CN105608153A - Universal POI information association method - Google Patents
Universal POI information association method Download PDFInfo
- Publication number
- CN105608153A CN105608153A CN201510957892.9A CN201510957892A CN105608153A CN 105608153 A CN105608153 A CN 105608153A CN 201510957892 A CN201510957892 A CN 201510957892A CN 105608153 A CN105608153 A CN 105608153A
- Authority
- CN
- China
- Prior art keywords
- data
- user
- poi
- longitude
- latitude
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2216/00—Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
- G06F2216/03—Data mining
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention belongs to the technical field of internet advertisements and provides a universal POI (Point Of Interest) information association method. The method comprises the steps of extracting longitude and latitude data in user information; cleaning the longitude and latitude data; supplementing the longitude and latitude data with historical data of a network access device to realize the expansion of the cleaned longitude and latitude data of a user; associating POI by adopting a projection formula based method and dynamically establishing a plane map index by using map planarization; comparing an index corresponding to a building in a map with an index generated by a POI database, and when a comparison result is that the index corresponding to the building in the map is consistent with the index generated by the POI database, associating the building in the map to GPS coordinates of the user and the POI database; and generating different POI information of the user. According to the method, the extracted longitude and latitude data is cleaned and expanded, so that the workload is reduced, the possibility of missing POI data is reduced, the POI of each user can be calculated in mass data reasonably and quickly, and current POI attribute and behavior data of the user can be judged.
Description
Technical field
The invention belongs to internet advertisement technology field, particularly a kind of general POI information association method.
Background technology
Internet advertising, because its interactive strong, advantage that accuracy is high proportion in advertisement structure rises year by year, andThere is the trend that continues expansion. 2011 to 2014, the market scale of Internet advertising exceeded newspaper advertisement scale,At the second place, market scale keeps rapid growth continuously.
That GPS is that one has is comprehensive, round-the-clock, all the period of time, high-precision satellite navigation system, can provide for userThe navigation informations such as low cost, high accuracy three-dimensional position, speed and accurate timing. And at present, answering in equipment for surfing the netAll can obtain from backstage user's positional information by program, for example: user's navigator used in everyday, they helpUser traverses into destination locations from starting point, and in whole process, what navigation application all can continue obtains user'sPositional information. In the whole process of therefore travelling user, different POI is along with occurring together, and POI is PointThe abbreviation of OfInterest, represents the interest of user on current this aspect. If we have a user's one dayAll POI data, the behavior track of cooking up this day of this user that we just can be general, even stops at each pointThe time of staying.
The point of POI has retractility for ten minutes, can be a house, an office building, a mailbox, a public transportStand, also can arrive a park, an airport etc. User can produce corresponding on any one physical locationPOI, can produce relevant intention, once it is intentional on certain point to know user, just can recommend to user,There is great business opportunity.
Current recommendation service is of a great variety, and people can pass through the various recommendation informations of Network Capture, recommends books, soundHappy, film, commodity etc., are POI information based on specific user but do not have a kind of, and recommend, lackWeary specific aim, applicability is not strong.
POI, can also be according to user POI in history except contributing to the recommendation service of user on current pointData, carry out some features of inference and recording user, and then are formed with recommendation targetedly. For example: if a useFamily appears at a workplace for a long time in the working time, and be probably just user's job site here, and asOn one's own time and often there will be the POI of a residential quarter class weekend, here probably for user of fruitBe user's place of abode, carry out association by some POI data and calculate, can pass through user's historical POI numberAccording to drawing corresponding knowledge.
Further, the quantity of POI and the transition of real world are closely related. In real world, when a new battalionWhen industry place is created, new POI along with and produced, and along with a large amount of generations, the renewal of POI, user's rowAlso can be along with upgrading together for attribute, thing followed problem is exactly, when getting more and more of data volume accumulation, to userPOI judgement and the difficulty of calculating can increase gradually. In addition, more data also can be brought more noise, how to closeThe POI that manages, calculates fast each user in mass data, being that internet advertisement technology field is to be solved asksTopic.
Therefore, a kind of general POI information association method is badly in need of in internet advertisement technology field, adopts POI system to carryingThe longitude and latitude data of taking out are cleaned, and reduce workload, can be rationally, in mass data, calculate fastEach user's POI, and judge user current POI attribute data and behavioral data.
Summary of the invention
The invention provides a kind of general POI information association method, technical scheme is as follows:
A general POI information association method, comprises the steps:
The first step, extracts the longitude and latitude data in user profile;
Second step, cleans the longitude and latitude data that extract in the first step;
The 3rd step, is used the historical data of equipment for surfing the net for not extracting the URL of longitude and latitude data in the first step, supplements warpLatitude data, realize the expansion of user's longitude and latitude data after cleaning;
The 4th step, adopts the associated POI of method based on projection formula, uses map plane Dynamic Establishing plane map ropeDraw;
The 5th step, compares the corresponding index of building in map in the 4th step and the index that POI database produces, in the time that comparative result is consistent, this building in map is associated with to user's gps coordinate and POI databaseIn;
The 6th step, generates the different POI information of user, comprises UAD and behavioral data.
Preferably, in a kind of above-mentioned general POI information association method, in the first step, extract the longitude and latitude in user profileThe concrete steps of degrees of data are:
First, in task start, load known extracting rule file, form key-value distributed storageThe key-value pair of system, is loaded in internal memory;
Further, in the data processing stage, scan original data on flows according to the established rule loading, from original flowExtracting data goes out and user-dependent positional information, for example: in original data on flows, the letter relevant to customer locationBreath is generally all stored in the URL of data on flows, and URL is the abbreviation of UniformResourceLocator, represents systemOne URLs, according to the server name in URL, looks for extracting rule corresponding in internal memory, uses and just findsExpression formula is carried out the extraction of longitude and latitude data from URL;
Further, after extraction, each user can form one according to the set rule of POI system and canonical tableReach user's latitude and longitude coordinates tables of data that formula extracts.
Preferably, in a kind of above-mentioned general POI information association method, in second step, use multiple resource files and systemThe longitude and latitude data that 2 kinds of modes of meter excavation extract the first step are respectively cleaned.
Preferably, in a kind of above-mentioned general POI information association method, in second step, use multiple resource files and systemThe concrete steps that the longitude and latitude data that 2 kinds of modes of meter excavation extract the first step are respectively cleaned are:
In the time adopting multiple resource files to clean the longitude and latitude data of extracting, because resource table has 2, be firstThe blacklist information of coordinate points, finds to exist a large amount of repetition longitude and latitude data in data by frequency analysis, andThe longitude and latitude data correlation repeating is to different users, because longitude and latitude data have all been accurate to four after decimal point to fiveMore than position, the registration of data itself is very little, but is all present in the data of extraction, draws by analysis this repetitionLongitude and latitude data are the down town points in each city, and the reason that occurs these points is the application software when a mobile terminalCan not obtain active user's positional information time, can be using the down town point at active user place as active user'sLatitude and longitude information, so class data are wrong; Therefore, this type of longitude and latitude data rows is entered to blacklist, complete to fromIn flow, extract the preliminary filtration of longitude and latitude data; The region, province that user belongs to is simultaneously fixed, if userLongitude and latitude data do not belong in this province, these longitude and latitude data are piped off;
In the time adopting statistics mining algorithm to clean the longitude and latitude data of extracting, first open equipment for surfing the net application,There will be the inaccurate phenomenon in location, this type of is located inaccurate data and has also been blended in data on flows; For onlineThe data on flows that equipment produces, this type of is located inaccurate data and can be eliminated, because equipment for surfing the net is notBecome, the positional information that user sends by equipment for surfing the net should be also constant or mobility scale is very little; Work as numberWhen being accumulated to a magnitude, service range formula, can calculate the distance between each anchor point, when occurring oneWhen the distance of individual point and other points is all greater than distance each other of other points, this point is just listed in abnormity point, because ofThis, delete this type of longitude and latitude data, and the longitude and latitude data that extract from flow are done further and filtered.
Preferably, in a kind of above-mentioned general POI information association method, in the 3rd step, use the historical number of equipment for surfing the netAccording to the URL for not extracting longitude and latitude data in the first step, supplement longitude and latitude data, realize and clean rear user's longitude and latitude dataThe concrete steps of expansion be:
In the time that user uses equipment for surfing the net online at every turn, all can on corresponding equipment for surfing the net, send URL, when extracting on thisWhen the longitude and latitude of net equipment, can preserve the longitude and latitude data of this equipment for surfing the net; Need to extract this online when some dayWhen the longitude and latitude of equipment, even can not extract longitude and latitude data from the URL of this online, but by going through of preservingHistory longitude and latitude data, still can draw and the longitude and latitude data of this equipment for surfing the net realize the expansion of user's longitude and latitude dataExhibition.
Preferably, in a kind of above-mentioned general POI information association method, in the 4th step, adopt the side based on projection formulaThe associated POI of method, uses the concrete steps of map plane Dynamic Establishing plane map index to be:
The map of world map, national map and each provinces and cities is all known, can know clearly each buildingInformation and geographical position; First, the earth, along development of a sphere, is become to a plane, and can obtain accuratelyThe longitude and latitude of each building on map, use projection formula in a plane, all can any one building projectionProduce an index, index is by (x, y) expression, and concrete projection formula belongs to prior art, refers in wikipediaAbout the knowledge of map projection, can obtain being positioned at the grid (x, y) in plane; Be the index of this building.
Preferably, in a kind of above-mentioned general POI information association method, in the 5th step, this building in map is closedThe concrete steps that are linked in user's gps coordinate and POI database are:
By the projection rectangle size of processing POI database be arranged to the 4th step in process the rectangle size of building in mapUnanimously, according to the method for building up of the 4th step midplane map reference, user's url data from be stored in POI databaseIn the longitude and latitude data that extract also can produce the index of (a, b), when the index producing in POI database andWhen known plane map index (x, y) is consistent in four steps, can think that this building is exactly this user's POI numberAccording to, i.e. position.
Preferably, in a kind of above-mentioned general POI information association method, in the 6th step, generate the POI letter that user is differentBreath, comprises that the concrete steps of UAD and behavioral data are:
The POI data of each user on diverse location are drawn by the 5th step, because user is mobile, so useFamily can produce many different POI data in one day, and the main Types of POI data is divided into Static and dynamic, and static state refers to useThe POI attribute data of working region, family or home area; Dynamically refer to the different location of process in user's a period of timePOI behavioral data;
(1) generate user POI attribute data according to user's historical behavior data:
Before calculating user POI attribute data, need advanced line number according to one's analysis, and then judge operation interval and non-workDo interval;
First, define multiple working times and non-working time, if user appears within the working time of current definitionClassification is on POI position, workplace, adds 1 so in the differential counting of this user job, contrary, in the time of inoperativeBetween appear on the POI position, residential quarter of family classification, the differential counting of this subscriber household adds 1; Last each user's meetingForm a statistic record, within the predefined time to this user the counting respectively under work and family sight;Use all users' count information, result is signed on a scatter diagram, transverse axis represents operative scenario counting, longitudinal axis generationTable family scene counting, the reduced time, interval classification dropped on anchor point on the position near the longitudinal axis, transverse axis; Finally selectOne to be positioned at data on the longitudinal axis and transverse axis maximum, and in the most clearly district of data boundary line delimitation in the longitudinal axis and transverse axis regionBetween, as the working time, the non-working time is in like manner; After judging working time and non-working time, if user existsIn working time, appear on certain POI position, the work POI data that this POI position is exactly this user so, asFruit is within the working time, and a user has appeared at multiple work POI position, selects so user's time of occurrence maximumPOI position is as the unique work POI of user, and the POI of family determines in like manner;
(2) user's current behavior data generate POI behavioral data:
Due in the 5th step, for each known map physical location, taking physical location as the center of circle, be defined in certainThe user who occurs in radius region is defined as and appears on this map physical location; When user is in apart from this buildingThing radius is considered this user's POI behavioral data during with interior region, and is exceeding this radius region, evenBe projected on same grid, can not serve as this user's POI behavioral data.
Beneficial effect of the present invention:
1, the present invention adopts multiple resource file modes to clean the longitude and latitude data of extracting, when occurring with each cityThe down town point in city is during as active user's latitude and longitude information, owing to having occurred a large amount of users on this aspect, suchData are inaccurate, therefore filter out such data, have not only protected the reasonability of business datum, after also having avoidedThe data long-tail phenomenon occurring when continuous data processing.
2, the present invention adopts statistics excavation mode to clean the longitude and latitude data of extracting, owing to there being a large amount of POI numbersCalculate according to needs, therefore for us, what face is not 1 or 2 positional information, neither tens onHundred positional information, but need to process 1,000,000, other position data of millions, and quantity is also progressively increasingLong, therefore, doing in map datum projection process, need to process the plane map after projection, when data accumulative totalDuring to a magnitude, service range formula, can calculate the distance between each point, when occurring point and otherDistance while being all greater than distance each other of other somes, this point, just as abnormity point processing, is deleted, to fromThe longitude and latitude data that extract in flow are done further and are filtered, and reduce workload.
3, the present invention adopts multiple resource files and statistics mining algorithm to clean the longitude and latitude data that extract,Reduce workload, can be rationally, in mass data, calculate fast each user's POI, and judge userAttribute data and behavioral data.
Brief description of the drawings
Describe the present invention in detail below in conjunction with the drawings and specific embodiments:
Fig. 1 is a kind of flow chart of general POI information association method.
Detailed description of the invention
For measure, creation characteristic that the technology of the present invention is realized, reach object and effect is easy to understand, general belowIn conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, aobviousSo, described embodiment is only the present invention's part embodiment, instead of whole embodiment. In the present inventionEmbodiment, those of ordinary skill in the art are not making the every other enforcement obtaining under creative work prerequisiteExample, all belongs to the scope of protection of the invention.
Embodiment 1:
Fig. 1 is a kind of flow chart of general POI information association method.
As shown in Figure 1, a kind of general POI information association method, comprises the steps:
The first step, extracts the longitude and latitude data in user profile;
Second step, cleans the longitude and latitude data that extract in the first step;
The 3rd step, is used the historical data of equipment for surfing the net for not extracting the URL of longitude and latitude data in the first step, supplements warpLatitude data, realize the expansion of user's longitude and latitude data after cleaning;
The 4th step, adopts the associated POI of method based on projection formula, uses map plane Dynamic Establishing plane map ropeDraw;
The 5th step, compares the corresponding index of building in map in the 4th step and the index that POI database produces, in the time that comparative result is consistent, this building in map is associated with to user's gps coordinate and POI databaseIn;
The 6th step, generates the different POI information of user, comprises UAD and behavioral data.
Embodiment 2:
As shown in Figure 1, a kind of general POI information association method, comprises the steps:
The first step, extracts the longitude and latitude data in user profile;
First, in task start, load known extracting rule file, form key-value distributed storageThe key-value pair of system, is loaded in internal memory;
Further, in the data processing stage, scan original data on flows according to the established rule loading, from original flowExtracting data goes out and user-dependent positional information, for example: in original data on flows, the letter relevant to customer locationBreath is generally all stored in the URL of data on flows, and URL is the abbreviation of UniformResourceLocator, represents systemOne URLs, according to the server name in URL, looks for extracting rule corresponding in internal memory, uses and just findsExpression formula is carried out the extraction of longitude and latitude data from URL;
Further, after extraction, each user can form one according to the set rule of POI system and canonical tableReach user's latitude and longitude coordinates tables of data that formula extracts;
Second step, cleans the longitude and latitude data that extract in the first step;
For the data of magnanimity, owing to being mingled with some noises in data, the data that therefore have not are truly available; Even if used the extracting rule in the first step to carry out Preliminary screening to data, regular expression is also through inspectionTest, but the quality of initial data can not ensure it must is satisfactory, so for the data of extracting, needThe conversion work of executing data cleaning and coordinate-system further;
In the time adopting multiple resource files to clean the longitude and latitude data of extracting, because resource table has 2, be firstThe blacklist information of coordinate points, finds to exist a large amount of repetition longitude and latitude data in data by frequency analysis, andThe longitude and latitude data correlation repeating is to different users, because longitude and latitude data have all been accurate to four after decimal point to fiveMore than position, the registration of data itself is very little, but is all present in the data of extraction, draws by analysis this repetitionLongitude and latitude data are the down town points in each city, and the reason that occurs these points is the application software when a mobile terminalCan not obtain active user's positional information time, can be using the down town point at active user place as active user'sLatitude and longitude information, so class data are wrong; Therefore, this type of longitude and latitude data rows is entered to blacklist, complete to fromIn flow, extract the preliminary filtration of longitude and latitude data; The region, province that user belongs to is simultaneously fixed, if userLongitude and latitude data do not belong in this province, these longitude and latitude data are piped off;
In the time adopting statistics mining algorithm to clean the longitude and latitude data of extracting, first open equipment for surfing the net application,There will be the inaccurate phenomenon in location, this type of is located inaccurate data and has also been blended in data on flows; For onlineThe data on flows that equipment produces, this type of is located inaccurate data and can be eliminated, because equipment for surfing the net is notBecome, the positional information that user sends by equipment for surfing the net should be also constant or mobility scale is very little; Work as numberWhen being accumulated to a magnitude, service range formula, can calculate the distance between each anchor point, when occurring oneWhen the distance of individual point and other points is all greater than distance each other of other points, this point is just listed in abnormity point, because ofThis, delete this type of longitude and latitude data, and the longitude and latitude data that extract from flow are done further and filtered.
The 3rd step, is used the historical data of equipment for surfing the net for not extracting the URL of longitude and latitude data in the first step, supplements warpLatitude data, realize the expansion of user's longitude and latitude data after cleaning;
In the time that user uses equipment for surfing the net online at every turn, all can on corresponding equipment for surfing the net, send URL, when extracting on thisWhen the longitude and latitude of net equipment, can preserve the longitude and latitude data of this equipment for surfing the net; Need to extract this online when some dayWhen the longitude and latitude of equipment, even can not extract longitude and latitude data from the URL of this online, but by going through of preservingHistory longitude and latitude data, still can draw and the longitude and latitude data of this equipment for surfing the net realize the expansion of user's longitude and latitude dataExhibition;
The 4th step, adopts the associated POI of method based on projection formula, uses map plane Dynamic Establishing plane map ropeDraw;
The map of world map, national map and each provinces and cities is all known, can know clearly each buildingInformation and geographical position; First, the earth, along development of a sphere, is become to a plane, and can obtain accuratelyThe longitude and latitude of each building on map, use projection formula in a plane, all can any one building projectionProduce an index, index is by (x, y) expression, and concrete projection formula belongs to prior art, refers in wikipediaAbout the knowledge of map projection, can obtain being positioned at the grid (x, y) in plane; Be the index of this building;
Owing to there being a large amount of POI data to calculate, owing to being a rectangle after map planeization, by this squareShape is divided into the multiple little rectangle of formed objects, and the POI information therefore covering in each little rectangle has just tailed off; Therefore rectangleDivision principle be: want to generate by changing the size of rectangle after projection in projection formula, can changing dynamicallyRectangle number, thereby can, according to different demands, dynamically divide square, accomplish the flexible operating to data;
The 5th step, compares the corresponding index of building in map in the 4th step and the index that POI database produces, in the time that comparative result is consistent, this building in map is associated with to user's gps coordinate and POI databaseIn;
By the projection rectangle size of processing POI database be arranged to the 4th step in process the rectangle size of building in mapUnanimously, according to the method for building up of the 4th step midplane map reference, user's url data from be stored in POI databaseIn the longitude and latitude data that extract also can produce the index of (a, b), when the index producing in POI database andWhen known plane map index (x, y) is consistent in four steps, can think that this building is exactly this user's POI numberAccording to, i.e. position;
The 6th step, generates the different POI information of user, comprises UAD and behavioral data;
The POI data of each user on diverse location are drawn by the 5th step, because user is mobile, so useFamily can produce many different POI data in one day, and the main Types of POI data is divided into Static and dynamic, and static state refers to useThe POI attribute data of working region, family or home area; Dynamically refer to the different location of process in user's a period of timePOI behavioral data;
(1) generate user POI attribute data according to user's historical behavior data:
Before calculating user POI attribute data, need advanced line number according to one's analysis, and then judge operation interval and non-workDo interval;
First, define multiple working times and non-working time, if user appears within the working time of current definitionClassification is on POI position, workplace, adds 1 so in the differential counting of this user job, contrary, in the time of inoperativeBetween appear on the POI position, residential quarter of family classification, the differential counting of this subscriber household adds 1; Last each user's meetingForm a statistic record, within the predefined time to this user the counting respectively under work and family sight;Use all users' count information, result is signed on a scatter diagram, transverse axis represents operative scenario counting, longitudinal axis generationTable family scene counting, the reduced time, interval classification dropped on anchor point on the position near the longitudinal axis, transverse axis; Finally selectOne to be positioned at data on the longitudinal axis and transverse axis maximum, and in the most clearly district of data boundary line delimitation in the longitudinal axis and transverse axis regionBetween, as the working time, the non-working time is in like manner; After judging working time and non-working time, if user existsIn working time, appear on certain POI position, the work POI data that this POI position is exactly this user so, asFruit is within the working time, and a user has appeared at multiple work POI position, selects so user's time of occurrence maximumPOI position is as the unique work POI of user, and the POI of family determines in like manner;
(2) user's current behavior data generate POI behavioral data:
Due in the 5th step, for each known map physical location, taking physical location as the center of circle, be defined in certainThe user who occurs in radius region is defined as and appears on this map physical location; When user is in apart from this buildingThing radius is considered this user's POI behavioral data during with interior region, and is exceeding this radius region, evenBe projected on same grid, can not serve as this user's POI behavioral data.
Below in conjunction with concrete URL and longitude and latitude data, the present invention is made an explanation:
The first step, extracts the longitude and latitude data in data;
Each user or equipment for surfing the net can correspond to specific longitude and latitude data, for example: url data ishttp:/diditaxi.com/mylocation.html？lat＝31.123412&lon＝121.231241；
Extract longitude and latitude for this url data, the result of extraction is: latitude=31.123412, longitude=121.231241;
Second step, the longitude and latitude data that the first step is extracted are cleaned;
(1) for example, in the time using the mode of multiple resource files to clean: resource file 1: from historical data,We can obtain some longitude and latitude positions, and the number of times that they occur is very many, because this information of longitude and latitude is to be accurate to rice, general registration can be too not high, if there is the point that the frequency abnormality of this appearance is large, must be problematic, with frequentlyInferior calculating, has produced a blacklist using each down town point longitude and latitude data as improper value, for example, in PekineseHeart point position: latitude 116.397428, longitude 39.90923, Shanghai center position:121.472607,31.232856; As long as user drops on such some position, down town, the longitude and latitude number of degrees at this user placeAccording to being cleaned.
Resource file 2: due to the corresponding specific longitude and latitude scope of each province; Because extract from the first stepLongitude and latitude data are divided economizing, and therefore, whether the longitude and latitude data that extract by judgement are positioned at the model of this provinceIn enclosing, and then which longitude and latitude data out of judging extraction are that reasonably which is irrational, for unreasonableData added blacklist, wash.
(2) in the time using the mode of statistics mining algorithm to clean, equipment for surfing the net, in the time obtaining positional information, can produceCertain deviation, first sets the error range threshold value allowing, and thinks the position letter obtaining in the time exceeding error range threshold valueBreath is mistake, and for example: 1,2,3,4 anchor points are all centered around together, 5 of anchor points do not have, by calculation level withDistance between point, point 5 and the distance value of other 4 points are all greater than other distance values between putting, and think anchor point 5Wrong, by anchor point 5 filterings. Utilize this kind of method can know a far point also can know 2 above far awayPoint, determines according to the distance value situation of Practical Calculation;
The 3rd step, is used the historical data of equipment for surfing the net for not extracting the URL of longitude and latitude data in the first step, supplements longitude and latitudeDegrees of data, realizes the expansion of user's longitude and latitude data;
In the time that user uses equipment for surfing the net online, can on corresponding equipment for surfing the net, send URL; For example: user is in and passes throughEquipment for surfing the net networking chauffeur, can produce a record, and content comprises, equipment for surfing the net name is called asd1243124asd, andThe URL:http:/diditaxi.com/mylocation.html producing? lat=31.123412&lon=121.231241,And then to extract the longitude and latitude that this family can draw equipment for surfing the net be 121.231241,31.123412; And this online is establishedStandby longitude and latitude data are preserved. Open certain webpage when some day by equipment for surfing the net, the page of for example Baidu,Can produce a record, content comprises, equipment for surfing the net name is called asd1243124asd, and the URL producing:Https:/baidu.com; Although do not extract positional information in internet records, by the historical longitude and latitude number of degrees of preservingAccording to, still can know and the longitude and latitude data of this family equipment for surfing the net realize the expansion of user's longitude and latitude data;
The 4th step, is used projection formula by map plane, and Dynamic Establishing plane map index;
The map of world map, national map and each provinces and cities is all known, can know clearly each buildingInformation and geographical position; Can obtain accurately the longitude and latitude of each building, wherein any one building is thrownIn shadow to plane, all can produce an index, index is by (x, y) expression simultaneously, and concrete projection formula belongs toPrior art, refers to the knowledge about map projection in wikipedia, can obtain being positioned at a grid in plane (x,Y); Be the index of this building;
The 5th step, based on plane map index, association goes out user's gps coordinate and POI data-base content;
According to the method for building up of the 4th step midplane index, be stored in the url data of user in POI database, extractLongitude and latitude data also can produce the index of (a, b), when in the index producing in POI database and the 4th stepWhen the plane map index known is consistent, can think that this building is exactly this user's POI data, i.e. position.The user's who for example extracts by URL longitude and latitude data are 121.231241,31.123412, and the index of generation is(77,91), and the longitude and latitude of building She Shan Silver Lake villa in POI database is 121.238486,31.119540 the index of generation is also (77,91), index is consistent, can determine that user's POI data are SheSilver Lake, mountain villa;
The 6th step, generates the different POI information of user;
The POI data of each user on diverse location are drawn by the 5th step, because user is mobile, so useFamily can produce many different POI data in one day, and the main Types of POI data is divided into Static and dynamic, and static state refers to useFamily POI attribute data, the POI that for example works, the POI of family; Dynamically refer to POI behavioral data, i.e. user one day processDifferent location;
(1) generate user POI attribute data according to user's historical behavior data:
Before calculating user POI attribute data, need advanced line number according to one's analysis, and then judge operation interval and non-workDo interval;
First will be divided into multiple combinations the time: combination 1 is the working time: 7～18 points, other times are non-working time,Combination 2 is the working time: 8～19 points, and other times are the combinations such as non-working time, for each combination, ifIt is workplace that user appears at classification in the working time of our current definition, for example, on the POI position of office building, thatDifferential counting at this user job adds 1, contrary, appears on one's own time the uptown of family's classificationOn POI position, the differential counting of this subscriber household adds 1; Last each user can form a statistic record, sevenThe statistics of this user under two kinds of scenes in it, field is user name, and meter respectively under work and family sightNumber; For example: user A, 9,3; User B, 3; 1. Use all users' count information, result is signed in to one and fall apartOn point diagram, transverse axis represents operative scenario counting, and the longitudinal axis represents family's scene counting, and the reduced time, interval classification can makeMore anchor point drops on the position near the longitudinal axis, transverse axis. Finally select one by this boundary line delimitation the most clearlyInterval,, the working time in current this time interval classification is exactly the working time that we calculate use, the non-working timeIn like manner. After obtaining rational time division, for the user POI data that obtain every day, if this user is beforeIn the working time of definition, appear on certain POI position, this POI is exactly this user job POI data so,If within a period of time of calculating, having there are multiple work POI in a user, selects so the conduct that time of occurrence is maximumThe work POI that user is unique, the deduction of the POI of family in like manner. For example check 30 days historical datas, user A is in the time of workBetween appear in office building A 3 hours in section, in office building B 1 hour, in office building C 1 hour, therefore from historyData it seems, this user's work POI is exactly office building A;
(2) user's current behavior data generate POI behavioral data:
Due in the 5th step, for each known map physical location, taking physical location as the center of circle, be defined in certainThe user who occurs in region in radius is defined as and appears on this physical location; When a user is from this building halfFootpath is considered this user's POI behavioral data during with interior distance, and is exceeding this radius region, even if be projected toOn same grid, can not serve as this user's POI behavioral data. For example: a user's longitude and latitude position is121.231241,31.123412, use projection formula, this user corresponding physics building on this position is She ShanyinPoly-two positions in lake villa and Shanghai, for She Shan Silver Lake villa, operating radius is 30 meters, and Shanghai is poly-Operating radius be 10 meters, this user's position is in the poly-poly-sphere of action in Shanghai, and not at She Shan Silver Lake villaScope in, therefore the current POI behavioral data of user is that Shanghai is poly-poly-.
The present invention adopts multiple resource file modes to clean the longitude and latitude data of extracting, when occurring with each cityDown town point is during as active user's latitude and longitude information, owing to having occurred a large amount of users on this aspect, such dataBe inaccurate, therefore filter out such data, not only protected the reasonability of business datum, also avoided follow-up numberThe data long-tail phenomenon occurring during according to processing.
The present invention adopts statistics excavation mode to clean the longitude and latitude data of extracting, owing to there being a large amount of POI data to needCalculate, therefore for us, what face is not 1 or 2 positional information, neither tens up to a hundredPositional information, but need to process 1,000,000, other position data of millions, and quantity also progressively increasing,Therefore doing in map datum projection process, need to process the plane map after projection, when data are accumulated to oneWhen individual magnitude, service range formula, can calculate the distance between each point, when occurring a point and other distanceWhen being all greater than distance each other of other somes, this point, just as abnormity point processing, is deleted, to from flowIn the longitude and latitude data that extract do further and filter, reduce workload.
The present invention adopts multiple resource files and statistics mining algorithm to clean the longitude and latitude data that extract, and reducesWorkload, can be rationally, in mass data, calculate fast each user's POI, and judge user propertyData and behavioral data.
More than show and described general principle of the present invention, principal character and advantage of the present invention. The technology people of the industryMember should understand, and the present invention is not restricted to the described embodiments, and the just explanation of describing in above-described embodiment and description originallyThe principle of invention, the present invention also has various changes and modifications without departing from the spirit and scope of the present invention, theseChanges and improvements all fall in the claimed scope of the invention. The claimed scope of the present invention is by appending claimsAnd equivalent defines.
Claims (8)
1. a general POI information association method, is characterized in that, comprises the steps:
The first step, extracts the longitude and latitude data in user profile;
Second step, cleans the longitude and latitude data that extract in the described first step;
The 3rd step, using the historical data of equipment for surfing the net is the URL that does not extract longitude and latitude data in the described first step, supplements longitude and latitude data, realizes the expansion of user's longitude and latitude data after cleaning;
The 4th step, adopts the associated POI of method based on projection formula, uses map plane Dynamic Establishing plane map index;
The 5th step, compares the index of the corresponding index of building in map in described the 4th step and the generation of POI database, in the time that comparative result is consistent, this building in map is associated with in user's gps coordinate and POI database;
The 6th step, generates the different POI information of user, comprises UAD and behavioral data.
2. a kind of general POI information association method according to claim 1, is characterized in that, the concrete steps of extracting the longitude and latitude data in user profile in the described first step are:
First, in task start, load known extracting rule file, form the key-value pair of key-value distributed memory system, be loaded in internal memory;
Further, in the data processing stage, scan original data on flows according to the established rule loading, go out and user-dependent positional information from original flow extracting data, in original data on flows, the information relevant to customer location is all stored in the URL of data on flows, URL is the abbreviation of UniformResourceLocator, represent URL, according to the server name in URL, look for extracting rule corresponding in internal memory, use the regular expression finding, from URL, carry out the extraction of longitude and latitude data;
Further, after extraction, each user can form a user's latitude and longitude coordinates tables of data extracting according to the set rule of POI system and regular expression.
3. a kind of general POI information association method according to claim 1, is characterized in that, the longitude and latitude data that use multiple resource files and 2 kinds of modes of statistics excavation respectively the first step to be extracted in described second step are cleaned.
4. a kind of general POI information association method according to claim 3, is characterized in that, the concrete steps that the longitude and latitude data that use multiple resource files and 2 kinds of modes of statistics excavation respectively the first step to be extracted in described second step are cleaned are:
In the time adopting multiple resource files to clean the longitude and latitude data of extracting, first be the blacklist information to coordinate points, find to exist a large amount of repetition longitude and latitude data by frequency analysis in data, and the different user of longitude and latitude data correlation who repeats, because longitude and latitude data have all been accurate to four after decimal point to more than five, the registration of data itself is very little, draw by analysis the down town point that these repetition longitude and latitude data are each cities, the reason that occurs these points is in the time that the application software of a mobile terminal can not be obtained active user's positional information, the latitude and longitude information of meeting using the down town point at active user place as active user, so class data are wrong, therefore, this type of longitude and latitude data rows is entered to blacklist, complete the preliminary filtration to extract longitude and latitude data from flow, the region, province that user belongs to is simultaneously fixed, if user's longitude and latitude data do not belong in this province, these longitude and latitude data is piped off,
In the time adopting statistics mining algorithm to clean the longitude and latitude data of extracting, first open equipment for surfing the net application, there will be the inaccurate phenomenon in location, this type of is located inaccurate data and has also been blended in data on flows; Because equipment for surfing the net is constant, the positional information that user sends by equipment for surfing the net should be also constant or mobility scale is very little; In the time that data are accumulated to a magnitude, service range formula, can calculate the distance between each anchor point, in the time occurring that the distance of a point and other points is all greater than other and puts distance each other, this point is just listed in abnormity point, therefore, delete this type of longitude and latitude data, the longitude and latitude data that extract from flow are done further and filtered.
5. a kind of general POI information association method according to claim 2, it is characterized in that, in described the 3rd step, use the historical data of equipment for surfing the net for not extracting the URL of longitude and latitude data in the first step, supplement longitude and latitude data, the concrete steps that realize the expansion of user's longitude and latitude data after cleaning are:
In the time that user uses equipment for surfing the net online at every turn, all can on corresponding equipment for surfing the net, send URL, in the time extracting the longitude and latitude of this equipment for surfing the net, can preserve the longitude and latitude data of this equipment for surfing the net; In the time need to extracting the longitude and latitude of this equipment for surfing the net some day, even can not extract longitude and latitude data from the URL of this online, but by the historical longitude and latitude data of preserving, still can draw the longitude and latitude data of this equipment for surfing the net, realize the expansion of user's longitude and latitude data.
6. a kind of general POI information association method according to claim 1, is characterized in that, adopts the associated POI of method based on projection formula in described the 4th step, uses the concrete steps of map plane Dynamic Establishing plane map index to be:
The map of world map, national map and each provinces and cities is all known, can know clearly information and the geographical position of each building; First, by the earth along development of a sphere, become a plane, and can obtain accurately the longitude and latitude of each building on map, use projection formula by any one building projection in a plane, capital produces an index, index represents by (x, y), can obtain being positioned at a grid (x in plane, y), be the index of this building.
7. a kind of general POI information association method according to claim 6, is characterized in that, the concrete steps that in described the 5th step, this building in map are associated with in user's gps coordinate and POI database are:
The projection rectangle size of processing POI database is arranged to process in map the rectangle size of building in described the 4th step consistent, according to the method for building up of described the 4th step midplane map reference, the longitude and latitude data that extract in user's url data from be stored in POI database also can produce (an a, b) index, as plane map index (x known in the index producing in POI database and the 4th step, y) when consistent, can think that this building is exactly this user's POI data, i.e. position.
8. a kind of general POI information association method according to claim 7, is characterized in that, in described the 6th step, generates the POI information that user is different, comprises that the concrete steps of UAD and behavioral data are:
The POI data of each user on diverse location are drawn by described the 5th step, because user is mobile, so user can produce many different POI data in one day, the type of POI data is divided into Static and dynamic, and static state refers to the POI attribute data of user job region or home area; Dynamically refer to the POI behavioral data of the different location of process in user's a period of time;
Generate user POI attribute data according to user's historical behavior data:
Before calculating user POI attribute data, need advanced line number according to one's analysis, and then judge between operation interval and nonclient area;
First, define multiple working times and non-working time, if it is on POI position, workplace that user appears at classification within the working time of current definition, in the differential counting of this user job, add 1 so, contrary, appear on one's own time on the POI position, residential quarter of family's classification, in the differential counting of this subscriber household, add 1 so; Last each user can form a statistic record, within the predefined time to this user the counting respectively under work and family sight; Use all users' count information, result is signed on a scatter diagram, transverse axis represents operative scenario counting, and the longitudinal axis represents family's scene counting, and the reduced time, interval classification dropped on anchor point on the position near the longitudinal axis, transverse axis; Finally selecting one, to be positioned at data on the longitudinal axis and transverse axis maximum, and the most interval at the data boundary line delimitation in the longitudinal axis and transverse axis region, and as the working time, the non-working time is determined according to the method described above, and then judges working time and non-working time; If user appears on certain POI position within the working time, the work POI data that this POI position is exactly this user so, if within the working time, a user has appeared at multiple work POI position, select so POI position that user's time of occurrence is maximum as the unique work POI of user, the POI of family determines in the manner described above;
User's current behavior data generate POI behavioral data:
Due in described the 5th step, for each known map physical location, taking physical location as the center of circle, be defined in certain radius region the user who occurs and be defined as and appear on this map physical location; When a user is in being considered this user's POI behavioral data during with interior region apart from this building radius, and exceeding this radius region, even if POI data have been projected on same grid, can not serve as this user's POI behavioral data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510957892.9A CN105608153A (en) | 2015-12-18 | 2015-12-18 | Universal POI information association method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510957892.9A CN105608153A (en) | 2015-12-18 | 2015-12-18 | Universal POI information association method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105608153A true CN105608153A (en) | 2016-05-25 |
Family
ID=55988093
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510957892.9A Pending CN105608153A (en) | 2015-12-18 | 2015-12-18 | Universal POI information association method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105608153A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358459A (en) * | 2017-06-15 | 2017-11-17 | 浙江启冠网络股份有限公司 | Advertisement placement method and system based on orientation house property interest crowd |
| CN107436925A (en) * | 2017-07-07 | 2017-12-05 | 杭州冉驰科技有限公司 | A kind of POI data search method and correlating method |
| CN107480246A (en) * | 2017-08-10 | 2017-12-15 | 北京中航安通科技有限公司 | A kind of recognition methods of associate people and device |
| CN107580069A (en) * | 2017-09-22 | 2018-01-12 | 挖财网络技术有限公司 | The determination method and device of station address |
| CN108021638A (en) * | 2017-11-28 | 2018-05-11 | 上海电科智能系统股份有限公司 | A kind of unstructured address resolution system of offline geocoding |
| CN109657148A (en) * | 2018-12-24 | 2019-04-19 | 北京百度网讯科技有限公司 | For abnormal operation recognition methods, device, server and the medium for reporting POI |
| CN109977179A (en) * | 2019-02-19 | 2019-07-05 | 中电海康集团有限公司 | A kind of poly- scattered method for visualizing of grid dividing realization magnanimity atural object mark point of layering |
| CN110096645A (en) * | 2019-05-07 | 2019-08-06 | 北京百度网讯科技有限公司 | Information recommendation method, device, equipment and medium |
| CN110324845A (en) * | 2019-07-16 | 2019-10-11 | 秒针信息技术有限公司 | A kind of method and device of resource allocation |
| WO2020172954A1 (en) * | 2019-02-28 | 2020-09-03 | 东南大学 | Living circle identification method based on positioning data |
| CN112925773A (en) * | 2019-12-10 | 2021-06-08 | 中国再保险(集团)股份有限公司 | POI (Point of interest) data cleaning and fusing method and device for constructing industry risk exposure database |
| CN114495484A (en) * | 2021-12-17 | 2022-05-13 | 北京航空航天大学杭州创新研究院 | Multi-source data hierarchical graph clustering algorithm-based bus station position recommendation method |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101452463A (en) * | 2007-12-05 | 2009-06-10 | 浙江大学 | Method and apparatus for directionally grabbing page resource |
| CN103258057A (en) * | 2013-06-03 | 2013-08-21 | 北京奇虎科技有限公司 | Method and device for displaying point of interest on electronic map interface |
| CN103577678A (en) * | 2012-08-07 | 2014-02-12 | 上海市测绘院 | Method of accurately measuring any deformation map |
| CN103886082A (en) * | 2014-03-26 | 2014-06-25 | 百度在线网络技术（北京）有限公司 | Method and device for verifying position information of interest points |
| CN103884345A (en) * | 2012-12-20 | 2014-06-25 | 中国移动通信集团辽宁有限公司 | Interest point information collecting method, interest point information displaying method, interest point information collecting device, interest point information displaying device, and interest point retrieval system |
| CN104581622A (en) * | 2013-10-28 | 2015-04-29 | 华为技术有限公司 | Mobile user position prediction method and device |
| CN104572956A (en) * | 2014-12-29 | 2015-04-29 | 北京奇虎科技有限公司 | System and method for confirming POI information effectiveness |
| CN104657506A (en) * | 2015-03-13 | 2015-05-27 | 百度在线网络技术（北京）有限公司 | Data processing method and device based on user scenario |
| CN104750751A (en) * | 2013-12-31 | 2015-07-01 | 华为技术有限公司 | Method and device for annotating trace data |
| CN104866542A (en) * | 2015-05-05 | 2015-08-26 | 腾讯科技（深圳）有限公司 | POI data verification method and device |
| CN105069079A (en) * | 2015-07-31 | 2015-11-18 | 北京奇虎科技有限公司 | Method and device for screening point of interest POI data |
| CN105159921A (en) * | 2015-07-29 | 2015-12-16 | 北京奇虎科技有限公司 | Method and apparatus for de-duplicating point-of-interest (POI) data in map |
| CN105160031A (en) * | 2015-09-30 | 2015-12-16 | 北京奇虎科技有限公司 | Mining method and device for map point of interest (POI) data |
-
2015
- 2015-12-18 CN CN201510957892.9A patent/CN105608153A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101452463A (en) * | 2007-12-05 | 2009-06-10 | 浙江大学 | Method and apparatus for directionally grabbing page resource |
| CN103577678A (en) * | 2012-08-07 | 2014-02-12 | 上海市测绘院 | Method of accurately measuring any deformation map |
| CN103884345A (en) * | 2012-12-20 | 2014-06-25 | 中国移动通信集团辽宁有限公司 | Interest point information collecting method, interest point information displaying method, interest point information collecting device, interest point information displaying device, and interest point retrieval system |
| CN103258057A (en) * | 2013-06-03 | 2013-08-21 | 北京奇虎科技有限公司 | Method and device for displaying point of interest on electronic map interface |
| CN104581622A (en) * | 2013-10-28 | 2015-04-29 | 华为技术有限公司 | Mobile user position prediction method and device |
| CN104750751A (en) * | 2013-12-31 | 2015-07-01 | 华为技术有限公司 | Method and device for annotating trace data |
| CN103886082A (en) * | 2014-03-26 | 2014-06-25 | 百度在线网络技术（北京）有限公司 | Method and device for verifying position information of interest points |
| CN104572956A (en) * | 2014-12-29 | 2015-04-29 | 北京奇虎科技有限公司 | System and method for confirming POI information effectiveness |
| CN104657506A (en) * | 2015-03-13 | 2015-05-27 | 百度在线网络技术（北京）有限公司 | Data processing method and device based on user scenario |
| CN104866542A (en) * | 2015-05-05 | 2015-08-26 | 腾讯科技（深圳）有限公司 | POI data verification method and device |
| CN105159921A (en) * | 2015-07-29 | 2015-12-16 | 北京奇虎科技有限公司 | Method and apparatus for de-duplicating point-of-interest (POI) data in map |
| CN105069079A (en) * | 2015-07-31 | 2015-11-18 | 北京奇虎科技有限公司 | Method and device for screening point of interest POI data |
| CN105160031A (en) * | 2015-09-30 | 2015-12-16 | 北京奇虎科技有限公司 | Mining method and device for map point of interest (POI) data |
Non-Patent Citations (1)
| Title |
|---|
| 刘耀林: "《地理信息系统》", 30 November 2004, 中国农业出版社 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358459A (en) * | 2017-06-15 | 2017-11-17 | 浙江启冠网络股份有限公司 | Advertisement placement method and system based on orientation house property interest crowd |
| CN107436925A (en) * | 2017-07-07 | 2017-12-05 | 杭州冉驰科技有限公司 | A kind of POI data search method and correlating method |
| CN107436925B (en) * | 2017-07-07 | 2023-11-17 | 江苏艾德锐电子科技有限公司 | POI data retrieval method and association method |
| CN107480246A (en) * | 2017-08-10 | 2017-12-15 | 北京中航安通科技有限公司 | A kind of recognition methods of associate people and device |
| CN107580069A (en) * | 2017-09-22 | 2018-01-12 | 挖财网络技术有限公司 | The determination method and device of station address |
| CN108021638B (en) * | 2017-11-28 | 2022-01-14 | 上海电科智能系统股份有限公司 | Offline geocoding unstructured address resolution system |
| CN108021638A (en) * | 2017-11-28 | 2018-05-11 | 上海电科智能系统股份有限公司 | A kind of unstructured address resolution system of offline geocoding |
| CN109657148B (en) * | 2018-12-24 | 2020-10-13 | 北京百度网讯科技有限公司 | Abnormal operation identification method, device, server and medium for reported POI |
| CN109657148A (en) * | 2018-12-24 | 2019-04-19 | 北京百度网讯科技有限公司 | For abnormal operation recognition methods, device, server and the medium for reporting POI |
| CN109977179A (en) * | 2019-02-19 | 2019-07-05 | 中电海康集团有限公司 | A kind of poly- scattered method for visualizing of grid dividing realization magnanimity atural object mark point of layering |
| WO2020172954A1 (en) * | 2019-02-28 | 2020-09-03 | 东南大学 | Living circle identification method based on positioning data |
| CN110096645A (en) * | 2019-05-07 | 2019-08-06 | 北京百度网讯科技有限公司 | Information recommendation method, device, equipment and medium |
| CN110324845A (en) * | 2019-07-16 | 2019-10-11 | 秒针信息技术有限公司 | A kind of method and device of resource allocation |
| CN110324845B (en) * | 2019-07-16 | 2022-08-12 | 秒针信息技术有限公司 | Resource allocation method and device |
| CN112925773A (en) * | 2019-12-10 | 2021-06-08 | 中国再保险(集团)股份有限公司 | POI (Point of interest) data cleaning and fusing method and device for constructing industry risk exposure database |
| CN114495484A (en) * | 2021-12-17 | 2022-05-13 | 北京航空航天大学杭州创新研究院 | Multi-source data hierarchical graph clustering algorithm-based bus station position recommendation method |
| CN114495484B (en) * | 2021-12-17 | 2023-10-27 | 北京航空航天大学杭州创新研究院 | Bus stop position recommendation method based on multi-source data hierarchical graph clustering algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105608153A (en) | Universal POI information association method | |
| US10455363B2 (en) | Systems and methods for using geo-blocks and geo-fences to discover lookalike mobile devices | |
| CN103795613B (en) | Method for predicting friend relationships in online social network | |
| KR102121361B1 (en) | Method and device for identifying the type of geographic location where the user is located | |
| Brughmans et al. | Understanding inter-settlement visibility in Iron Age and Roman Southern Spain with exponential random graph models for visibility networks | |
| US9401100B2 (en) | Selective map marker aggregation | |
| Huang et al. | A 100 m population grid in the CONUS by disaggregating census data with open-source Microsoft building footprints | |
| CN103647800A (en) | Method and system of recommending application resources | |
| CN103839290A (en) | Information processing apparatus, terminal apparatus, information processing method, and program | |
| US20170214752A1 (en) | Systems and methods for providing geographically delineated content author information | |
| US9690007B2 (en) | Device and method for comparison of multiple tropical cyclone routes | |
| CN108259638A (en) | Personal group list intelligent sorting method, intelligent terminal and storage medium | |
| CN107038620A (en) | Based on user call a taxi preference information push and device | |
| CN103744916A (en) | Method and equipment for determining heat information of target video | |
| CN107395680A (en) | Shop group's information push and output intent and device, equipment | |
| CN104750716A (en) | Social network site friend recommendation method and social network site friend recommendation device | |
| Pitombo et al. | Proposal of a sequential method for spatial interpolation of mode choice | |
| CN108038734B (en) | Urban commercial facility spatial distribution detection method and system based on comment data | |
| CN109977324A (en) | A kind of point of interest method for digging and system | |
| CN105630842A (en) | Method and device for establishing air quality information database and determining pollution information | |
| CN103838861B (en) | A kind of method and display system based on three-dimension GIS dynamic display information | |
| CN110738734B (en) | Three-dimensional topography model generation method and device, processor and electronic device | |
| Evans et al. | Was Thebes Necessary? Contingency in Spatial Modeling | |
| US20140358897A1 (en) | Location awareness using local semantic scoring | |
| CN113989629A (en) | Memory, high consequence area determination method, device and equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160525 |
|
| RJ01 | Rejection of invention patent application after publication |