CN108537373A - Travel information recommends method and apparatus - Google Patents
Travel information recommends method and apparatus Download PDFInfo
- Publication number
- CN108537373A CN108537373A CN201810259835.7A CN201810259835A CN108537373A CN 108537373 A CN108537373 A CN 108537373A CN 201810259835 A CN201810259835 A CN 201810259835A CN 108537373 A CN108537373 A CN 108537373A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- sight spot
- neural network
- module
- 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.)
- Withdrawn
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"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
-
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
This application discloses a kind of travel informations to recommend method and apparatus.This method includes：User information obtaining step：The first user data is obtained, first user data includes the identity information and user history information of the first user；With recommendation information generation step：By the trained deep neural network of the first user data input, travel information recommendation results are generated, the travel information recommendation results characterize the probability that first user will go at least one sight spot tourism；Wherein, deep neural network is obtained by training step, and the training step includes：User data annotation step, scene data obtaining step, model prediction step and Modifying model step.The method that this method uses machine learning recommends tourist attractions for user automatically, saves the time of user, bring completely new better user experience by obtaining and analyzing the behavioral data of user.
Description
Technical field
This application involves field of information processing, recommend method and apparatus more particularly to a kind of travel information.
Background technology
Currently, the application program (APP) in terms of tourism for user when recommending tourist attractions, it is general using following several
Kind method.A kind of method is the target scenic spot for needing user voluntarily to select to want to go to, and recommends sight spot according to the user's choice and makes
Make travel schedule, this method calculates after needing user to input target scenic spot, when user does not know which sight spot or not wanted to go to
When understanding very much sight spot information, this method is simultaneously not suitable for, therefore practicability is relatively low, is not suitable for general public.There is a method in which by
The profession customization teacher of travel agency completes by manually, and first, customization teacher is it should be understood that a large amount of travel information and information, to each
The characteristics of ground, is well understood by；Secondly, customization teacher, which requires a great deal of time, links up with user and according to communication result to recommending
Sight spot be adjusted, this method needs to spend a large amount of human resources, efficiency very low.
Invention content
The application's aims to overcome that the above problem or solves or extenuate to solve the above problems at least partly.
According to the one side of the application, a kind of travel information recommendation method is provided, this method includes：
User information obtaining step：The first user data is obtained, first user data includes the identity of the first user
Information and user history information；
Recommendation information generation step：By the trained deep neural network of the first user data input, trip is generated
Trip information recommendation will go the general of at least one sight spot tourism as a result, the travel information recommendation results characterize first user
Rate；
Wherein, deep neural network is obtained by training step, and the training step includes：
User data annotation step：Each user in gathering for user, obtains the travel history data of the user,
The data before time point T will be happened in the travel history data as the input data of training data, by the time
Other sight spots after point T except the user to first tourist attractions and the tourist attractions crossed are marked respectively, obtain reality
Border tourism result data；
Scene data obtaining step：Each sight spot in gathering for sight spot obtains scene data, the scene data
The behavioral data of essential information data and each user including the sight spot to the sight spot；
Model prediction step：By the scene data, the identity information of each user and the instruction of the user
Practice data to be input in deep neural network, obtains the Making by Probability Sets that each described user goes each sight spot tourism；
Modifying model step：The practical tourism result data and Making by Probability Sets are compared, to depth nerve
Network is modified, and obtains the trained deep neural network.
The method that this method uses machine learning the methods of is shared by sight spot picture and social activity and obtains and analyze user
Behavioral data, recommend tourist attractions automatically for user, user without selecting the tourist attractions to be gone that can be met in advance
User's heart is desired as a result, save the time of user, brings completely new better user experience.Social activity is shared etc. at this
It occupies an important position in the data analysis of application, data source can be provided for model, and provide to capture user behavior
Good platform.
Optionally, the model prediction step includes：
Input step：The training data and the scene data are input to the input layer of the deep neural network；
Processing step：By each item data difference in each item data and the scene data in the training data
It is converted into corresponding characteristic；
Step of converting：Convert characteristic corresponding with the training data to user characteristics matrix, and will be with institute
It states the corresponding characteristic of scene data and is converted into sight spot eigenmatrix；
Associated steps：The user characteristics matrix is associated with the sight spot eigenmatrix at each sight spot, described in calculating
User representated by user characteristics matrix removes the probability at the sight spot；
Export step：Calculating each user in user's set goes the probability set of each sight spot tourism to merge output
The Making by Probability Sets.
It can be trained by this method and be more in line with user's custom, the depth consistent with the recent behavior of user and psychology
Neural network model, so that more accurate by the result that the model is predicted.
Optionally, the Modifying model step includes：
Data type step of converting：Probability data in the Making by Probability Sets is compared with preset threshold value, by institute
State the data that probability data is converted into integer type；
Residual computations step：The data of the integer type are compared with the practical tourism result data, are obtained
Residual error data；
Amendment step：The deep neural network is modified by reversed neural propagation using the residual error data.
This method is analyzed by the agenda to user and is used as feedback result and is modified to model, to make
Model is more accurate, calculates the true requirement for being more nearly user and desired tourism recommendation results.
Optionally, this method further includes after the recommendation information generation step：
Sight spot preferred steps：Interval or the route planning in response to first user instruct at every predetermined time,
At least one sight spot for including according to the travel information recommendation results and first user are at least one sight spot
Behavioral data determine preferred sight spot；
Route planning step：According to object function and constraints, path planning is carried out to the preferred sight spot, generates trip
Row line.
This method also adds route planning step after recommending tourist attractions, and the viewpoint of tourist psychology is incorporated the party
In the design of method so that result is more in line with user demand.
Optionally, this method further includes after the route planning step：
User intent judgment step：Judge whether the itinerary meets the expection of first user, if it is,
It executes circuit and determines step, if it is not, then executing sight spot annotation step；
Route determination step：Judge whether the itinerary needs to change, if necessary to change, then according to described first
Modification of the user to the sight spot in the itinerary, again planning path；
Sight spot markers step：According to the feedback of first user, the preferred sight spot is labeled as first user
Behavioral data, so as to as the data needed for the training step.
The true feedback data of user can be obtained by this method, can be made the data as the data of correction model
It is more accurate to obtain model.
According to further aspect of the application, a kind of travel information recommendation apparatus is provided, including：
User profile acquisition module is configured to obtain the first user data, and first user data includes the first use
The identity information and user history information at family；
Recommendation information generation module is configured to the trained depth nerve net of first user data input
Network generates travel information recommendation results, and the travel information recommendation results, which characterize first user, will go at least one sight spot
The probability of tourism；
Wherein, deep neural network is obtained by training module, and the training module includes：
User data labeling module, each user being configured in gathering for user, obtains the tourism of the user
Historical data, will be happened at the data before time point T as the input data of training data in the travel history data, will
Other sight spots after the time point T except the user to first tourist attractions and the tourist attractions crossed are respectively into rower
Note, obtains practical tourism result data；
Scene data acquisition module is configured to each sight spot in gathering for sight spot, obtains scene data, described
Scene data includes behavioral data of the essential information data at the sight spot with the user to the sight spot；
Model prediction module is configured to the training data and the scene data being input to deep neural network
In, obtain the Making by Probability Sets that each user in user's set goes each sight spot tourism；
Modifying model module is configured to the practical tourism result data and Making by Probability Sets being compared, to described
Deep neural network is modified, and obtains the trained deep neural network.
The device is pushed away for user using the method for machine learning by obtaining and analyzing the behavioral data of user automatically
Recommend tourist attractions, user without select in advance the tourist attractions to be gone can obtain meeting user's heart it is desired as a result, saving
The time of user, bring completely new better user experience.
Optionally, the model prediction module includes：
Input module is configured to the training data and the scene data being input to the deep neural network
Input layer；
Processing module is configured to each single item in each item data and the scene data in the training data
Data are separately converted to corresponding characteristic；
Conversion module is configured to convert characteristic corresponding with the training data to user characteristics matrix, and
And convert characteristic corresponding with the scene data to sight spot eigenmatrix；
Relating module, be configured to the user characteristics matrix is associated with the sight spot eigenmatrix at each sight spot,
Calculate the probability that the user representated by the user characteristics matrix goes to the sight spot；
Output module is configured to calculate the probability set that each user in user's set goes each sight spot tourism
Merge and exports the Making by Probability Sets.
Optionally, which is also associated with after the recommendation information generating means：
Sight spot preferred module is configured to be spaced at every predetermined time or be advised in response to the route of first user
Instruction is drawn, at least one sight spot for including according to the travel information recommendation results and first user are to described at least one
The behavioral data at a sight spot determines preferred sight spot；
Route planning module, is configured to according to object function and constraints, and path rule are carried out to the preferred sight spot
It draws, generates itinerary.
According to further aspect of the application, a kind of computer equipment is provided, including memory, processor and be stored in
In the memory and the computer program that can be run by the processor, wherein the processor execution computer journey
Realize that travel information as described above recommends method when sequence.
According to further aspect of the application, provide a kind of computer readable storage medium, it is preferably non-volatile can
Storage medium is read, is stored with computer program, the computer program is realized as described above when executed by the processor
Travel information recommends method.
According to the accompanying drawings to the detailed description of the specific embodiment of the application, those skilled in the art will be more
Above-mentioned and other purposes, the advantages and features of the application are illustrated.
Description of the drawings
Some specific embodiments of the application are described in detail by way of example rather than limitation with reference to the accompanying drawings hereinafter.
Identical reference numeral denotes same or similar component or part in attached drawing.It should be appreciated by those skilled in the art that these
What attached drawing was not necessarily drawn to scale.In attached drawing：
Fig. 1 is the flow chart of the one embodiment for recommending method according to the travel information of the application；
Fig. 2 is the flow chart according to one embodiment of the training step of the present processes；
Fig. 3 is the flow chart according to one embodiment of the model prediction step of the present processes；
Fig. 4 is the flow chart according to one embodiment of the Modifying model step of the present processes；
Fig. 5 is the flow chart for another embodiment for recommending method according to the travel information of the application；
Fig. 6 is the block diagram according to one embodiment of the deep neural network of the application；
Fig. 7 is the block diagram according to one embodiment of the travel information recommendation apparatus of the application；
Fig. 8 is the block diagram according to one embodiment of the training module of the device of the application；
Fig. 9 is the block diagram according to one embodiment of the model prediction module of the device of the application；
Figure 10 is the block diagram according to another embodiment of the travel information recommendation apparatus of the application.
Specific implementation mode
According to the accompanying drawings to the detailed description of the specific embodiment of the application, those skilled in the art will be more
Above-mentioned and other purposes, the advantages and features of the application are illustrated.
According to the one side of the application, a kind of travel information recommendation method is provided.Fig. 1 is the tourism according to the application
The flow chart of one embodiment of information recommendation method.This method includes：
User information obtaining step：The first user data is obtained, first user data includes the identity of the first user
Information and user history information；
Recommendation information generation step：By the trained deep neural network of the first user data input, trip is generated
Trip information recommendation will go the general of at least one sight spot tourism as a result, the travel information recommendation results characterize first user
Rate.
First user data is used to describe the essential information and historical information of user, and essential information can be identity information,
Essential information may include one or more of following data：User identity information (User ID), name, gender, the age,
Occupation, family status, user tag.User tag may include tourism favor and personal preference for example：Exploration, movement, sandy beach,
Leisure, sightseeing, activity, culture, exploration, party, people, food, ocean, the mountains and rivers, nature, city, museum, travelling, peace and quiet, the back of the body
Packet, open air, building, lake, sunset, forest, sunrise, warm, cold etc..Historical information be used for describe user social platform
On behavior history for example, social platform may include social network sites and/or application program (APP).The historical information can wrap
Include one or more of following data：User's travel history data, user and relevant retrieval record of travelling, consumption are practised
Used and attentinal contents.In a preferred embodiment, user's travel history data can be used for describing user in upper tourism
History.For example, " feature " at each sight spot that user went.For example, being gone about the user of being delivered in social platform of user
The record of which sight spot tourism, the data can be picture and/or word, and word may include daily record, the state leaving a message, deliver
Deng.User search data may include that user examines to some sight spot in social platform or with the relevant information in the sight spot
The record of rope.May include one or more of following data with the relevant information in the sight spot：U.S. food, shelter around sight spot
Place, traffic, ticket and other relevant sight spots of the sight spot.
Then, it by the trained deep neural network of the first user data input, generates travel information and recommends knot
Fruit, the travel information recommendation results characterize the probability that first user will go at least one sight spot tourism.If used
The probability at multiple sight spots will be removed in family, then probability sorts according to sequence from big to small.It can be with preferential recommendation and indicating probability most
High sight spot, several sight spots before can also showing.
Deep neural network model is also referred to as multilayer perceptron model, be it is a kind of there is the preceding artificial neural network to structure,
It maps one group and inputs data into one group of output data.It can be counted as a digraph, be made of multiple node layers, each
Layer is connected to next layer entirely.In addition to input node, each node be one with nonlinear activation function neuron (or
Processing unit).Deep neural network generally comprises input layer, hidden layer and output layer, and wherein hidden layer includes at least one layer,
It can be two layers or more.The flow of neural network is divided into forward process and reverse procedure.Forward process is generally used for predicting, reversely
Process is generally used for training.
There is weights, biasing and the activation primitive for an input in forward process, on each neuron.Activation
Function may include one or more of identity, sigmoid, ReLU and its variant.Using the deep neural network
During being predicted, the data are inputted into input layer, i.e. first layer, exported after neural n ary operation as a result,
Then, using the output result of first layer as the input of the second layer, and so on, until output layer exports final result.If
The deep neural network is trained, then weights and biasing has determined.For a new input, by the above process,
Prediction result can be exported.In this application, being associated between user data and scene data is established in deep neural network
Relationship, therefore by the first user data input deep neural network, travel information recommendation results can be obtained, which can
To characterize the probability that first user will go at least one sight spot tourism.
In reverse procedure, deep neural network is obtained by following training step, and Fig. 2 is according to the present processes
The flow chart of one embodiment of training step.The training step includes：
User data annotation step：Each user in gathering for user, obtains the travel history data of the user,
The data before time point T will be happened in the travel history data as the input data of training data, by the time
Other sight spots after point T except the user to first tourist attractions and the tourist attractions crossed are marked respectively, obtain reality
Border tourism result data.Training data can also include the essential information of user.
User's set can register all users of one or more and relevant social platform of travelling.It can at one
In the embodiment of choosing, which can be the particular platform researched and developed according to the demand of the application, the platform can be with
The platform that tourism is the theme, user can deliver on the platform and the relevant article of tourism, picture, comment, in addition, user is also
Can thumb up, click platform recommendation connection, payment etc..
The travel history data of user can be generated after user operates on the platform it is related to tourism
Data.With some time point, such as time point T is node, which is divided into two parts, by the tourism
Data before being happened at time point T in historical data extract, using the data as the input data of training data.To institute
State after time point T until the travel history data between current time are analyzed, can obtain the user whether go some or
Certain sight spot tourisms.If user, to excessively at least one tourist attractions, first tourist attractions which is gone to mark
It is 1, other tourist attractions is labeled as 0, other sight spots can be that user went after going to first field sight spot,
Can be that user did not go.A certain range of scape around the tourist attractions that the tourist attractions that do not go can be
Point.It is understood that other identification means can also be used, as long as can be by went first tourist attractions and other
Tourist attractions distinguish.It is obtained the result is that practical tourism result data after each sight spot is marked.For
One user, can be arranged several time points T, for example, T can take on January 1st, 2015, on July 1st, 2016 etc., pass through
The entire data of one user can be generated multiple training datas by such mode, multiple samples be formed, to increase training
The data volume of data set so that the training result of deep neural network model is more accurate.
The training step further includes：Scene data obtaining step.Each sight spot in gathering for sight spot obtains scape
Point data, the scene data include behavioral data of the essential information data at the sight spot with each user to the sight spot；
Gather at sight spot：The sight spot in range of countries, worldwide sight spot where the user and activity
Place.Wherein, space for activities includes but not limited to：Friendship ties assembly place, concert place, fashion show place etc..The base at sight spot
This information data may include the one or several kinds in following data：Sight spot identification information (sight spot ID), sight spot type, text
Chapter, picture, scoring, comment.Wherein, article includes description and/or introduces the article at the sight spot, introduces shop, quotient in the sight spot
The article of family.Article can be daily record.Comment includes the comment to sight spot, article, picture, sight spot periphery.Daily record and comment
Semantic analysis technology realization may be used in intellectual analysis.The essential information data at sight spot be for user set in all users into
Capable statistical analysis.
User in user's set can also be referred to as user to the behavioral data at the sight spot and rely on data, the behavior
Degree of dependence of the data characterization user for the sight spot.For example, for the first user, scene data obtaining step is being carried out
When, for some sight spot, behavior data include that the first user thumbs up the sight spot, browsing article related with the sight spot,
The duration of picture, number, comment.Behavioral data is the statistical carried out for the behavior of a certain user in data markers step
Analysis.
The training step further includes：Model prediction step.The identity of the scene data, each user are believed
Breath and the training data of the user are input in deep neural network, obtain each described user and each sight spot is gone to travel
Making by Probability Sets.
Fig. 3 is the flow chart according to one embodiment of the model prediction step of the present processes.In an optional reality
It applies in scheme, the model prediction step may include：Input step：The training data and the scene data are input to
The input layer of the deep neural network.
The model prediction step can also include processing step：By each item data in the training data and described
Each item data in scene data is separately converted to corresponding characteristic.Characteristic can be the form of vector.It is each
Item data and characteristic are one-to-one relationships.For example, converting id information to ID features, it is converting article to article
Feature.After input layer is handled the data of input, obtained characteristic is transferred to hidden layer.Tourism for user
Historical data and for travel it is relevant retrieval record both data, input when, multiple records can be inputted, handled
The characteristic obtained in step is the average characteristics data of multiple records.With the travel history data instance of user, due to
In some period, travel history data volume may be very big, when carrying out feature conversion to each travel history data, equal energy
Obtain a vector, the limitation due to model to data entry format, when carrying out model training, by the multiple every of user
Average characteristics data that the corresponding characteristic of one travel history data obtains after being averaged substitute into deep neural network into
Row operation." user's travel history information characteristics ", gained knot are obtained in such a way that the summation of travel history data is taken mean value again
Fruit can both meet requirement of the model to data format, can also embody all travel histories.
The model prediction step can also include step of converting：It will characteristic conversion corresponding with the training data
For user characteristics matrix, and convert characteristic corresponding with the scene data to sight spot eigenmatrix.Step of converting
It can be carried out in hidden layer, hidden layer is further calculated and handled to characteristic, and eigenmatrix is obtained, and calculating is used
Hidden layer can be one layer, can also be multilayer.Optionally, all characteristics related to user can be converted into one
A user characteristics matrix can be converted into a sight spot eigenmatrix with the relevant all characteristics in sight spot.
The model prediction step can also include associated steps：By the scape of the user characteristics matrix and each sight spot
Point feature matrix correlation joins, and calculates the probability that the user representated by the user characteristics matrix goes to the sight spot.Associated steps also may be used
To be carried out in hidden layer.
The model prediction step can also include output step：Each user calculated in user's set goes often
The probability set of a sight spot tourism, which merges, exports the Making by Probability Sets.The step can be carried out in output layer.In the training process, deep
The output layer output for spending neural network is probability, and which characterizes users to go the possibility that the sight spot is traveled.
Fig. 6 be according to the block diagram of one embodiment of the deep neural network of the application, optionally, deep neural network mould
Type includes three layers, and in the figure, lowermost layer indicates that input layer, sublevel second from the bottom are hidden layer, and layer third from the bottom is output layer.
The data handled in partial data and hidden layer in input layer are connected by full articulamentum, for example, the user in training data
Scape in ID, gender, occupation, age, tourism favor, the travel history data of user, user search record and scene data
Point ID, it sight spot type, sight name, scoring, comments on, thumb up the thumbing up, to picture or Log Browser to sight spot with the first user
Duration and number.For the daily record in scene data, between input layer and hidden layer even using convolutional neural networks (CNN)
It connects.For the picture in scene data, connected using recurrent neural network (RNN) before input layer and hidden layer.It " is used in the figure
The family sight spot degree of correlation " is probability value.The loss function of training pattern is cross entropy.After inputting N number of sight spot information, it can obtain N number of
" user sight spot degree of correlation feature " probability value, is ranked up according to the probability value.User characteristics and user sight spot degree of correlation feature
Between using softmax activation primitives realize, it is not of the same race which can be applied to multilayer neural network, convolutional neural networks etc.
In the neural network of class.
The application extracts feature by deep learning, such as the pass of user and sight spot are calculated using deep neural network model
Connection property.Data similar in feature will move closer in model hidden layer, finally obtain eigenmatrix, and output layer can be to scape
Point and/or activity are ranked up and recommend to user.
The training step further includes：Modifying model step.Fig. 4 is the Modifying model step according to the present processes
The flow chart of one embodiment.In amendment step, the practical tourism result data and Making by Probability Sets are compared, to institute
It states deep neural network to be modified, obtains the trained deep neural network.
Wherein, the Modifying model step includes data type step of converting：By the probability data in the Making by Probability Sets
It is compared with preset threshold value, converts the probability data to the data of integer type.The purpose of the step is by probability
Data are consistent with the practical tourism type of result data.
The Modifying model step further includes residual computations step：By the data of the integer type and the practical tourism
Result data is compared, and obtains residual error data.Probability data can be converted to after integer and make with practical tourism result poor, obtained
To residual error data.
The Modifying model step further includes amendment step：Using the residual error data by reversed neural propagation to described
Deep neural network is modified.Modification method may include：Test of outlier, homogeneity test of variance, the normality inspection of error
It tests, correlation test and the variance-stabilizing transformation accompanied, normal transformations etc..
In this way, not needing user inputs the target scenic spot wanted, planner that also should not be special to user into
Row consulting, by being analyzed the behavior of user the tourist attractions that can be automatically analyzed out client and want.This method is being set
Timing considers the knowledge of tourist psychology, therefore more intelligence and hommization, and recommendation results are also more accurate.
Referring to Fig. 1, optionally, this method after the recommendation information generation step can also include sight spot preferred steps and
Route planning step.Sight spot preferred steps include：Interval or the route in response to first user at every predetermined time
Planning instruction, at least one sight spot for including according to the travel information recommendation results and first user to it is described at least
The behavioral data at one sight spot determines preferred sight spot.The route planning step is according to object function and constraints, to described excellent
It selects sight spot to carry out path planning, generates itinerary.
Two methods may be used after recommending tourist attractions step and carry out path planning.The first is using periodically recommendation
The method of tourism route.For example, being spaced at every fixed time, for example, a week or one month, push away according to travel information
It recommends as a result, carrying out path planning according to one of those or several tourist attractions.When being planned using multiple sight spots, this
A little sight spots are preferably belonging to a city or apart from closer.For example, according to weekend, liberty, the time span of long vacation
Number, the sight name at sight spot are selected for user, and carry out path planning.Using this method, user time can be saved, and
And practicable trip scheme can be provided to the user.
In another optional embodiment, the method that compartmentalization path is recommended may be used.In this embodiment,
User can select at least one city or a regional extent, and then system selection from recommendation tourist attractions belongs to the city
City or the sight spot of the regional extent carry out path planning.Using this method, the travelling route made is more in line with user's heart
Reason is expected.
Optionally, this method can also directly carry out route planning step in the recommendation information generation step.The route
Planning step includes：The whole for include to the travel information recommendation results according to object function and constraints or portion
Divide sight spot to carry out path planning, generates itinerary.
Using this method, intellectual analysis can be carried out according to the behavioral data of user and obtain tourist attractions probability, according to trip
It swims sight spot probability and carries out path planning, this method is combined user behavior and tourist psychology, is more in line with user's
Demand and in-mind anticipation.
In an optional embodiment, the object function of path planning can be as shown in formula (1)：
Wherein, R is the path of the overall distance minimum of planning, cijFor sight spot i to the distance of sight spot j, xijIt indicates from sight spot
I to sight spot j whether there is path, if xij=1, indicate there is the path being connected between two sight spots, if xij=0, then table
Show, path is not present between two sight spots, and n indicates the sum at sight spot.
The bound for objective function includes：
0≤xij≤ 1, i=1 ..., n, j=1 ..., n (2)
Wherein, formula (3) indicates that only there are one sight spots to reach sight spot j in other sight spots, and formula (4) expression goes out from sight spot i
Hair can only reach a sight spot in other sight spots.
Optionally, which can also include at least one of the following conditions：
(1) the sight spot open hour, article which can deliver from user in the social platform of the application, comment,
Other websites or channel obtain；
(time needed for 2 visit sight spots, the time can be go sight-seeing under normal conditions time needed for some sight spot or
Using the average value under statistical conditions, calculating multiple users and browsing the sight spot time；
(3) vehicles, the constraints can determine according to the user's choice, can also divide according to other users
Analysis determines, or obtains according to analyses such as the travel history data of the user, retrieval records the tourism of the user and go on a journey custom；
(4) the maximum bearing capacity of user, maximum bearing capacity show the user daily go out the time maximum time.The data
It can be calculated by machine learning and user's history tourism favor is learnt and predicts to obtain, can also such as navigated by routing
Class's information obtains.For example, if calculated by machine learning, it is leisurely and carefree type to analyze the user, then shows its maximum carrying
It measures less than normal, then can arrange less sight spot to the user daily；If the user is the type of checking card, maximum bearing capacity is inclined
Greatly, then more sight spot can be arranged to the user daily.
(5) starting point and terminal of stroke, it is to be understood that starting point and terminal can be the same or different, should
Data can be obtained by user's definition；
(6) one days path lengths or the total amount of distance time are no more than first threshold.The total amount of path length or time
It does not include distance and the time of sight-seeing spot.First threshold user's history tourism data can be predicted by machine learning and
It obtains, or is obtained by User Defined.
Sight-seeing spot time total amount is no more than second threshold within (7) one day.Sight-seeing spot time total amount does not include disappearing in distance
Time-consuming.Second threshold can predict to obtain to user's history tourism data by machine learning or User Defined obtains.It can be with
Understand, the 6th point and the 7th point can also be combined and form a constraints.
(8) process context degree.Specifically, such as：If user visits before requiring sight spot A that must come sight spot B, system
Then judge restrictive condition.For example, determining the priority at sight spot according to application scenarios.Wherein, application scenarios can be the time.If
Limited time, then sight spot A compared with sight spot B, the priority of sight spot A is higher than sight spot B, and sight spot B can not be in route programming result
In the middle.The data of priority can be by obtaining the sequence at sight spot in travel information recommendation results.
Fig. 5 is the flow chart for another embodiment for recommending method according to the travel information of the application.In the embodiment, trip
Swimming information recommendation method can also include after the route planning step：
User intent judgment step：Judge whether the itinerary meets the expection of first user, if it is,
It executes circuit and determines step, if it is not, then executing sight spot annotation step；
Route determination step：Judge whether the itinerary needs to change, if necessary to change, then according to described first
Modification of the user to the sight spot in the itinerary, again planning path；
Sight spot markers step：According to the feedback of first user, the preferred sight spot is labeled as first user
Behavioral data, so as to as the data needed for the training step.
According to further aspect of the application, a kind of travel information recommendation apparatus is additionally provided, Fig. 7 is according to the application
The block diagram of one embodiment of travel information recommendation apparatus.The device includes：
User profile acquisition module is configured to obtain the first user data, and first user data includes the first use
The identity information and user history information at family；
Recommendation information generation module is configured to the trained depth nerve net of first user data input
Network generates travel information recommendation results, and the travel information recommendation results, which characterize first user, will go at least one sight spot
The probability of tourism；
Wherein, deep neural network is obtained by training module, and the training module includes：
User data labeling module, each user being configured in gathering for user, obtains the tourism of the user
Historical data, will be happened at the data before time point T as the input data of training data in the travel history data, will
Other sight spots after the time point T except the user to first tourist attractions and the tourist attractions crossed are respectively into rower
Note, obtains practical tourism result data；
Scene data acquisition module is configured to each sight spot in gathering for sight spot, obtains scene data, described
Scene data includes behavioral data of the essential information data at the sight spot with the user to the sight spot；
Model prediction module is configured to the training data and the scene data being input to deep neural network
In, obtain the Making by Probability Sets that each user in user's set goes each sight spot tourism；
Modifying model module is configured to the practical tourism result data and Making by Probability Sets being compared, to described
Deep neural network is modified, and obtains the trained deep neural network.
The device is pushed away for user using the method for machine learning by obtaining and analyzing the behavioral data of user automatically
Recommend tourist attractions, user without select in advance the tourist attractions to be gone can obtain meeting user's heart it is desired as a result, saving
The time of user, bring completely new better user experience.
Fig. 8 is the block diagram according to one embodiment of the training device of the device of the application.In this embodiment, model is repaiied
Positive module includes：
Data type conversion module, be disposed for by probability data and the preset threshold value in the Making by Probability Sets into
Row compares, and converts the probability data to the data of integer type；
Residual computations module, be disposed for by the data of the integer type with it is described it is practical travel result data into
Row compares, and obtains residual error data；
Correcting module is disposed for being modified the deep neural network using the residual error data.
Fig. 9 is the block diagram according to one embodiment of the model training apparatus of the device of the application.Optionally, the model
Prediction module includes：
Input module is configured to the training data and the scene data being input to the deep neural network
Input layer；
Processing module is configured to each single item in each item data and the scene data in the training data
Data are separately converted to corresponding characteristic；
Conversion module is configured to convert characteristic corresponding with the training data to user characteristics matrix, and
And convert characteristic corresponding with the scene data to sight spot eigenmatrix；
Relating module, be configured to the user characteristics matrix is associated with the sight spot eigenmatrix at each sight spot,
Calculate the probability that the user representated by the user characteristics matrix goes to the sight spot；
Output module is configured to calculate the probability set that each user in user's set goes each sight spot tourism
Merge and exports the Making by Probability Sets.
Referring to Fig. 7, optionally, which is also associated with after the recommendation information generating means：
Sight spot preferred module is configured to be spaced at every predetermined time or be advised in response to the route of first user
Instruction is drawn, at least one sight spot for including according to the travel information recommendation results and first user are to described at least one
The behavioral data at a sight spot determines preferred sight spot；
Route planning module, is configured to according to object function and constraints, and path rule are carried out to the preferred sight spot
It draws, generates itinerary.
Figure 10 is the block diagram according to another embodiment of the travel information recommendation apparatus of the application.Optionally, in the dress
In setting, the route planning module is also associated with：
User intent judgment module, judges whether the itinerary meets the expection of first user, if it is,
Circuit determining module is executed, if it is not, then executing sight spot labeling module；
Route determination module is disposed for judging whether the itinerary needs to change, if necessary to change, then
Modification according to first user to the sight spot in the itinerary, again planning path；With
Sight spot mark module is disposed for the feedback according to first user, and the preferred sight spot is labeled as
The behavioral data of first user, so as to as the data needed for the training step.
The application is used replaces people to carry out decision using artificial intelligence, first passes through specific channel and obtains data information, finally
Realize the recommendation function of the full-automatic personalized itinerary for generating and meeting user images.
The application obtains the trip behavior feature of user, tourist image design and society of these information from user by social platform
Bank of Communications is middle acquisition.The application mainly predicts and recommends using artificial intelligence, and is carried out using linear programming algorithm
Path planning.The application in recommending scenery spot part, using tourist psychology basic theories to the trip behavior feature of user into
The prediction and recommendation that tourist image design is carried out to user are realized in row verification and explanation using machine learning algorithm.Use the present invention's
Method and apparatus can save the time of user, and can recommend sight spot information and road automatically for it according to the behavior of user
Diameter information, more convenient, quick and hommization.The application is in terms of path planning, by integral linear programming algorithm for use
Traveling mode of family during travelling formulates itinerary.Since artificial intelligence does not have very by force in terms of accurate calculation
Advantage, therefore the application calculates the part for needing accurate operation in conjunction with linear planning algorithm, " fixed to be truly realized
The effect of system ".Algorithm consider major parameter include：Intelligence is pre- at the sight spot that user terminal voluntarily selects or manually by user
The sight spot of the user's optimal selection measured, the type of logical tool that will be selected, the sight spot of arrival, the most possible option of user
Path, user's time requirement etc. that sight spot is watched.Path planning is carried out using object function and constraints, designs mould
Quasi- route.Again by the design of this platform, obtain feedback and the requirement of corresponding user, for example whether need to change the route,
Rationally whether arrangement of time, whether user be ready to buy sight spot admission ticket on the route or traffic ticket and go on a journey with reference to the route
Deng initial determinations.Since existing technology is in path planning, use only linear programming method, and the application be by
Artificial intelligence and path planning are combined, therefore can obtain more accurately result of calculation, to meet different type
Client needs.
By mobile Internet application in this platform, a large amount of user feedback can be obtained, then characteristic is changed with this,
Algorithm parameter is optimized.Once user determines final route and carries out tourist activity with reference to route, platform also will
It is collected by big data, such as evaluation of the user to sight spot, the restaurant on the way that other users are recommended, more accurate traffic information
Deng.
According to further aspect of the application, a kind of computer equipment, including memory, processor and storage are additionally provided
In the memory and the computer program that can be run by the processor, wherein the processor executes the computer
Any one in travel information recommendation method as described above is realized when program.
According to further aspect of the application, a kind of computer readable storage medium is additionally provided, it is preferably non-volatile
Readable storage medium storing program for executing, is stored with computer program, and the computer program is realized as described above when executed by the processor
Travel information recommendation method in any one.
According to further aspect of the application, a kind of computer program product including instruction is additionally provided.When the calculating
When machine program product is run on computers so that computer executes any one in above-mentioned travel information recommendation method.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or its arbitrary combination real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When computer loads and executes the computer program instructions, whole or portion
Ground is divided to generate according to the flow or function described in the embodiment of the present application.The computer can be all-purpose computer, dedicated computing
Machine, computer network obtain other programmable devices.The computer instruction can be stored in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or
Data center is transmitted.The computer readable storage medium can be any usable medium that computer can access or
It is comprising data storage devices such as one or more usable mediums integrated server, data centers.The usable medium can be with
It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state disk
Solid State Disk (SSD)) etc..
Professional should further appreciate that, described in conjunction with the examples disclosed in the embodiments of the present disclosure
Unit and algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, hard in order to clearly demonstrate
The interchangeability of part and software generally describes each exemplary composition and step according to function in the above description.
These functions are implemented in hardware or software actually, depend on the specific application and design constraint of technical solution.
Professional technician can use different methods to achieve the described function each specific application, but this realization
It is not considered that exceeding scope of the present application.
One of ordinary skill in the art will appreciate that implement the method for the above embodiments be can be with
It is completed come instruction processing unit by program, the program can be stored in computer readable storage medium, and the storage is situated between
Matter is non-transitory (English：Non-transitory) medium, such as random access memory, read-only memory, flash
Device, hard disk, solid state disk, tape (English：Magnetic tape), floppy disk (English：Floppy disk), CD (English：
Optical disc) and its arbitrary combination.
The preferable specific implementation mode of the above, only the application, but the protection domain of the application is not limited thereto,
Any one skilled in the art is in the technical scope that the application discloses, the change or replacement that can be readily occurred in,
It should all cover within the protection domain of the application.Therefore, the protection domain of the application should be with scope of the claims
Subject to.
Claims (10)
1. a kind of travel information recommendation method, this method include：
User information obtaining step：The first user data is obtained, first user data includes the identity information of the first user
And user history information；With
Recommendation information generation step：By the trained deep neural network of the first user data input, tourism letter is generated
Recommendation results are ceased, the travel information recommendation results characterize the probability that first user will go at least one sight spot tourism；
Wherein, deep neural network is obtained by training step, and the training step includes：
User data annotation step：Each user in gathering for user, obtains the travel history data of the user, by institute
Input data of the data as training data before being happened at time point T in travel history data is stated, after the time point T
Other sight spots except the user to first tourist attractions and the tourist attractions crossed are marked respectively, obtain practical tourism
Result data；
Scene data obtaining step：Each sight spot in gathering for sight spot, obtains scene data, and the scene data includes
The behavioral data of the essential information data at the sight spot and each user to the sight spot；
Model prediction step：By the trained number of the scene data, the identity information of each user and the user
According to being input in deep neural network, the Making by Probability Sets that each described user goes each sight spot tourism is obtained；With
Modifying model step：The practical tourism result data and Making by Probability Sets are compared, to the deep neural network
It is modified, obtains the trained deep neural network.
2. according to the method described in claim 1, it is characterized in that, the model prediction step includes：
Input step：The training data and the scene data are input to the input layer of the deep neural network；
Processing step：Each item data in each item data and the scene data in the training data is converted respectively
For corresponding characteristic；
Step of converting：Convert characteristic corresponding with the training data to user characteristics matrix, and will be with the scape
The corresponding characteristic of point data is converted into sight spot eigenmatrix；
Associated steps：The user characteristics matrix is associated with the sight spot eigenmatrix at each sight spot, calculate the user
User representated by eigenmatrix removes the probability at the sight spot；With
Export step：Calculating each user in user's set goes the probability set of each sight spot tourism to merge described in output
Making by Probability Sets.
3. according to the method described in claim 1, it is characterized in that, the Modifying model step includes：
Data type step of converting：Probability data in the Making by Probability Sets is compared with preset threshold value, it will be described general
Rate data are converted into the data of integer type；
Residual computations step：The data of the integer type are compared with the practical tourism result data, obtain residual error
Data；With
Amendment step：The deep neural network is modified by reversed neural propagation using the residual error data.
4. according to any one of them method of claims 1 to 3, which is characterized in that this method is generated in the recommendation information
Further include after step：
Sight spot preferred steps：Interval or the route planning in response to first user instruct at every predetermined time, according to
The row of at least one sight spot that the travel information recommendation results include and first user at least one sight spot
Preferred sight spot is determined for data；With
Route planning step：According to object function and constraints, path planning is carried out to the preferred sight spot, generates travelling line
Road.
5. according to the method described in claim 4, it is characterized in that, this method further includes after the route planning step：
User intent judgment step：Judge whether the itinerary meets the expection of first user, if it is, executing
Circuit determines step, if it is not, then executing sight spot annotation step；
Route determination step：Judge whether the itinerary needs to change, if necessary to change, then according to first user
Modification to the sight spot in the itinerary, again planning path；With
Sight spot markers step：According to the feedback of first user, the preferred sight spot is labeled as to the row of first user
For data, so as to as the data needed for the training step.
6. a kind of travel information recommendation apparatus, including：
User profile acquisition module is configured to obtain the first user data, and first user data includes the first user
Identity information and user history information；With
Recommendation information generation module is configured to the trained deep neural network of the first user data input, raw
At travel information recommendation results, the travel information recommendation results, which characterize first user, will go at least one sight spot tourism
Probability；
Wherein, deep neural network is obtained by training module, and the training module includes：
User data labeling module, each user being configured in gathering for user, obtains the travel history of the user
Data, will be happened at the data before time point T as the input data of training data in the travel history data, will be described
Other sight spots after time point T except the user to first tourist attractions and the tourist attractions crossed are marked respectively, obtain
To practical result data of travelling；
Scene data acquisition module is configured to each sight spot in gathering for sight spot, obtains scene data, the sight spot
Data include behavioral data of the essential information data at the sight spot with the user to the sight spot；
Model prediction module is configured to the training data and the scene data being input in deep neural network, obtain
Each user in gathering to the user goes the Making by Probability Sets of each sight spot tourism；With
Modifying model module is configured to the practical tourism result data and Making by Probability Sets being compared, to the depth
Neural network is modified, and obtains the trained deep neural network.
7. device according to claim 6, which is characterized in that the model prediction module includes：
Input module is configured to for the training data and the scene data to be input to the input of the deep neural network
Layer；
Processing module is configured to each item data in each item data and the scene data in the training data
It is separately converted to corresponding characteristic；
Conversion module is configured to convert characteristic corresponding with the training data to user characteristics matrix, and will
Characteristic corresponding with the scene data is converted into sight spot eigenmatrix；
Relating module is configured to the user characteristics matrix is associated with the sight spot eigenmatrix at each sight spot, calculating
User representated by the user characteristics matrix removes the probability at the sight spot；With
Output module is configured to calculate the probability set merging that each user in user's set goes each sight spot tourism
Export the Making by Probability Sets.
8. the device described according to claim 6 or 7, which is characterized in that the device is gone back after the recommendation information generating means
It is connected with：
Sight spot preferred module is configured to be spaced at every predetermined time or refer in response to the route planning of first user
It enables, at least one sight spot for including according to the travel information recommendation results and first user are at least one scape
The behavioral data of point determines preferred sight spot；With
Route planning module, is configured to according to object function and constraints, and path planning is carried out to the preferred sight spot, raw
At itinerary.
9. a kind of computer equipment, including memory, processor and storage can be transported in the memory and by the processor
Capable computer program, wherein the processor is realized when executing the computer program such as any one of claim 1 to 5
The method.
10. a kind of computer readable storage medium, preferably non-volatile readable storage medium, are stored with computer journey
Sequence, the computer program realize the method as described in any one of claim 1 to 5 when executed by the processor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810259835.7A CN108537373A (en) | 2018-03-27 | 2018-03-27 | Travel information recommends method and apparatus |
| PCT/CN2019/079335 WO2019184833A1 (en) | 2018-03-27 | 2019-03-22 | Tourism information recommending method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810259835.7A CN108537373A (en) | 2018-03-27 | 2018-03-27 | Travel information recommends method and apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108537373A true CN108537373A (en) | 2018-09-14 |
Family
ID=63485190
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810259835.7A Withdrawn CN108537373A (en) | 2018-03-27 | 2018-03-27 | Travel information recommends method and apparatus |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108537373A (en) |
| WO (1) | WO2019184833A1 (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109710848A (en) * | 2018-12-26 | 2019-05-03 | 巨轮智能装备股份有限公司 | A kind of recommendation method for personalized information based on Fuzzy Optimum Neural Network |
| CN109815309A (en) * | 2018-12-21 | 2019-05-28 | 航天信息股份有限公司 | A kind of user information recommended method and system based on personalization |
| CN109872065A (en) * | 2019-02-15 | 2019-06-11 | 深圳易至科技有限公司 | A kind of method of the preferred house ornamentation personnel of intelligence |
| CN109977283A (en) * | 2019-03-14 | 2019-07-05 | 中国人民大学 | A kind of the tourism recommended method and system of knowledge based map and user's footprint |
| CN109993668A (en) * | 2019-04-09 | 2019-07-09 | 桂林电子科技大学 | A kind of recommending scenery spot method based on gating cycle unit neural network |
| CN110222902A (en) * | 2019-06-13 | 2019-09-10 | 衢州学院 | Tourist attractions recommender system and paths planning method |
| CN110297964A (en) * | 2019-03-28 | 2019-10-01 | 特斯联（北京）科技有限公司 | A kind of tourist attractions recommended method and device based on big data analysis |
| WO2019184833A1 (en) * | 2018-03-27 | 2019-10-03 | 上海程向信息科技有限公司 | Tourism information recommending method and device |
| CN110321516A (en) * | 2019-07-12 | 2019-10-11 | 四川亨通网智科技有限公司 | A kind of universe tourism public service platform and its management system |
| CN110335099A (en) * | 2019-05-06 | 2019-10-15 | 北京盛威时代科技有限公司 | A kind of ticket purchase circuit recommendation method based on user's history data |
| CN111104614A (en) * | 2019-12-11 | 2020-05-05 | 上海携旅信息技术有限公司 | Method for generating recall information for tourist destination recommendation system |
| CN111489131A (en) * | 2019-01-25 | 2020-08-04 | 北京搜狗科技发展有限公司 | Information recommendation method and device |
| CN111522979A (en) * | 2020-04-20 | 2020-08-11 | 携程旅游网络技术（上海）有限公司 | Picture sorting recommendation method and device, electronic equipment and storage medium |
| CN111612590A (en) * | 2020-03-19 | 2020-09-01 | 江苏智檬智能科技有限公司 | Scenic spot recommendation method and device based on artificial intelligence big data |
| CN111695920A (en) * | 2019-03-11 | 2020-09-22 | 新疆丝路大道信息科技有限责任公司 | Tourist attraction recommendation system and method of automobile leasing platform and electronic equipment |
| CN111753182A (en) * | 2019-03-28 | 2020-10-09 | 北京达佳互联信息技术有限公司 | Multimedia information recommendation method and device, electronic equipment and readable storage medium |
| CN112200334A (en) * | 2020-10-28 | 2021-01-08 | 上海伯瑞斯健康管理发展有限公司 | Self-driving public platform |
| CN112465276A (en) * | 2021-01-28 | 2021-03-09 | 湖南惠旅云网络科技有限公司 | Method for automatically generating scenic spot playing recommended route |
| CN113724069A (en) * | 2021-08-31 | 2021-11-30 | 平安科技（深圳）有限公司 | Pricing method and device based on deep learning, electronic equipment and storage medium |
| CN114065014A (en) * | 2020-07-31 | 2022-02-18 | 北京达佳互联信息技术有限公司 | Information matching method, device, equipment and storage medium |
| CN114374854A (en) * | 2021-12-20 | 2022-04-19 | 广西壮族自治区公众信息产业有限公司 | Cloud tourism live broadcasting method and system |
| CN114543831A (en) * | 2022-04-18 | 2022-05-27 | 季华实验室 | Route planning method, device and equipment based on driving style and storage medium |
| CN115292630A (en) * | 2022-10-09 | 2022-11-04 | 深圳市明源云客电子商务有限公司 | Data processing method, device and computer readable storage medium |
| CN116433269A (en) * | 2023-06-13 | 2023-07-14 | 四川交通职业技术学院 | Method and device for charging parking lot of zone type unmanned vehicle based on big data |
| CN116457810A (en) * | 2020-08-06 | 2023-07-18 | Ipass科技股份有限公司 | Method, system, equipment and device for realizing fluency of travel program |
| CN111753182B (en) * | 2019-03-28 | 2024-05-17 | 北京达佳互联信息技术有限公司 | Multimedia information recommendation method and device, electronic equipment and readable storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103886368A (en) * | 2014-03-26 | 2014-06-25 | 南京航空航天大学 | Satellite accurate orbit prediction method |
| CN107436950B (en) * | 2017-08-07 | 2020-12-29 | 苏州大学 | Travel route recommendation method and system |
| CN108537373A (en) * | 2018-03-27 | 2018-09-14 | 黄晓鸣 | Travel information recommends method and apparatus |
-
2018
- 2018-03-27 CN CN201810259835.7A patent/CN108537373A/en not_active Withdrawn
-
2019
- 2019-03-22 WO PCT/CN2019/079335 patent/WO2019184833A1/en active Application Filing
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2019184833A1 (en) * | 2018-03-27 | 2019-10-03 | 上海程向信息科技有限公司 | Tourism information recommending method and device |
| CN109815309A (en) * | 2018-12-21 | 2019-05-28 | 航天信息股份有限公司 | A kind of user information recommended method and system based on personalization |
| CN109710848A (en) * | 2018-12-26 | 2019-05-03 | 巨轮智能装备股份有限公司 | A kind of recommendation method for personalized information based on Fuzzy Optimum Neural Network |
| CN111489131A (en) * | 2019-01-25 | 2020-08-04 | 北京搜狗科技发展有限公司 | Information recommendation method and device |
| CN109872065A (en) * | 2019-02-15 | 2019-06-11 | 深圳易至科技有限公司 | A kind of method of the preferred house ornamentation personnel of intelligence |
| CN111695920B (en) * | 2019-03-11 | 2023-06-13 | 新疆丝路大道信息科技有限责任公司 | Tourist attraction recommendation system and method of automobile leasing platform and electronic equipment |
| CN111695920A (en) * | 2019-03-11 | 2020-09-22 | 新疆丝路大道信息科技有限责任公司 | Tourist attraction recommendation system and method of automobile leasing platform and electronic equipment |
| CN109977283A (en) * | 2019-03-14 | 2019-07-05 | 中国人民大学 | A kind of the tourism recommended method and system of knowledge based map and user's footprint |
| CN109977283B (en) * | 2019-03-14 | 2021-03-05 | 中国人民大学 | Tourism recommendation method and system based on knowledge graph and user footprint |
| CN110297964A (en) * | 2019-03-28 | 2019-10-01 | 特斯联（北京）科技有限公司 | A kind of tourist attractions recommended method and device based on big data analysis |
| CN111753182B (en) * | 2019-03-28 | 2024-05-17 | 北京达佳互联信息技术有限公司 | Multimedia information recommendation method and device, electronic equipment and readable storage medium |
| CN111753182A (en) * | 2019-03-28 | 2020-10-09 | 北京达佳互联信息技术有限公司 | Multimedia information recommendation method and device, electronic equipment and readable storage medium |
| CN109993668A (en) * | 2019-04-09 | 2019-07-09 | 桂林电子科技大学 | A kind of recommending scenery spot method based on gating cycle unit neural network |
| CN109993668B (en) * | 2019-04-09 | 2021-07-13 | 桂林电子科技大学 | Scenic spot recommendation method based on gated cyclic unit neural network |
| CN110335099A (en) * | 2019-05-06 | 2019-10-15 | 北京盛威时代科技有限公司 | A kind of ticket purchase circuit recommendation method based on user's history data |
| CN110335099B (en) * | 2019-05-06 | 2021-01-01 | 盛威时代科技集团有限公司 | Method for recommending ticket purchase line based on user historical data |
| CN110222902A (en) * | 2019-06-13 | 2019-09-10 | 衢州学院 | Tourist attractions recommender system and paths planning method |
| CN110321516A (en) * | 2019-07-12 | 2019-10-11 | 四川亨通网智科技有限公司 | A kind of universe tourism public service platform and its management system |
| CN110321516B (en) * | 2019-07-12 | 2022-01-28 | 四川亨通网智科技有限公司 | Universal tourism public service platform and management system thereof |
| CN111104614A (en) * | 2019-12-11 | 2020-05-05 | 上海携旅信息技术有限公司 | Method for generating recall information for tourist destination recommendation system |
| CN111612590A (en) * | 2020-03-19 | 2020-09-01 | 江苏智檬智能科技有限公司 | Scenic spot recommendation method and device based on artificial intelligence big data |
| CN111522979A (en) * | 2020-04-20 | 2020-08-11 | 携程旅游网络技术（上海）有限公司 | Picture sorting recommendation method and device, electronic equipment and storage medium |
| CN111522979B (en) * | 2020-04-20 | 2023-09-29 | 携程旅游网络技术（上海）有限公司 | Picture sorting recommendation method and device, electronic equipment and storage medium |
| CN114065014A (en) * | 2020-07-31 | 2022-02-18 | 北京达佳互联信息技术有限公司 | Information matching method, device, equipment and storage medium |
| CN116457810A (en) * | 2020-08-06 | 2023-07-18 | Ipass科技股份有限公司 | Method, system, equipment and device for realizing fluency of travel program |
| CN112200334A (en) * | 2020-10-28 | 2021-01-08 | 上海伯瑞斯健康管理发展有限公司 | Self-driving public platform |
| CN112465276A (en) * | 2021-01-28 | 2021-03-09 | 湖南惠旅云网络科技有限公司 | Method for automatically generating scenic spot playing recommended route |
| CN113724069A (en) * | 2021-08-31 | 2021-11-30 | 平安科技（深圳）有限公司 | Pricing method and device based on deep learning, electronic equipment and storage medium |
| CN113724069B (en) * | 2021-08-31 | 2024-02-13 | 平安科技（深圳）有限公司 | Deep learning-based pricing method, device, electronic equipment and storage medium |
| CN114374854A (en) * | 2021-12-20 | 2022-04-19 | 广西壮族自治区公众信息产业有限公司 | Cloud tourism live broadcasting method and system |
| CN114543831A (en) * | 2022-04-18 | 2022-05-27 | 季华实验室 | Route planning method, device and equipment based on driving style and storage medium |
| CN114543831B (en) * | 2022-04-18 | 2022-10-18 | 季华实验室 | Route planning method, device and equipment based on driving style and storage medium |
| CN115292630A (en) * | 2022-10-09 | 2022-11-04 | 深圳市明源云客电子商务有限公司 | Data processing method, device and computer readable storage medium |
| CN116433269A (en) * | 2023-06-13 | 2023-07-14 | 四川交通职业技术学院 | Method and device for charging parking lot of zone type unmanned vehicle based on big data |
| CN116433269B (en) * | 2023-06-13 | 2023-08-18 | 四川交通职业技术学院 | Method and device for charging parking lot of zone type unmanned vehicle based on big data |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2019184833A1 (en) | 2019-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108537373A (en) | Travel information recommends method and apparatus | |
| Jin et al. | Using user-generated content to explore the temporal heterogeneity in tourist mobility | |
| CN109977283B (en) | Tourism recommendation method and system based on knowledge graph and user footprint | |
| CN110276008A (en) | A kind of recommending scenery spot method and device based on user's travelling decision process | |
| Long et al. | Moving ahead with computational movement analysis | |
| CN113139140B (en) | Tourist attraction recommendation method based on space-time perception GRU and combined with user relationship preference | |
| KR101094484B1 (en) | Tour information guide system using the recommendation algorithm by experts and method thereof | |
| MirHassani et al. | Quantum binary particle swarm optimization-based algorithm for solving a class of bi-level competitive facility location problems | |
| Chen et al. | Multi-objective reinforcement learning approach for trip recommendation | |
| Ectors et al. | Developing an optimised activity type annotation method based on classification accuracy and entropy indices | |
| Yin et al. | ADPR: An attention-based deep learning point-of-interest recommendation framework | |
| KR102592129B1 (en) | system for recommending clothing based on artificial intelligence | |
| Liu | Rank-based ant colony optimization applied to dynamic traveling salesman problems | |
| Qian et al. | Vehicle trajectory modelling with consideration of distant neighbouring dependencies for destination prediction | |
| Zhang | RETRACTED ARTICLE: Artificial intelligence assisted cyber threat assessment and applications for the tourism industry | |
| Chen et al. | An efficient and practical approach to obtain a better optimum solution for structural optimization | |
| KR20210047147A (en) | Artificial intelligence based hotel recommendation device and method | |
| Wang et al. | etour: A two-layer framework for tour recommendation with super-pois | |
| CN115952355A (en) | Next interest point recommendation system based on space-time power law attention | |
| CN114417166A (en) | Continuous interest point recommendation method based on behavior sequence and dynamic social influence | |
| Mathias et al. | Personalized sightseeing tours: a model for visits in art museums | |
| Turktarhan et al. | Big Data and Business Intelligence in Hospitality and Tourism | |
| Liu et al. | Data analysis of the development status of Basketball National Fitness based on FOG Computing | |
| Natapov et al. | Visuospatial search in urban environment simulated by random walks | |
| Shao et al. | A Combinatorial optimization framework for scoring students in University Admissions |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20190211 Address after: Room 501-3, North Building, No. 1 Building, 951 Jianchuan Road, Minhang District, Shanghai, 200240 Applicant after: Shanghai Chengxiang Information Technology Co., Ltd. Address before: 350000 Yili Mingju, 221 Xiufeng Road, Fuzhou City, Fujian Province, 2 buildings, 1006 units Applicant before: Huang Xiaoming |
|
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180914 |