest house type in multiple house types；If the cheapest house type
It is not the house type of lowest price in the target hotel, then the recommending module is according to the cheapest house type and the mesh
Composite services type in mark hotel corresponding to the house type of lowest price, recommendation information of the output on the optimal house type of service.
9. commending system as claimed in claim 7, it is characterised in that the searching modul is additionally operable in the house type for finding
When lowest price is higher than default price, continuation searches whether the house type that there is next composite services type in target hotel；
If the lowest price of the house type that the searching modul finds is above default price, the recommending module output is on institute
State the recommendation information of the house type of lowest price in target hotel；If the house type of lowest price is multiple, basis in the target hotel
Composite services type corresponding to multiple house types, recommendation information of the output on the optimal house type of service.
10. commending system as claimed in claim 7, it is characterised in that the selecting module is used for the maximum probability in selection
When the price of corresponding house type orders the product of highest price and the preset multiple of house type higher than the user's history, selection time is high general
The corresponding house type of rate, until the house type of selection is inexpensive in the product；
If the price of each house type of the selecting module selection is above the product, the recommending module does not export any
Recommendation information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710150879.1A CN106934498A (en) | 2017-03-14 | 2017-03-14 | The recommendation method and system of hotel's house type in OTA websites |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710150879.1A CN106934498A (en) | 2017-03-14 | 2017-03-14 | The recommendation method and system of hotel's house type in OTA websites |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106934498A true CN106934498A (en) | 2017-07-07 |
Family
ID=59432356
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710150879.1A Pending CN106934498A (en) | 2017-03-14 | 2017-03-14 | The recommendation method and system of hotel's house type in OTA websites |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106934498A (en) |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107563816A (en) * | 2017-09-08 | 2018-01-09 | 携程计算机技术（上海）有限公司 | The Forecasting Methodology and system of the customer loss of e-commerce website |
| CN107704485A (en) * | 2017-07-31 | 2018-02-16 | 北京拉勾科技有限公司 | A kind of position recommends method and computing device |
| CN107944583A (en) * | 2017-12-08 | 2018-04-20 | 携程旅游网络技术（上海）有限公司 | Secondary distribution scene method for pushing, system, equipment and storage medium |
| CN108053275A (en) * | 2017-12-01 | 2018-05-18 | 深圳市天下房仓科技有限公司 | A kind of online product screening method, system and storage medium |
| CN108256067A (en) * | 2018-01-16 | 2018-07-06 | 平安好房（上海）电子商务有限公司 | Calculate method, apparatus, equipment and the storage medium of source of houses similarity |
| CN108334568A (en) * | 2018-01-16 | 2018-07-27 | 平安好房（上海）电子商务有限公司 | Source of houses method for pushing, device, equipment and computer readable storage medium |
| CN108446351A (en) * | 2018-03-08 | 2018-08-24 | 携程计算机技术（上海）有限公司 | The hotel's screening technique and system based on user preference of OTA platforms |
| CN108491395A (en) * | 2018-02-06 | 2018-09-04 | 北京三快在线科技有限公司 | Hotel's house type polymerization, device, electronic equipment and storage medium |
| CN109146096A (en) * | 2018-09-07 | 2019-01-04 | 上海携程商务有限公司 | Report processing method, device, storage medium and the electronic equipment of event for repairment |
| CN109325806A (en) * | 2018-09-20 | 2019-02-12 | 北京小度信息科技有限公司 | A kind of processing method and processing device of user information |
| CN109345319A (en) * | 2018-10-31 | 2019-02-15 | 深圳市元征科技股份有限公司 | A kind of house intelligent recommendation method, apparatus, equipment and storage medium |
| CN109740066A (en) * | 2019-01-28 | 2019-05-10 | 北京三快在线科技有限公司 | Information recommendation method, information recommending apparatus, storage medium and electronic equipment |
| CN109960777A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | Goods review personalized recommendation method, system, electronic equipment and storage medium |
| WO2019218654A1 (en) * | 2018-05-15 | 2019-11-21 | 北京三快在线科技有限公司 | Product ordering method |
| CN111340553A (en) * | 2020-02-28 | 2020-06-26 | 山东爱城市网信息技术有限公司 | Financial service platform product personalized recommendation method and system |
| CN112418464A (en) * | 2020-09-23 | 2021-02-26 | 深圳市金溢科技股份有限公司 | Expressway service area reservation method, RSU, OBU, platform and system |
| CN112446514A (en) * | 2019-09-03 | 2021-03-05 | 阿里巴巴集团控股有限公司 | Information booking method and device and electronic equipment |
| CN113139667A (en) * | 2021-05-07 | 2021-07-20 | 深圳他米科技有限公司 | Hotel room recommendation method, device, equipment and storage medium based on artificial intelligence |
| CN113724039A (en) * | 2021-08-03 | 2021-11-30 | 携程计算机技术（上海）有限公司 | Hotel recommendation method, system and device for OTA website and readable storage medium |
| CN113763134A (en) * | 2021-09-22 | 2021-12-07 | 携程旅游信息技术（上海）有限公司 | Information recommendation method, system, device and storage medium |
| CN117436709A (en) * | 2023-12-20 | 2024-01-23 | 四川宽窄智慧物流有限责任公司 | Cross-region order data overall warning method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101271558A (en) * | 2008-05-16 | 2008-09-24 | 华东师范大学 | Multi-policy commercial product recommending system based on context information |
| CN105184618A (en) * | 2015-10-20 | 2015-12-23 | 广州唯品会信息科技有限公司 | Commodity individual recommendation method for new users and system |
| CN105678570A (en) * | 2015-12-31 | 2016-06-15 | 北京京东尚科信息技术有限公司 | Method and apparatus for identifying potential users of E-commerce |
| CN106485562A (en) * | 2015-09-01 | 2017-03-08 | 苏宁云商集团股份有限公司 | A kind of commodity information recommendation method based on user's history behavior and system |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
-
2017
- 2017-03-14 CN CN201710150879.1A patent/CN106934498A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101271558A (en) * | 2008-05-16 | 2008-09-24 | 华东师范大学 | Multi-policy commercial product recommending system based on context information |
| CN106485562A (en) * | 2015-09-01 | 2017-03-08 | 苏宁云商集团股份有限公司 | A kind of commodity information recommendation method based on user's history behavior and system |
| CN105184618A (en) * | 2015-10-20 | 2015-12-23 | 广州唯品会信息科技有限公司 | Commodity individual recommendation method for new users and system |
| CN105678570A (en) * | 2015-12-31 | 2016-06-15 | 北京京东尚科信息技术有限公司 | Method and apparatus for identifying potential users of E-commerce |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107704485A (en) * | 2017-07-31 | 2018-02-16 | 北京拉勾科技有限公司 | A kind of position recommends method and computing device |
| CN107563816A (en) * | 2017-09-08 | 2018-01-09 | 携程计算机技术（上海）有限公司 | The Forecasting Methodology and system of the customer loss of e-commerce website |
| CN108053275A (en) * | 2017-12-01 | 2018-05-18 | 深圳市天下房仓科技有限公司 | A kind of online product screening method, system and storage medium |
| CN107944583A (en) * | 2017-12-08 | 2018-04-20 | 携程旅游网络技术（上海）有限公司 | Secondary distribution scene method for pushing, system, equipment and storage medium |
| CN107944583B (en) * | 2017-12-08 | 2021-02-02 | 携程旅游网络技术（上海）有限公司 | Secondary selling scene pushing method, system, equipment and storage medium |
| CN109960777A (en) * | 2017-12-25 | 2019-07-02 | 北京京东尚科信息技术有限公司 | Goods review personalized recommendation method, system, electronic equipment and storage medium |
| CN109960777B (en) * | 2017-12-25 | 2022-12-02 | 北京京东尚科信息技术有限公司 | Personalized recommendation method and system for article comment, electronic equipment and storage medium |
| CN108334568B (en) * | 2018-01-16 | 2020-08-21 | 平安好房（上海）电子商务有限公司 | House resource pushing method, device, equipment and computer readable storage medium |
| CN108256067A (en) * | 2018-01-16 | 2018-07-06 | 平安好房（上海）电子商务有限公司 | Calculate method, apparatus, equipment and the storage medium of source of houses similarity |
| CN108334568A (en) * | 2018-01-16 | 2018-07-27 | 平安好房（上海）电子商务有限公司 | Source of houses method for pushing, device, equipment and computer readable storage medium |
| CN108491395A (en) * | 2018-02-06 | 2018-09-04 | 北京三快在线科技有限公司 | Hotel's house type polymerization, device, electronic equipment and storage medium |
| CN108491395B (en) * | 2018-02-06 | 2022-03-29 | 北京三快在线科技有限公司 | Hotel room type aggregation method and device, electronic equipment and storage medium |
| CN108446351A (en) * | 2018-03-08 | 2018-08-24 | 携程计算机技术（上海）有限公司 | The hotel's screening technique and system based on user preference of OTA platforms |
| CN108446351B (en) * | 2018-03-08 | 2022-03-22 | 携程计算机技术（上海）有限公司 | Hotel screening method and system based on user preference of OTA platform |
| WO2019218654A1 (en) * | 2018-05-15 | 2019-11-21 | 北京三快在线科技有限公司 | Product ordering method |
| CN109146096A (en) * | 2018-09-07 | 2019-01-04 | 上海携程商务有限公司 | Report processing method, device, storage medium and the electronic equipment of event for repairment |
| CN109146096B (en) * | 2018-09-07 | 2020-11-17 | 上海携程商务有限公司 | Processing method and device for repair event, storage medium and electronic equipment |
| CN109325806A (en) * | 2018-09-20 | 2019-02-12 | 北京小度信息科技有限公司 | A kind of processing method and processing device of user information |
| CN109345319A (en) * | 2018-10-31 | 2019-02-15 | 深圳市元征科技股份有限公司 | A kind of house intelligent recommendation method, apparatus, equipment and storage medium |
| CN109740066A (en) * | 2019-01-28 | 2019-05-10 | 北京三快在线科技有限公司 | Information recommendation method, information recommending apparatus, storage medium and electronic equipment |
| CN109740066B (en) * | 2019-01-28 | 2021-05-28 | 北京三快在线科技有限公司 | Information recommendation method, information recommendation device, storage medium and electronic equipment |
| CN112446514A (en) * | 2019-09-03 | 2021-03-05 | 阿里巴巴集团控股有限公司 | Information booking method and device and electronic equipment |
| CN111340553A (en) * | 2020-02-28 | 2020-06-26 | 山东爱城市网信息技术有限公司 | Financial service platform product personalized recommendation method and system |
| CN112418464A (en) * | 2020-09-23 | 2021-02-26 | 深圳市金溢科技股份有限公司 | Expressway service area reservation method, RSU, OBU, platform and system |
| CN113139667A (en) * | 2021-05-07 | 2021-07-20 | 深圳他米科技有限公司 | Hotel room recommendation method, device, equipment and storage medium based on artificial intelligence |
| CN113139667B (en) * | 2021-05-07 | 2024-02-20 | 深圳他米科技有限公司 | Hotel room recommending method, device, equipment and storage medium based on artificial intelligence |
| CN113724039A (en) * | 2021-08-03 | 2021-11-30 | 携程计算机技术（上海）有限公司 | Hotel recommendation method, system and device for OTA website and readable storage medium |
| CN113763134A (en) * | 2021-09-22 | 2021-12-07 | 携程旅游信息技术（上海）有限公司 | Information recommendation method, system, device and storage medium |
| CN117436709A (en) * | 2023-12-20 | 2024-01-23 | 四川宽窄智慧物流有限责任公司 | Cross-region order data overall warning method |
| CN117436709B (en) * | 2023-12-20 | 2024-03-19 | 四川宽窄智慧物流有限责任公司 | Cross-region order data overall warning method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106934498A (en) | The recommendation method and system of hotel's house type in OTA websites | |
| Chetan Panse et al. | Understanding consumer behaviour towards utilization of online food delivery platforms | |
| CN104281961B (en) | For the advertisement in on-line system and the quality score system of content | |
| TWI603273B (en) | Method and device for placing information search | |
| US20170097763A1 (en) | Sports and concert event ticket pricing and visualization system | |
| CN104281962B (en) | For the advertisement in on-line system and the united market of content | |
| CN107194774A (en) | Personalized hotel's commending system and method in OTA websites | |
| AU2012378630B2 (en) | Categorizing and ranking travel-related search results | |
| US20130290324A1 (en) | Categorizing and ranking travel-related database query results | |
| US20120173310A1 (en) | Deal quality for event tickets | |
| CN106991576A (en) | A kind of heating power of geographic area shows method and apparatus | |
| CN108205768A (en) | Database building method and data recommendation method and device, equipment and storage medium | |
| CN110428298A (en) | A kind of shop recommended method, device and equipment | |
| CN107230098A (en) | Method and system is recommended in a kind of timesharing of business object | |
| CN101563701A (en) | Information providing system | |
| CN109509039A (en) | Method for building up and system, the Method of Commodity Recommendation and system of price expectation model | |
| CN109582876A (en) | Tourism industry user portrait building method, device and computer equipment | |
| CN108053282A (en) | A kind of method for pushing of combined information, device and terminal | |
| CN109461030A (en) | A kind of O2O evaluation method and device based on block chain | |
| US20130066800A1 (en) | Method of aggregating consumer reviews | |
| CN108648058A (en) | Model sequencing method and device, electronic equipment, storage medium | |
| CN107742245A (en) | A kind of merchant information recommends method, apparatus and equipment | |
| CN107481093A (en) | Personalized shop Forecasting Methodology and device | |
| CN103843026B (en) | Information processor, information processing method | |
| CN108446351A (en) | The hotel's screening technique and system based on user preference of OTA platforms |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170707 |