CN110288023A - Fusion method and device, detection method, acquisition methods, server and vehicle - Google Patents
Fusion method and device, detection method, acquisition methods, server and vehicle Download PDFInfo
- Publication number
- CN110288023A CN110288023A CN201910562056.9A CN201910562056A CN110288023A CN 110288023 A CN110288023 A CN 110288023A CN 201910562056 A CN201910562056 A CN 201910562056A CN 110288023 A CN110288023 A CN 110288023A
- Authority
- CN
- China
- Prior art keywords
- data
- pond
- matching
- fusion method
- pairing
- 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
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Remote Sensing (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
This application discloses a kind of fusion methods of multi-data source based on geography information.Fusion method includes: two data sources chosen in multiple data sources；Preliminary matches are carried out to the data in two data sources, and matching result is placed in pairing pond to be selected；It treats apolegamy accurately to match the data in pond, and matching result is placed in final result pond；The fused data of two data sources is obtained with according to the data in final result pond.In the fusion method of the application embodiment, information in multiple data sources can be merged, the effective information of the fused fully compliant each data source of data, user when carrying out acquisition of information each data source of obtained data compatibility, more comprehensively effectively.Disclosed herein as well is a kind of fusing device, detection method, geography information acquisition methods, vehicle and servers.
Description
Technical field
This application involves data processing field, in particular to a kind of multi-data source fusion method based on geographical location is melted
Attach together set, detection method, geography information acquisition methods, server and vehicle.
Background technique
In general, user can inquire traffic route and Locale information by map application, such as location or destination
The information such as neighbouring gas station, charging pile and parking lot.However the cartographic information of the application program of different developers' offers
Information not intercommunication between data source, in this way, the information of compatible multiple data sources can not be obtained by single application program.
Summary of the invention
In view of this, embodiments herein provides a kind of multi-data source fusion method based on geographical location, fusion
Device, detection method, geography information acquisition methods, server and vehicle.
This application provides a kind of fusion methods of multi-data source based on geography information, comprising:
Choose two data sources in multiple data sources；
Preliminary matches are carried out to the data in described two data sources, and matching result is placed in pairing pond to be selected；
Data in the pairing pond to be selected are accurately matched, and matching result is placed in final result pond；
The fused data of described two data sources is obtained according to the data in the final result pond.
In some embodiments, the fusion method further include:
Data prediction is carried out to the data in each data source.
In some embodiments, the data in described two data sources carry out preliminary matches and by matching results
Being placed in pairing pond to be selected includes:
It is preliminary to carry out that keyword fusion is carried out to two data sources by the data prediction according to parameter preset
Matching.
In some embodiments, the parameter preset includes:
The geographical location information of the number of the longest common subsequence of the place name of the data and the data away from
From.
In some embodiments, the data in described two data sources carry out preliminary matches and by matching results
Being placed in pairing pond to be selected includes:
Make a reservation for when being greater than or equal in two data sources there are the number of the longest common subsequence of the place name of data
When number and the distance of geographical location information are less than or equal to preset distance, confirm data preliminary matches, be placed into pairing to be selected
Pond.
In some embodiments, the data in the pairing pond to be selected are accurately matched and by matching result
Merging final result pond includes:
When the corresponding relationship one-to-one there are two data sources of the data in the pairing pond to be selected, by the pair of one
Two data of corresponding relationship are placed in the final result pond.
In some embodiments, the data in the pairing pond to be selected are accurately matched and by matching result
Merging final result pond includes:
Building matching is carried out to the data of the pairing Chi Zhongfei one to be selected, the building will be passed through
The data that one is formed after object matching are placed in the final result pond.
In some embodiments, the data in the pairing pond to be selected are accurately matched and by matching result
Merging final result pond includes:
Carrying out title to the data of the pairing Chi Zhongfei one to be selected includes matching, will pass through the name
The data comprising forming one after matching are claimed to be placed in the final result pond.
In some embodiments, the fusion method method further include:
Judge whether the address name of two in described two data sources data matches；
Judge that the distance of the geographical location information of the data is less than predetermined threshold when address name matching；
Judge whether described two data are same building object when the distance is less than the predetermined threshold；
If so, matching result is placed in the final result pond.
The application provides a kind of detection method, for detecting the fused data obtained according to fusion method as described above
Accuracy is detected, and the detection method includes:
Set the characteristic value of logistic regression；
Logistic regression monitoring model is established according to the characteristic value；
Confidence level detection is carried out to the data in the final result pond according to the logistic regression monitoring model.
This application provides a kind of geographical position information acquisition methods, comprising:
Receive the geographic location information query request of user；
Corresponding geographical location information is returned to according to the inquiry request, the geographical location information is according to as described above
Fusion method generates.
The application provides a kind of fusing device of multi-data source based on geography information, comprising:
Module is chosen, for choosing two data sources in multiple data sources；
First matching module, for carrying out preliminary matches to the data in described two data sources and being placed in matching result
Pairing pond to be selected；
Second matching module for accurately being matched to the data in the pairing pond to be selected, and matching result is set
Enter final result pond；
Fusion Module, for obtaining the fused data of described two data sources according to the data in the final result pond.
This application provides a kind of servers, comprising:
One or more processors, memory；With
One or more programs, wherein one or more of programs are stored in the memory, and described
One or more processors execute, and described program includes for executing melting for the multi-data source based on geography information as described above
The instruction of conjunction method.
This application provides a kind of vehicles, comprising: middle control display screen and processor；
The middle control display screen is used to receive the geographic location information query request of user；
The processor is used to return corresponding geographical location information according to the inquiry request and is shown by the middle control
Display screen shows that the geographical location information, the geographical location information are generated according to fusion method as described above.
This application provides a kind of non-volatile computer readable storage medium storing program for executing, when the computer executable instructions are by one
When a or multiple processors execute, so that the processor executes the fusion of the multi-data source based on geography information as described above
The instruction of method.
The multi-data source fusion method based on geographical location, fusing device, the detection method, geography of the application embodiment
In information acquisition method, server, vehicle and computer readable storage medium, the information in multiple data sources can be melted
It closes, the effective information of the fused fully compliant each data source of data, user's obtained data when carrying out acquisition of information
Compatible each data source, more comprehensively effectively.
Detailed description of the invention
The application is above-mentioned and/or additional aspect and advantage will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, in which:
Fig. 1 is the flow diagram of the fusion method of multi-data source of the application certain embodiments based on geography information.
Fig. 2 is the module diagram of the fusing device of multi-data source of the application certain embodiments based on geography information.
Fig. 3 is the status diagram of the fusion method of the application certain embodiments.
Fig. 4-9 is the flow diagram of the fusion method of the application certain embodiments.
Figure 10 is the status diagram of the fusion method of the application certain embodiments.
Figure 11 is the flow diagram of the detection method of the application certain embodiments.
Figure 12 is the flow diagram of the geographical position information acquisition method of the application certain embodiments.
Figure 13 is the display partial schematic diagram of the fused data of the application certain embodiments.
Specific embodiment
Embodiments herein is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, it is intended to for explaining the application, and should not be understood as the limitation to the application.
Generally, user before travel can Locale information near the traffic route to destination on earth or destination or
Say that information point is inquired by map application, however the data of the cartographic information of the application program of different developer's offers
Therefore information not intercommunication between source may have loss of learning using single application program, and use multiple applications
Program one is relatively complicated to inquire repeatedly, and two come since Naming conventions of each data source for information point are different, will lead to pair
Returning the result in substantially identical information point may be different, the puzzlement that user is in use is caused, in addition, returning the result identical
Information point becomes redundancy, and user experience is bad.
Referring to Fig. 1, the application first aspect embodiment provides a kind of melting for multi-data source based on geography information
Conjunction method, comprising:
S10: two data sources in multiple data sources are chosen；
S20: preliminary matches are carried out to the data in two data sources, and matching result is placed in pairing pond to be selected；
S30: it treats apolegamy and the data in pond is accurately matched, and matching result is placed in final result pond；With
S40: the fused data of two data sources is obtained according to the data in final result pond.
Fig. 2 and Fig. 3 are please referred to, embodiment further provides a kind of fusions of multi-data source based on geography information by the application
Device 100.Fusing device 100 includes choosing module 10, the first matching module 20, the second matching module 30 and Fusion Module 40.
The fusion method of the multi-data source based on geography information of the application embodiment can be realized by fusing device 100.
Specifically, S10 can be realized by selection module 10, and S20 can be realized by the first matching module 20, and S30 can be by
Second matching module 30 realizes that S40 can be realized by Fusion Module 40.In other words, module 10 is chosen for choosing multiple data
Two data sources in source.First matching module 20 is used to carry out the data in two data sources preliminary matches, and will matching
As a result it is placed in pairing pond to be selected.Second matching module 30 accurately matches the data in pond for treating apolegamy, and general
Final result pond is placed in result.Fusion Module 40 is used to obtain the fusion of two data sources according to the data in final result pond
Data.
It, can be in the fusion method and fusing device 100 of the multi-data source based on geography information of the application embodiment
Information in multiple data sources is merged, the effective information of the fused fully compliant each data source of data, Yong Hu
Each data source of obtained data compatibility of institute when acquisition of information is carried out, it is more comprehensively effective.
Specifically, described that foundation of multiple data sources when carrying out data fusion is referred to based on geography information, it that is to say
Multiple data fusions actual geographic information is identical and that the form of expression is different are a data, or perhaps multiple reality are identical
Data be directed toward the same information point jointly, such as the parking lot for being A for actual geographic information, the name in data source I
For parking lot A, it is named as parking lot A ' in data source II, the two can be associated after fusion.
Two data sources are merged when data source is two, work as number by the unlimited number for determining data source in the application
According to source number more than two when, two data sources are merged first, data and third data source after goodbye will merge
It is merged, and so on, until the fusion of total data source finishes.
In an example it is assumed that data source I and data source II respectively have 50 geography information, wherein having 25 information is two
A platform is shared, and after the fusion method processing of present embodiment, obtained fused data will include 75 information, this 75
A information is sufficiently extracted and has been compatible with the effective information of data source I and data source II.
Referring to Fig. 4, in some embodiments, fusion method further include:
S00: data prediction is carried out to the data in each data source.
In some embodiments, fusing device 100 further includes preprocessing module 50.S00 can be by preprocessing module 50
It realizes, in other words, preprocessing module 50 is used to carry out data prediction to the data in each data source.
Specifically, S00 can be implemented before S10, in other words, first all be carried out the data in each data source pre-
Processing, is then merged in the data source that will be singled out.Certainly, S00 can also be implemented after S10, in other words, chosen
After the data source merged, then the data in two data sources are pre-processed, pretreatment is merged.
It is to be appreciated that it is different for the Naming conventions of same information point in different data sources, therefore first have to data
It is pre-processed.
The operation of data prediction may include but be not limited to following content:
I. the character boundary conversion in the title and address of information point corresponding to data.Can include by in practical operation
Upper case character be converted to lowercase character.
Ii. the specially treated of the character in the title and address of information point corresponding to data.For example, will capitalization or Chinese character
Number be converted to Arabic numerals, double byte character is converted to half-angle character etc..
Iii. the General adaptive of the title of information point corresponding to data and address.For example, for " charging station ", " power-up
Stand ", the characters such as " charging pile " be regarded as matched.
Iv. the title of information point corresponding to data and the filtering of the redundancy of address.It is to be appreciated that for information point
Matching, such as the matching of information point in unified city is carried out, then such as provinces and cities' information may have no for matching in address
It is practical to help, therefore can be used as redundancy filtering.Further, for the functional specification in some places, such as " hospital ",
" school ", " airport " etc. are not also of practical assistance usually in the matching process, and title has stronger uniqueness, therefore can be with
It can be filtered the functional specification in place as redundancy.Certainly, redundancy is not limited to the disclosure as set forth herein and shows
Example, such as can also have the restriction in orientation, such as the four corners of the world etc..Specifically can according to demand, equipment computing capability and accurate
Degree demand is adjusted.
In this way, carrying out data prediction to data is that follow-up data matching is ready.
Referring to Fig. 5, in some embodiments, S20 includes:
S21: it is preliminary to carry out that keyword fusion is carried out to two data sources by data prediction according to parameter preset
Matching.
In some embodiments, S21 can be realized by the first matching module 20.In other words, the first matching module 20 is used
It merges in carrying out keyword to two data sources by data prediction according to parameter preset to carry out preliminary matches.
Specifically, keyword will be carried out to treated data according to parameter preset after the pretreatment of aforementioned data to melt
It closes.Parameter preset that is to say formulates certain matching rules to by some restrictions of pretreated data progress in other words, thus
So that relevant data are able to carry out preliminary matches in two data sources.For example, whether the title of two data is identical, two numbers
According to address information it is whether consistent etc., it is not limited here.
In some embodiments, the parameter preset includes: the longest common subsequence of the place name of data
The distance of the geographical location information of the number and data of (Longest Common Subsequence, LCS).
Specifically, LCS length is the character length of longest subsequence in all sequences in the two sequences.In an example
In, certain information point is entitled " Guangzhou Red Star hotel parking lot " in data source I, certain information point is entitled in data source II
" Tianhe District Red Star underground parking (east gate) ".Remaining keyword is respectively " Red Star " and " Red Star is big " after pretreatment, is then counted
The LCS=2 for calculating " Red Star " and " Red Star is big " that is to say that the character length of longest subsequence is 2.
The distance of the geography information position of data is the geographic distance of above-mentioned two information point, for example, can pass through GPS information
Whether the position to judge the two is same or similar.
Referring to Fig. 6, in such an embodiment, S20 includes:
S22: when there are the numbers of the longest common subsequence of the place name of data to be greater than or equal in two data sources
When predetermined number and the distance of geographical location information are less than or equal to preset distance, confirm data preliminary matches, be placed into be selected
Match pond.
In some embodiments, S22 can be realized by the first matching module 20.In other words, the first matching module 20 is used
In in two data sources there are the number of the longest common subsequence of the place name of data be greater than or equal to predetermined number and
When the distance of geographical location information is less than or equal to preset distance, confirmation data preliminary matches are placed into pairing pond to be selected.
Specifically, by data prediction, the character quantity of information point title is usually fewer, therefore, the public son of longest
The predetermined number setting of sequence should not be too large, and can be adjusted conjunction according to the actual conditions of pretreated data in practical operation
Reason setting.For example, if being set as 1, it is contemplated that the probability in title with an identical characters is bigger, such preliminary matches
As a result more numerous and jumbled, follow-up work is relatively complicated.If being set as 5, character quantity is smaller after information point title process, can
A large amount of related data can be filtered out, and cause final fused data less.In the present embodiment, longest common subsequence
Predetermined number be set as 2.
The distance of geographical location information is by the distance between two information point GPS points.It is to be appreciated that for same letter
The position that breath point different data sources are identified it is different, therefore corresponding GPS point is different, such as the same parking
, data source I may using parking lot center as identification point, and data source II may using the entrance in parking lot as identification point,
Although the two essence is same information point, GPS coordinate is not identical, in this case, detects the GPS of two data
Distance.When the distance is less than preset distance, it is believed that the two preliminary matches.
It should be noted that due to parameter preset and without limitation, parameter preset meets condition also with default
The variation of parameter and change, when carrying out preliminary matches, data need to meet the condition of whole parameter presets.
Referring to Fig. 7, in some embodiments, S30 includes:
S31: it when the corresponding relationship one-to-one there are two data sources of the data in pairing pond to be selected, will correspond
Two data of relationship are placed in final result pond.
In some embodiments, S31 can be realized by the second matching module 30.In other words, the second matching module 30 is used
When data one-to-one there are two data sources corresponding relationship in pairing pond to be selected, by two of one
Data are placed in final result pond.
Specifically, after the preliminary matches of above-mentioned data, for matching result, there may be more for corresponding relationship in other words
Kind result.Situations such as specifically including one-to-one, one-to-many, many-one, multi-to-multi.It is to be appreciated that if by preliminary
It is one-to-one corresponding relationship there are two data after matching, it is believed that matched data confidence is higher, can be placed directly within final
In outcome pool.And the data of other non-ones need further progress screening until no longer depositing in pairing pond to be selected
In the data that can be matched.
Fig. 8-10 is please referred to, in some embodiments, S30 includes:
S32: it treats apolegamy and building matching is carried out to the data of Chi Zhongfei one, building will be passed through
The data that one is formed after matching are placed in final result pond.
In some embodiments, S32 can be realized by the second matching module 30.In other words, the second matching module 30 is used
Building matching is carried out to the data of Chi Zhongfei one in treating apolegamy, and will be after building matching
The data for forming one are placed in final result pond.
Specifically, when the data after tentatively matching in two databases form such as one-to-many, many-one, multi-to-multi pair
It when should be related to, needs further to match data, in present embodiment, the data for forming non-one-to-one relationship is carried out
Building matching, whether building matching that is to say judges information point in same building object.For example, " the Red Star in data pool I
" Red Star hotel " in restaurant ", " Red Star supermarket " and data pool II, the preliminary pairing through keyword pairings, preset condition
Enter pairing pond to be selected afterwards and form many-to-one corresponding relationship in pairing pond to be selected, " Red Star is big through building matching confirmation
Hotel " and " Red Star hotel " they are same building objects, then it is assumed that two data successful match, and can be by " Red Star supermarket " mistake
Filter.
Building matching can realize that inverse geocoding is to pass through HTTP/HTTPS using the method for GPS against geocoding
The interface of protocol access remote service provides the ability mutually converted between structuring address and longitude and latitude.Inverse geocoding
Longitude and latitude can be converted to the address of detailed construction, and return to the information point on periphery nearby.Such as: GPS coordinate
(116.480881,39.989410), after converting address descriptor: Chaoyang District, Beijing City Fu Tong East Street 6.The application embodiment party
In formula, the address such as road name that we can return after inverse geocoding according to two data sources of comparison judges whether it is same
Building, to judge whether there is the data of one.
In addition, in some embodiments, for providing the data source of address information, we can be according to comparing two data
Source address such as road name is matched.For example, " Red Star parking lot " " Red Star in the II of matched data source simultaneously in data source I
Parking lot (east gate) " and " Red Star parking lot (west gate) ".So, it is matched by link name, it can be seen that in data source I
" Red Star parking lot " which parking lot in data source II be equal to actually.
In some embodiments, building matching can realize that information point search is using the method for search information point
Data are filtered by inputting the information of information point.Generally, each information point includes four aspect information, title, class
Not, latitude, longitude, peripheral information.Still in data pool I " Red Star restaurant ", in " Red Star supermarket " and data pool II
For " Red Star hotel " is matched.In matching, restaurant can be inputted to carry out classification search, return value is " Red Star restaurant "
And " Red Star hotel ", then it is believed that the two forms one-to-one matching relationship, and " Red Star supermarket " is filtered.
It is to be appreciated that matching by building, confirm that two data form one-to-one matching relationship, it is believed that number
According to accurate matching is completed, matching result is placed in final result pond, the data without successful matching may then be filtered or be waited
To further screen.
In some embodiments, S30 includes:
S33: treating apolegamy and carrying out title to the data of Chi Zhongfei one includes matching, will pass through title packet
Final result pond is placed in containing the data for forming one after matching.
In some embodiments, S33 can be realized by the second matching module 30, and in other words, the second matching module 30 is used
It include matching in treating apolegamy to carry out title to the data of Chi Zhongfei one, and will be after title includes matching
The data for forming one are placed in final result pond.
Specifically, rearward, priority matches title inclusion relation priority lower than building in other words.In other words, name
Inclusion relation is claimed to be used to carry out the data filtered after building matches matching again.For example, being wrapped in data source I
Include data: " white clouds parking lot " and " white clouds parking lot east gate ".It include data " A mouthfuls of white clouds parking lot " and " white in data source II
B mouthfuls of cloud parking lot ".The corresponding relationship of multi-to-multi is formed by preliminary matches, four.If " white clouds parking lot east gate " passes through building
Object matching forms one-to-one corresponding relationship, but " white clouds parking lot " and " A mouthfuls of white clouds parking lot " with " B mouthfuls of white clouds parking lot "
It matches to form one-to-one corresponding relationship not over building.So, by title inclusion relation to " white clouds parking lot " with
" A mouthfuls of white clouds parking lot " is accurately matched.It is appreciated that white clouds parking lot is to include for A mouthfuls by white clouds parking lot, therefore,
They can still be considered as forming one-to-one corresponding relationship and successful match because of inclusion relation.
Wherein, title inclusion relation is it is to be understood that the character string of one of data name is another or multiple numbers
According to the subset of name character string.
The data for including the still not formed one-one relationship of matching by title may then be filtered or be waited further
Screening.
In some embodiments, S30 includes:
S34: it treats apolegamy and minimum distance match is carried out to the data of Chi Zhongfei one, most narrow spacing will be passed through
Final result pond is placed in from the data for forming one after matching.
In some embodiments, S34 can be realized by the second matching module 30.In other words, the second matching module 30 is used
Minimum distance match is carried out to the data of Chi Zhongfei one in treating apolegamy, and will be after minimum distance match
The data for forming one are placed in final result pond.
Specifically, when the data for still having non-one after above-mentioned screening twice, most narrow spacing is carried out
From matching, for example, being judged at a distance from data of multiple data and this respectively, and select in many-to-one corresponding relationship
One-to-one corresponding relationship is formed apart from the smallest data and a data with a data in multiple data, is placed into
Final result pond.And other data are filtered.
In this way, above embodiment on the basis of guaranteeing accuracy, while being improved and being called together by repeatedly accurate matching
The rate of returning.Above-mentioned accurate matched condition can be regarded as or relationship, that is to say that meeting either condition is believed that successful match.
In other embodiments, fusion method further include:
Judge whether the address name of two data in two data sources matches；
Judge that the distance of the geographical location information of data is less than predetermined threshold when address name-matches；
Judge whether two data are same building object when distance is less than the predetermined threshold；
If so, matching result is placed in the final result pond.
It is to be appreciated that in such an embodiment, have any one link to be not matched to, data all can be by mistake
It filters, so, it is ensured that accuracy rate, but recall rate is lower.
In actual operation, different matching ways can be set according to actual needs to balance recall rate and accuracy rate.
In conclusion the multi-data source fusion method and device of the application embodiment, can pass through multiple matching conditions
Data in multiple data sources are carried out matching to realize that information merges, the fused fully compliant each data of data
The effective information in source, user when carrying out acquisition of information each data source of obtained data compatibility, more comprehensively effectively.
The embodiment of the application second aspect provides a kind of detection method.Detection method is used for according to the application first
The accuracy of fused data acquired in the fusion method of aspect embodiment is detected.
Figure 11 is please referred to, detection method includes:
S50: the characteristic value of logistic regression is set；
S60: logistic regression monitoring model is established according to characteristic value；
S70: confidence level detection is carried out to the data in final result pond according to logistic regression monitoring model.
Specifically, it is verified using logistic regression, the cost of artificial detection can be mitigated, certainly, if it is desired, still can be regular
Manual verification, selective examination.Satellite map, artificial observation information point title etc. are put by on-the-spot investigation, by information point first to institute
There is fused data to carry out result correctness manually to mark.Such as, it is believed that result correctly marks 1, it is believed that the mark of result mistake
0, in this way, the matching accuracy rate of the fused data obtained by above embodiment fusion method can be obtained.In annotation process,
The parameter preset that above-mentioned fusion method carries out preliminary matches can be adjusted at any time according to judging result, such as the longest of title is public
The distance of the geographical location information of the number and data of subsequence observes the accuracy rate of result, to obtain most suitable parameter
Combination.Parameter can also include but is not limited to: the length of data name longest common subsequence, the public son of the longest of data address
Whether identical sequence length, GPS distance between two information points, the operator of building, building orientation etc. be specific without limitation.
After determining parameter, the characteristic value of logistic regression is set, characteristic value can be normalized as logistic regression
Input.Judge that matching result is regarded as two classification problems in itself.In other words, there are two the end values of prediction
Value 0 or 1, wherein positive sample is that matching result is correct, and negative sample is matching result mistake.In some instances, it is believed that will
The output valve of logistic regression be more than or equal to 0.5 corresponding to input be classified as positive sample, less than 0.5 corresponding to input be classified as negative sample
This.A function is set, judging result is returned when inputting a characteristic value that is to say that the data are positive sample or negative sample.
Logistic regression monitoring model is used to export the probability and probability of failure of the successful match of machine prediction, by final result
In data in take a certain number of data as training set at random, using the result manually marked as true value, by characteristic value
As trained input.For example, logistic regression judgement is positive sample after two characteristic value inputs, and artificial annotation results are one
A correct mistake, then the result accuracy of machine prediction is 50%, error rate 50%.
In this way, allowing machine which kind of knows the result is that correct, which kind of result by a certain number of training sets of machine learning
It is wrong.The success of machine oneself prediction and matching and the probability that it fails to match are allowed by comparing with true value.
The output of model is to meet the matching result of characteristic value in fused data to be the probability correctly matched is how many, example
Such as, the output result obtained are as follows: 1 → 90%, 0 → 10%, indicate that matching correct probability is 90%.When confidence level that is to say just
True rate is greater than the threshold value of setting, such as 80% when exports fused data.It in actual operation can be according to the requirement to accuracy
Adjust confidence threshold value.In this way, the accuracy rate of pairing result can be monitored by the logistic regression monitoring model.
The embodiment of third aspect present invention provides a kind of geographical position information acquisition method.
Figure 12 and Figure 13 are please referred to, acquisition methods include:
S80: the geographic location information query request of user is received；
S90: corresponding geographical location information is returned to according to inquiry request.
Wherein, the geographical location information is raw according to the fusion method of the multi-data source of the application first aspect embodiment
At.
Embodiment further provides a kind of vehicles by the application.Vehicle includes middle control display screen and processor.The application is implemented
The reason location information acquisition method of mode can have the vehicle of application embodiment to realize.In other words, middle control display screen is for connecing
Receive the geographic location information query request of user.Processor is used to return to corresponding geographical location information according to the inquiry request
And geographical location information is shown by middle control display screen.Wherein, geographical location information is according to the application first aspect embodiment
Multi-data source fusion method generate.
Specifically, in user using the electronic device for loading multiple map applications, such as mobile phone, tablet computer, vehicle-mounted
When controlling display screen in liquid crystal, when carrying out information inquiry by any map application, it can return to based on multiple map application journeys
The fused data of sequence data source is to user, to get through the association between multiple data sources, provides for user more comprehensively effective
Information.
Embodiment further provides a kind of computer readable storage mediums by the application.One or more can be held comprising computer
The non-volatile computer readable storage medium storing program for executing of row instruction, when computer executable instructions are executed by one or more processors
When, so that processor executes the instruction of the fusion method of the multi-data source based on geography information of any of the above-described embodiment.
Embodiment further provides a kind of servers by the application.Server includes one or more processors, and one or more
A program is stored in memory, and is configured to be performed by one or more processors.Program includes for executing
State the fusion method of the multi-data source described in any one embodiment based on geography information.
Processor can be used for providing calculating and control ability, support the operation of entire server.The memory of server is
Memory computer-readable instruction operation therein provides environment.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Instruct relevant hardware to complete by computer program, program can be stored in a non-volatile computer readable storage medium
In matter, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, storage medium can for magnetic disk,
CD, read-only memory (Read-Only Memory, ROM) etc..
Above embodiments only express the several embodiments of the application, and the description thereof is more specific and detailed, but can not
Therefore it is interpreted as the limitation to the application the scope of the patents.It should be pointed out that for those of ordinary skill in the art,
Without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection model of the application
It encloses.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (15)
1. a kind of fusion method of the multi-data source based on geography information characterized by comprising
Choose two data sources in multiple data sources；
Preliminary matches are carried out to the data in described two data sources, and matching result is placed in pairing pond to be selected；
Data in the pairing pond to be selected are accurately matched, and matching result is placed in final result pond；
The fused data of described two data sources is obtained according to the data in the final result pond.
2. fusion method according to claim 1, which is characterized in that the fusion method further include:
Data prediction is carried out to the data in each data source.
3. fusion method according to claim 2, which is characterized in that the data in described two data sources carry out
Matching result is simultaneously placed in pairing pond to be selected and includes: by preliminary matches
Keyword is carried out to two data sources by the data prediction according to parameter preset to merge to carry out preliminary matches.
4. fusion method according to claim 3, which is characterized in that the parameter preset includes:
The distance of the geographical location information of the number of the longest common subsequence of the place name of the data and the data.
5. fusion method according to claim 4, which is characterized in that the data in described two data sources carry out
Matching result is simultaneously placed in pairing pond to be selected and includes: by preliminary matches
When there are the numbers of the longest common subsequence of the place name of data to be greater than or equal to predetermined number in two data sources
And the distance of geographical location information be less than or equal to preset distance when, confirm data preliminary matches, be placed into pairing pond to be selected.
6. fusion method according to claim 3, which is characterized in that the data in the pairing pond to be selected carry out
It is accurate to match and include: by matching result merging final result pond
It is corresponding by the pair of one when the corresponding relationship one-to-one there are two data sources of the data in the pairing pond to be selected
Two data of relationship are placed in the final result pond.
7. fusion method according to claim 3, which is characterized in that the data in the pairing pond to be selected carry out
It is accurate to match and include: by matching result merging final result pond
Building matching is carried out to the data of the pairing Chi Zhongfei one to be selected, the building will be passed through
The data that one is formed after matching are placed in the final result pond.
8. fusion method according to claim 3 or 7, which is characterized in that the data in the pairing pond to be selected
It is accurately matched and includes: by matching result merging final result pond
Carrying out title to the data of the pairing Chi Zhongfei one to be selected includes matching, will pass through the title packet
The final result pond is placed in containing the data for forming one after matching.
9. fusion method according to claim 1, which is characterized in that the fusion method method further include:
Judge whether the address name of two in described two data sources data matches；
Judge that the distance of the geographical location information of the data is less than predetermined threshold when address name matching；
Judge whether described two data are same building object when the distance is less than the predetermined threshold；
If so, matching result is placed in the final result pond.
10. a kind of detection method, for detecting the fused data that -9 described in any item fusion methods obtain according to claim 1
Accuracy detected, which is characterized in that the detection method includes:
Set the characteristic value of logistic regression；
Logistic regression monitoring model is established according to the characteristic value；
Confidence level detection is carried out to the data in the final result pond according to the logistic regression monitoring model.
11. a kind of geographical position information acquisition method characterized by comprising
Receive the geographic location information query request of user；
Corresponding geographical location information is returned to according to the inquiry request, the geographical location information is according to such as claim 1-9
Described in any item fusion methods generate.
12. a kind of fusing device of the multi-data source based on geography information characterized by comprising
Module is chosen, for choosing two data sources in multiple data sources；
First matching module, for carrying out preliminary matches to the data in described two data sources and being placed in matching result to be selected
Match pond；
Second matching module, for accurately being matched to the data in the pairing pond to be selected, and most by matching result merging
Whole outcome pool；
Fusion Module, for obtaining the fused data of described two data sources according to the data in the final result pond.
13. a kind of server characterized by comprising
One or more processors, memory；With
One or more programs, wherein one or more of programs are stored in the memory, and one
Or multiple processors execute, described program includes being believed described in -9 any one according to claim 1 based on geography for executing
The instruction of the fusion method of the multi-data source of breath.
14. a kind of vehicle characterized by comprising middle control display screen and processor；
The middle control display screen is used to receive the geographic location information query request of user；
The processor is used to return to corresponding geographical location information according to the inquiry request and passes through the middle control display screen
Show that the geographical location information, the geographical location information are raw according to the described in any item fusion methods of such as claim 1-9
At.
15. a kind of non-volatile computer readable storage medium storing program for executing, when the computer executable instructions are handled by one or more
When device executes, so that the processor perform claim requires the multi-data source based on geography information described in any one of 1-9
The instruction of fusion method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910562056.9A CN110288023A (en) | 2019-06-26 | 2019-06-26 | Fusion method and device, detection method, acquisition methods, server and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910562056.9A CN110288023A (en) | 2019-06-26 | 2019-06-26 | Fusion method and device, detection method, acquisition methods, server and vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110288023A true CN110288023A (en) | 2019-09-27 |
Family
ID=68006256
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910562056.9A Pending CN110288023A (en) | 2019-06-26 | 2019-06-26 | Fusion method and device, detection method, acquisition methods, server and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110288023A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191733A (en) * | 2020-01-02 | 2020-05-22 | 平安科技（深圳）有限公司 | Data fusion method and device for multiple data sources, electronic equipment and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102789467A (en) * | 2011-05-20 | 2012-11-21 | 腾讯科技（深圳）有限公司 | Data fusion method, data fusion device and data processing system |
| CN104572645A (en) * | 2013-10-11 | 2015-04-29 | 高德软件有限公司 | Method and device for POI (Point Of Interest) data association |
| CN105808609A (en) * | 2014-12-31 | 2016-07-27 | 高德软件有限公司 | Discrimination method and equipment of point-of-information data redundancy |
| CN107256230A (en) * | 2017-05-03 | 2017-10-17 | 昆明理工大学 | A kind of fusion method based on diversification geography information point |
| CN108345609A (en) * | 2017-01-24 | 2018-07-31 | 腾讯科技（深圳）有限公司 | A kind of method and apparatus of processing POI information |
| CN109284446A (en) * | 2018-08-01 | 2019-01-29 | 昆明理工大学 | A kind of POI information fusion method |
| CN109299203A (en) * | 2018-08-30 | 2019-02-01 | 深圳市赛为智能股份有限公司 | Based on BIM and GIS integrated digital map data storage method and device |
-
2019
- 2019-06-26 CN CN201910562056.9A patent/CN110288023A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102789467A (en) * | 2011-05-20 | 2012-11-21 | 腾讯科技（深圳）有限公司 | Data fusion method, data fusion device and data processing system |
| CN104572645A (en) * | 2013-10-11 | 2015-04-29 | 高德软件有限公司 | Method and device for POI (Point Of Interest) data association |
| CN105808609A (en) * | 2014-12-31 | 2016-07-27 | 高德软件有限公司 | Discrimination method and equipment of point-of-information data redundancy |
| CN108345609A (en) * | 2017-01-24 | 2018-07-31 | 腾讯科技（深圳）有限公司 | A kind of method and apparatus of processing POI information |
| CN107256230A (en) * | 2017-05-03 | 2017-10-17 | 昆明理工大学 | A kind of fusion method based on diversification geography information point |
| CN109284446A (en) * | 2018-08-01 | 2019-01-29 | 昆明理工大学 | A kind of POI information fusion method |
| CN109299203A (en) * | 2018-08-30 | 2019-02-01 | 深圳市赛为智能股份有限公司 | Based on BIM and GIS integrated digital map data storage method and device |
Non-Patent Citations (1)
| Title |
|---|
| 刘晓梅: "《笃行华文·专业汉语系列 应用现代汉语词汇学》", 31 October 2018 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191733A (en) * | 2020-01-02 | 2020-05-22 | 平安科技（深圳）有限公司 | Data fusion method and device for multiple data sources, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11740102B2 (en) | Method, apparatus, device and storage medium for determining point of interest area | |
| CN111951144B (en) | Method and device for determining violation road section and computer readable storage medium | |
| US10101165B2 (en) | Electronic device and method for planning tour route | |
| US9672240B2 (en) | Apparatus and method to update geographic database | |
| US8521128B1 (en) | Method, system, and computer program product for obtaining crowd-sourced location information | |
| US9359880B2 (en) | Methods and systems for managing underground assets | |
| US20150285656A1 (en) | Method of determining trajectories through one or more junctions of a transportation network | |
| US10107642B2 (en) | Map information management and correction of geodata | |
| US10234299B2 (en) | Geo-location tracking system and method | |
| US20130317746A1 (en) | Systems and methods for collecting and using user-contributed map data | |
| US10852921B2 (en) | Method of gathering, storing, and distributing user defined geographic location identities | |
| KR20190093811A (en) | Method And System For Providing Traveler-Generated Contents As Recommended Travel Information | |
| CN110597940A (en) | WeChat-based natural resource field check method and system, server and medium | |
| US10079888B2 (en) | Generation and use of numeric identifiers for locating objects and navigating in spatial maps | |
| Velosa et al. | Edge Solution with Machine Learning and Open Data to Interpret Signs for People with Visual Disability. | |
| CN104182486B (en) | Electronic map shows method and apparatus | |
| CN110288023A (en) | Fusion method and device, detection method, acquisition methods, server and vehicle | |
| WO2022033247A1 (en) | Road side positioning method and apparatus, and device and storage medium | |
| US20170064512A1 (en) | System and method for facilitating user connections in travel locations | |
| US8358201B1 (en) | Storage and recall of buried asset data over communications networks for damage avoidance and mapping | |
| US20160034515A1 (en) | Assessing Risk of Inaccuracies in Address Components of Map Features | |
| US10068299B1 (en) | System for providing tourist tracking on a mobile device | |
| Mahendra et al. | Geographic Information System of Public Complaint Testing Based On Mobile Web (Public Complaint) | |
| CN112699318A (en) | Man-machine interaction system and method for rapidly searching parking lot | |
| CN114743395B (en) | Signal lamp detection method, device, equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190927 |