CN104281577B - The sort method of data file - Google Patents

The sort method of data file Download PDF

Info

Publication number
CN104281577B
CN104281577B CN201310273231.5A CN201310273231A CN104281577B CN 104281577 B CN104281577 B CN 104281577B CN 201310273231 A CN201310273231 A CN 201310273231A CN 104281577 B CN104281577 B CN 104281577B
Authority
CN
China
Prior art keywords
keyword
ranking
keywords
data file
sequence algorithm
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.)
Active
Application number
CN201310273231.5A
Other languages
Chinese (zh)
Other versions
CN104281577A (en
Inventor
张国峰
朱逸斐
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Via Technologies Inc
Original Assignee
Via Technologies Inc
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Via Technologies Inc filed Critical Via Technologies Inc
Priority to CN201310273231.5A priority Critical patent/CN104281577B/en
Priority to TW102125770A priority patent/TWI610257B/en
Priority to US14/271,458 priority patent/US9558262B2/en
Publication of CN104281577A publication Critical patent/CN104281577A/en
Priority to US15/361,015 priority patent/US10083241B2/en
Application granted granted Critical
Publication of CN104281577B publication Critical patent/CN104281577B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]

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)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

The present invention provides a kind of sort method of data file, is suitable for electronic device, and sort method includes：Multiple keywords are captured from the content of multiple data files；Keyword ranking corresponding to multiple keywords is retrieved by Search engine；Search keyword categories corresponding to multiple keywords；And according to multiple keywords, the respective keyword ranking of multiple keywords and keyword categories and the respective current ranking of multiple data files, to generate the algorithm that sorts, wherein, prediction ranking of the sequence algorithm to calculate another data file, with this another data file that sorts.

Description

The sort method of data file
Technical field
The present invention relates to a kind of data processing methods, more particularly to a kind of sort method of data file.
Background technique
By the development of science and technology, network have become in modern life can not or scarce information obtain medium, especially The development of Internet news and universal, may replace paper not only to meet trend environmentally friendly now, can also in response to it is fast changing when Thing and update gio signal immediately.
The epoch of information explosion arrive, when huge Internet news quantity can also cause user to read and search for message Puzzlement.In order to allow user rapidly to capture important information, the dealer of Internet news usually carries out weight by human-edited Point news sequence, this measure is not only time-consuming and laborious, the sequence cis-position of highlight more likely by human-edited it is subjective because Element influences and loses its objectivity.
However, to carry out highlight by machine since news content itself has complicated data message The automation of cis-position sequence is not an easy thing.
Summary of the invention
The present invention provides a kind of sort method of data file, by the data file to known ranking results content into Row analysis, to produce the prediction model of ranking results, so that the burden for carrying out sorting data file by human-edited is reduced, or It is that can avoid subjective factor of the human-edited in sorting data file.
The present invention provides a kind of sort method of data file, is suitable for electronic device, and sort method includes：From multiple numbers According to capturing multiple keywords in the content of file；Keyword ranking corresponding to multiple keywords is retrieved by Search engine；It searches Keyword categories corresponding to the keyword of Zadoi；And according to multiple keywords, the respective keyword ranking of multiple keywords With keyword categories and the respective current ranking of multiple data files, to generate sequence algorithm, wherein sequence algorithm use To calculate the prediction ranking of another data file, with another data file that sorts.
The present invention provides a kind of sort method of data file, is suitable for electronic device, including：From the first data file At least one first keyword is captured in content；The row of keyword corresponding at least one first keyword is retrieved by Search engine Name；Search keyword categories corresponding at least one first keyword；And it will at least one first keyword, at least one first pass The respective keyword ranking of keyword and keyword categories input sequencing algorithm, to export the prediction ranking of the first data file, With first data file that sorts, wherein sequence algorithm is the content and multiple second numbers according to multiple second data files It is generated according to the respective current ranking of file.
Based on above-mentioned, the present invention is by keyword, keyword ranking, the keyword categories and more in multiple data files The known current ranking of a data file to generate sequence algorithm, and can calculate another data text using sequence algorithm The prediction ranking of part, with another data file that sorts.
To enable features described above and advantage of the invention to be clearer and more comprehensible, special embodiment below, and it is detailed in conjunction with attached drawing It is described as follows.
Detailed description of the invention
Fig. 1 is the electronic device of multiple embodiments according to the present invention and the block diagram of servomechanism.
Fig. 2 is the block diagram that device is notified according to the message of one embodiment of the invention.
Fig. 3 is the flow chart according to the message notification method of one embodiment of the invention.
Fig. 4 is the flow chart according to the message notification method of another embodiment of the present invention.
Fig. 5 is the flow chart of the display methods of the landmark data of an embodiment according to the present invention.
Fig. 6 is the flow chart of the display methods of landmark data according to another embodiment of the present invention.
Fig. 7 is the flow chart according to the display methods of the landmark data of another embodiment of the present invention.
Fig. 8 is the flow chart according to the display methods of the landmark data of another embodiment of the present invention.
Fig. 9 is the block diagram according to the region labelling apparatus of the data file of one embodiment of the invention.
Figure 10 is the flow chart according to the region labeling method of the data file of one embodiment of the invention.
Figure 11 is the schematic diagram according to the tree of one embodiment of the invention.
Figure 12 is the flow chart according to the region labeling method of the data file of another embodiment of the present invention.
Figure 13 A~Figure 13 D is the schematic diagram according to the construction process of the tree of one embodiment of the invention.
Figure 14 is the flow chart according to the sort method of the data file of one embodiment of the invention.
Figure 15 is the flow chart according to the sort method of the data file of one embodiment of the invention.
Accompanying drawings symbol description
101：Electronic device
103：Servomechanism
200：Message notifies device
210：Communication unit
230：Storage element
250：Broadcast unit
270：Gyroscope
290：Control unit
S310~S330：The step of message notification method
S401, S402, S310~S330：The step of message notification method
S510~S550, S541~S542, S710~S740, S731~S733：The step of display methods of landmark data
900：Region labelling apparatus
910：Taxon
930：Acquisition unit
950：Comparing unit
970：Marking unit
990：Store database
S1010~S1040：The step of region labeling method
S1011~S1013, S1021~S1022, S1031~S1033 and S1041~S1042：The step of region labeling method Suddenly
1301~1304：Second node
S1410~S1430, S1440, S1441, S1442 and S1450：The step of sort method of data file
Specific embodiment
Fig. 1 is according to the electronic device 101 of one embodiment of the invention and the block diagram of servomechanism 103.Servomechanism 103 It can be the computer or processor of personal computer, work station, host computer or various other types.Electronic device 101 can be notebook computer, tablet computer, personal digital assistant, smartphone or various other types just Portable electronic apparatus.Electronic device 101 can be communicated by network with servomechanism 103.It, will be with news in the narration of the present embodiment Breath notifies device 200 to represent the citing of electronic device 101.In other words, electronic device 101 and message notice device 200 can To be substantially equal same and interchangeable device.
When user is intended to notify device 200 to set the information category to be paid close attention to message, user can be first by news Breath notifies device 200 to issue solicited message.For example, " if there is newest Japanese nuclear accident news, tell at once I ", " such as Having advance versus decline in my self-selected stock of fruit is more than 2%, with regard to notifying me " either " is if the 36th phase colour film is left at once My " etc. is notified at once.In embodiments of the present invention, user can input solicited message to message by voice utterance Notify device 200.What message notice device 200 can differentiate solicited message by various types of natural language processing modules can It can be intended to, or can be further by being retrieved the structured database for storing a large amount of words to differentiate and be captured The attribute of crucial words out, is used after this solicited message is analyzed and understood, it can be seen that corresponding command condition, and will This command condition is sent to servomechanism 103.For example, command condition, which can be, " has newest Japanese nuclear accident news to produce Raw ", " designated speculative stock amount of increase be more than that 2% " either " the 36th phase colour film is announced the winners in a lottery ".Servomechanism 103 can be according to this order item Whether part has corresponding prompting message to inquire, for example, " content of newest Japan's nuclear accident news ", " designated speculative stock The 36th phase colour film of share price " either " prize-winning number ".In embodiments of the present invention, device 200 can be notified by message To be analyzed solicited message and be understood.In another embodiment of the invention, message notice device 200 can also believe request Breath is sent to servomechanism 103 to differentiate solicited message by various types of natural language processing modules in servomechanism 103 Possibility be intended to.It either, can be further by being retrieved the structured database for storing a large amount of words to differentiate The attribute of the crucial words captured is used and the crucial words captured is analyzed and understood.Servomechanism 103 is being looked into It askes after whether having the correspondence prompting message for meeting command condition, if judgement has the correlated condition for having and meeting order, news Corresponding prompting message then can be downloaded and be stored by breath notice device 200, and the person of connecing can be played back.In the embodiment of the present invention Concept in, due to being by its information category to be paid close attention to set by servomechanism 103 periodically (or immediately) record user Up-to-date information, use and inquire corresponding prompting message when receiving command condition, accordingly, with respect to by message notify fill For setting the up-to-date information that 200 periodically (or immediately) record its information category to be paid close attention to set by user, the present invention can Further decrease the power consumption and workload of message notice device 200.
In embodiments of the present invention, servomechanism 103, which can be inquired, meets the correspondence of command condition in specified time interval and mentions Show message.This specified time interval can be certain interval of time of user's setting, be also possible to leave news in user The time interval of breath notice device 200 therebetween.For example, user the factors such as can go out because having a bath or having forgotten band, and will news Breath notice device 200 stays on the table；After it have passed through specified time interval, user returns to table side once again and has picked up message again Notify device 200.At this point, due in leave message notice device 200 information category that may have user of interest therebetween Latest news occurs, therefore message notice device 200 can correspond to the initial time of this specified time interval and terminate the time, will Its corresponding prompting message is downloaded and is stored from servomechanism 103, and is then played back to remind user.It is following will to this into One step is described in detail.
Fig. 2 is the block diagram that device is notified according to the message of one embodiment of the invention.As shown in Fig. 2, message notice dress Setting 200 includes communication unit 210, storage element 230, broadcast unit 250, gyroscope 270 and control unit 290.Control is single Member 290 is coupled to communication unit 210, storage element 230, broadcast unit 250 and gyroscope 270.Communication unit 210 to Servomechanism 103 communicates, and storage element 230 is to store data, and broadcast unit 250 is to play message, and gyroscope 270 is to examine Survey the angular speed of message notice device 200.Communication unit 210 can be wireless communications chips or module or other with net The chip or module of network line function.Storage element 230 can be various types of data storage media.Broadcast unit 250 can To be various types of data replay apparatus, such as loudspeaker, display or other data output devices.Control unit 290 can be various types of functional modules, chip or microprocessor.Fig. 3 is logical according to the message of one embodiment of the invention The flow chart of perception method.As shown in figure 3, message notification method according to an embodiment of the present invention includes step S310~S330.Please Referring concurrently to Fig. 2 and Fig. 3.
In step s310, control unit 290 judges whether that having message notice device 200 initially enters stationary state First time point (i.e. above-mentioned initial time) and the second time point (i.e. above-mentioned termination time) for terminating stationary state.It lifts For example, user the factors such as can go out because having a bath or having forgotten band, and message notice device 200 is stayed on the table, thus message Device 200 is notified to enter stationary state in first time point, at this point, first time point can be recorded in storage by control unit 290 In memory cell 230.After having crossed specified time interval, user returns to table side once again again and has picked up message notice device 200, because And message notifies that device 200 terminates stationary state at the second time point, at this point, control unit 290 also can be by the second time point It is recorded in storage element 230.Control unit 290 can be by inquiry storage element 230 to determine whether having first time point With the second time point.
In step s 320, if control unit 290 judges that it has first time point and the second time point, servomechanism Whether 103 inquiries have first time point at least prompting message between the second time point.For example, if control unit 290 judge that it has first time point and the second time point, this represents user and notifies device 200 possibly off message For a period of time, the first time point learnt is sent to servomechanism 103 with the second time point and looks by control unit 290 at this time It askes, whether judgement during this period of time has prompting message generation.For example, if user has missed call or has in this period Message is not read, then its prompting message can be " you have a logical missed call " or " first you, which have, does not read message ".The prompt Message can also be the concern information of user's setting, for example, hot news, stock or lottery ticket etc..
In step S330, if servomechanism 103 has an at least prompting message, message notifies device 200 to download this extremely A few prompting message is simultaneously stored to storage element 230, and this at least prompting message is played by broadcast unit 250.Citing For, if user has missed call or has not between the first time point that user is left and the second time point Message is read, or has and produces the concern information of stock, lottery ticket or hot news, then message notice device 200 can will be prompted to Message " you have a logical missed call " or " first you, which have, does not read message ", or corresponding generated stock, lottery ticket or heat The specifying information of point news is downloaded and is stored to storage element 230, and is played by broadcast unit 250.Implement in the present invention In example, the mode of 250 play cuing message of broadcast unit, which can be, plays text or playing video, without restriction herein.
Fig. 4 is the flow chart according to the message notification method of another embodiment of the present invention.As shown in figure 4, according to this hair The message notification method of bright embodiment includes step S401, S402, S410, S420 and S430.Referring to Fig. 2 and Fig. 4.
In step S401, user can notify device 200 to receive solicited message by message.For example, user Solicited message can be " if there is newest Japanese nuclear accident news, tell at once I ", " if had in my self-selected stock strand Ticket ups and downs are more than 2%, with regard to notifying me " either " if the 36th phase colour film has been left to notify me at once at once ".Institute as above It states, in embodiments of the present invention, user can input solicited message to message by voice utterance and notify device 200.In another embodiment of the invention, user can also ask by specific software interface or various other ways to input Information is sought, it is without restriction herein.
In step S402, at least one crucial words in solicited message is captured, to differentiate the command condition of solicited message And given threshold.In embodiments of the present invention, can be captured by control unit 290 the crucial words in solicited message with into Row analysis and understanding.In another embodiment of the invention, solicited message can be also transmitted to servomechanism 103, then in servo The crucial words in solicited message is captured in device 103 to be analyzed and be understood.Crucial words can be the concern letter of user The type of breath, the words to express order or other predefined words that can be used to analyze and understand.Implement in the present invention In example, control unit 290 can differentiate that the possibility of solicited message is intended to by various types of natural language processing modules, or Being can be further by retrieving to differentiate captured keyword the structured database for storing a large amount of words The attribute of word is used and the crucial words captured is analyzed and understood, in the hope of the corresponding order in solicited message Condition, such as, if there is newest Japanese nuclear accident news to generate, whether designated speculative stock amount of increase is more than 2% or the 30th Whether six phase colour films announce the winners in a lottery.In addition, control unit 290 also can by crucial words analysis and understand result come acquire threshold value with It is used in step S410.
In step S410, control unit 290 notifies the entrance of device 200 static according to threshold value to determine whether having message First time point after state, and terminate the second time point of stationary state.Threshold value can be time threshold or angular speed Threshold value.In embodiments of the present invention, control unit 290 can detect the angle speed of message notice device 200 by gyroscope 270 Degree, to differentiate that message notifies device 200 whether to enter stationary state and whether terminates stationary state, to obtain at the first time Point and the second time point.For example, when the angular speed of message notice device 200 is less than angular speed threshold value, then control unit 290 Can determine that message notice device 200 enter stationary state, and when message notice device 200 angular speed be less than angular speed threshold value and Its duration is more than time threshold, then control unit 290 can determine that user has had left message notice device 200, To which this time point is set as first time point.If again after a period of time, when the angular speed of message notice device 200 is big When angular speed threshold value, then control unit 290 can determine that message notice device 200 terminates stationary state and (enters Moving condition), in other words, control unit 290 can determine that user has picked up message notice device 200 once again, thus will This time point was set as the second time point.As described above, first time point and the second time point can record in storage element 230 In.In embodiments of the present invention, if message notice device 200 itself has vibrating mode (for example, having incoming call or having news in brief When, then message notice device 200 can enter vibrating mode), then angular speed threshold value can be greater than message notice device 200 in vibration mould Angular speed caused by under formula, thereby, then the message under vibrating mode just will not be notified device 200 to sentence by control unit 290 It is set to and enters moving condition.In another embodiment of the invention, control unit 290 can notify device by detection message Whether 200 enter whether suspend mode either receives touch-control input signal by detection message notice device 200, to judge to interrogate Whether breath notice device 200 enters or terminates stationary state.
In the step s 420, if control unit 290 judges that it has first time point and the second time point, servomechanism 103 inquire whether it has first time point at least prompting message for meeting command condition between the second time point.Citing For, if having occurred in specified time interval between first time point and the second time point, " newest Japan's nuclear accident is new It hears and generates ", " designated speculative stock amount of increase be more than that 2% " either " the 36th phase colour film is announced the winners in a lottery ", then servomechanism 103 can have and mention Show message.In embodiments of the present invention, prompting message can be the interior of the latest news of corresponding user information category of interest Hold itself, for example, " the report content of newest Japan's nuclear accident news ", " share price of designated speculative stock " the either the " the 36th Phase colour film prize-winning number ".In another embodiment of the invention, prompting message can also be to prompt user to generate There is the message of the latest news content of its information category of interest, for example, " you is reminded, existing newest Japanese nuclear accident news ", " Dear user, your XXX stock " either " the 36th phase colour film that risen sharply have been announced the winners in a lottery Lei ".
In step S430, if servomechanism 103 has an at least prompting message, message notifies device 200 to download this extremely A few prompting message is simultaneously stored to storage element 230, and this corresponding at least prompting message is played by broadcast unit 250 Voice.For example, if between the first time point that user is left and the second time point, control unit 290 passes through Servomechanism 103 judges that it has prompting message " the report content of newest Japan's nuclear accident news ", " stock of designated speculative stock The 36th phase colour film of valence " either " prize-winning number ", then this prompting message is downloaded and is stored to storage element 230, and It is played by broadcast unit 250.In embodiments of the present invention, broadcast unit 250 can will be prompted to the content of message with voice side Formula plays back.
In conclusion the present invention can receive the solicited message that user is inputted with voice mode, and capture its crucial words Differentiate that command condition and given threshold are used, device is notified therebetween in user's leave message, it can be according to this threshold value （Time threshold or angular speed threshold value）To detect first time point and the end that message notice device initially enters stationary state Only the second time point of stationary state, and then when user brings back message notice device, satisfaction life can be inquired in servomechanism The prompting message of condition is enabled, and will be prompted to message downloading and store to message notify device, by voice mode play cuing Message is to remind user, to reduce its probability for ignoring important message.
It is following from another technological standpoint and to realize the embodiment of this technological standpoint, come describe electronic device 101 with And the running between servomechanism 103.As shown in Figure 1, in another embodiment of the invention, electronic device 101 and servomechanism 103 also can be used for executing the display methods of landmark data.
In embodiments of the present invention, it when user executes the search of specific landmark by electronic device 101, can input Place name keyword, and the place name keyword inputted can be sent in servomechanism 103 by electronic device 101, servomechanism 103 is right This place name keyword executes search, and is then ranked up to the landmark data searched, so as to may meet use The landmark data of the searching demand of person sequence cis-position with higher.Finally, servomechanism again passes the ranking results of landmark data Electronic device 101 is sent back to show, user is just able to therefrom look for most beneficial message.Then, in embodiments of the present invention, Electronic device 101 can be shown the respective landmark names of the landmark data in ranking results by map application Come.In another embodiment of the invention, electronic device 101 itself is with to the function of searching of specific place name and to being searched The function that the landmark data sought is ranked up, therefore just no longer need to can be to user by servomechanism 103 for electronic device 101 Show the ranking results of landmark data.
Landmark data can have specific characterized parameter.For example, different landmark datas, which can have, different knows Name degree.Therefore, corresponding cis-position relationship can be generated during the sequence of landmark data.Therefore, in the embodiment of the present invention In, before user inputs search of the place name keyword to execute specific landmark by electronic device 101, servomechanism 103 can The calculating of popularity is carried out to existing landmark data.It is following this to be described in detail.
Fig. 5 is the flow chart of the display methods of the landmark data of an embodiment according to the present invention.Table 1 is according to the present invention An embodiment landmark data schematic diagram.As shown in figure 5, the display methods of landmark data according to an embodiment of the present invention Step includes S510~S550.As shown in table 1, landmark data can have landmark names, objective hierarchical category, address, guide With number and the corresponding calculated popularity of institute.It is following referring to Fig. 5 and table 1.
[table 1]
In step S510, multiple landmark datas are obtained.For example, servomechanism 103 can be by database or search Engine obtains multiple landmark datas, without restriction herein.Acquired multiple landmark datas can be stored in specific storage Medium is as landmark data library.
In step S520, the address reference number of the respective address of multiple landmark datas on the internet is counted.Citing For, servomechanism 103 can count address " the Pudong New Area, Shanghai century avenue 1 of " Oriental Pearl " by Search engine Number " on the internet altogether be cited 852318 times.Due to randomness with height of landmark names itself, if not making It is the reference using the landmark names of landmark data instead with address reference number as the characterized parameter of corresponding popularity If number, it would be possible to generate great error.It for example, is " Nanjing Xuanwu District Beijing East Road 31 calculating address Number 9 floor of industrial art building " businessman " Oriental Pearl " popularity characterized parameter when, if selection uses landmark names " Oriental Pearl " reference number, then great error will be generated because of the presence of sight spot " Oriental Pearl ".Conversely, as ground Address corresponding to mark data is usually unique, thus the characterization with address reference number as popularity is joined in this step Number is quite objective standard.
In step S530, the respective objective hierarchical category of multiple landmark datas is searched.In embodiments of the present invention, objective Hierarchical category can be generally acknowledged sight spot and (for example, 1A grades~5A grades) or generally acknowledged hotel owner such as comment to comment etc. (for example, a star~six stars Grade).For example, servomechanism 103 hunts out the sight spot that " Shanghai Wild Life Park " is " 3A grades ", and " Falls at Hukou of the Yellow River " are " 4A grades " Sight spot, " Falls at Hukou of the Yellow River hotel " is the hotel owner of " three-star ", and " South Beauty food and drink " is the hotel owner of " two stars ".At this In inventive embodiments, objective hierarchical category is also possible to class divide attribute.For example, " South Beauty industry " and " pretty good People's medium " is all " businessman " scale, and " Beijing Hua Lian comprehensive supermarket " is then the scale in " market "；" 217 " road is to belong to " National highway ", " 373 " are to belong to " provincial highway ", and " 048 " is to belong to " county road ".Above-mentioned objective hierarchical category can have a variety of different Other objective definitions, it is without restriction herein.
In step S540, calculated according to the respective objective hierarchical category of multiple landmark datas and address reference number Multiple respective popularity of landmark data.According to embodiments of the present invention, address reference number is higher, then servomechanism 103 is calculated The popularity of its corresponding landmark data can be higher out.For example, " South Beauty Art Design " and " the pretty river of businessman are similarly Southern industry " address reference number be respectively " 293 " with " 531 ", therefore " South Beauty industry " corresponding calculated popularity " South Beauty Art Design " can be greater than.According to embodiments of the present invention, the rank of objective hierarchical category is higher, then 103 institute of servomechanism The popularity for calculating its corresponding landmark data can be higher.For example, address is similarly " Fuwai Avenue, Xicheng District, Beijing No. 1 Sichuan mansion The East Pagoda building 5 floor 515 " and address reference number is similarly " 5236 " " the comprehensive supermarket's " with " in Beijing Hua Lian McDonald ", since " the comprehensive supermarket in Beijing Hua Lian " is " market ", and " McDonald " is " businessman " in this market, therefore " Beijing Hua Lian integrates supermarket " corresponding calculated popularity can be greater than " McDonald ".Similarly, if road, then " 217 national highway " Popularity can be greater than " 373 provincial highway " and " 048 county road ".
In step S550, according to the respective popularity of multiple landmark datas, multiple landmark datas are shown in electronics dress Set 101.For example, after the completion of servomechanism 103 calculates landmark data and corresponding popularity, result can be sent back Electronic device 101, then electronic device 101 can show landmark data further according to the sequence of popularity.
According to another embodiment of the present invention, step S510~S550 can be all implemented in electronic device 101 or step Part steps in S510~S550 can be performed in electronic device 101, and other parts step can be performed in servomechanism 103 In, the two can reach communication and coordination to each other by internet therebetween, without restriction herein.
Fig. 6 is the flow chart of the display methods of the landmark data of an embodiment according to the present invention.As shown in fig. 6, according to The step of display methods of the landmark data of the embodiment of the present invention includes S510~S530, S541, S542 and S550.It is following by needle To explaining with above-mentioned difference.Under be listed in calculate landmark data popularity when, can further execute step S541 and S542。
In step S541, the respective objective hierarchical category of multiple landmark datas and address reference number are converted to pair The conversion value answered.For example, in an embodiment of the present invention, if objective hierarchical category is to generally acknowledge that sight spot is commented, " 1A Grade ", " 2A grades ", " 3A grades ", " 4A grades " and " 5A grades " corresponding conversion value can be respectively 20,40,60,80 and 100；If objective Hierarchical category is to generally acknowledge that hotel owner comments, then " star ", " two stars ", " three-star ", " four-star ", " five-star " and " six stars " corresponding conversion value can be respectively 20,40,60,80,100 and 120 to grade.If objective hierarchical category is class divide attribute, Then " businessman " and " market " corresponding conversion value can be respectively 20 and 80；" national highway ", " provincial highway " and " county road " corresponding conversion Value can be respectively 30,60 and 90.In an embodiment of the present invention, address reference number be converted to corresponding conversion value can be by Operation is done in natural logrithm function (ln x) × 10.For example, the address reference number at " 4A grades " sight spot " Oriental Pearl " is 852318, then its corresponding conversion value can be (ln852318) × 10=136.56；If zero-address data, corresponding turn It shifts to and can be 0.The calculation of above-mentioned respective value and respective value can be adjusted and be changed according to various situations, herein not It limits.
In step S542, according to objective hierarchical category and the corresponding conversion value of address reference number and weighted value, To calculate the popularity of landmark data.In an embodiment of the present invention, the corresponding weighted value of objective hierarchical category can be 0.4, The corresponding weighted value of address reference number can be 0.6, and the calculation formula of the popularity of landmark data can be：(objective level The conversion value of classification) × 0.4+ (conversion value of address reference number) × 0.6.For example, " 4A grades " sight spot " Oriental Pearl " Popularity be (80) × 0.4+ ((ln852318) × 10) × 0.6=113.94, " South Beauty industry " of " businessman " scale Popularity is (20) × 0.4+ ((ln531) × 10) × 0.6=45.66.
As described above, after the landmark data library construction in electronic device 101 or servomechanism 103 is completed, user The search of i.e. executable specific landmark.It is following this to be described in detail.
Fig. 7 is the flow chart according to the display methods of the landmark data of one embodiment of the invention.As shown in fig. 7, according to The display methods of the landmark data of the embodiment of the present invention includes step S710~S740.
In step S710, place name keyword is received.It for example, can be by when user's specific landmark to be searched Place name keyword is inputted with voice or manual mode by electronic device 101.
In step S720, at least landmark data for whether having corresponding place name keyword searched.For example, electric at this time Sub-device 101 can search whether have relevant landmark data by built-in landmark data library, or by place name keyword Whether be sent to servomechanism 103 has relevant landmark data to search.If searching, step S730 is then executed, if not It searches, then continues waiting for receiving another place name keyword.
It is respective well-known according to an at least landmark data if searching an at least landmark data in step S730 Degree, matching degree and apart from score, come at least landmark data that sorts.For example, when searching related landmark data, by May be very more in related terrestrial reference data bulk, therefore in order to be practised close to user for the general sense organ of terrestrial reference or cognition It is used, it can sort by characterized parameter corresponding to related landmark data, inquire spent mental and physical efforts to save user. In embodiments of the present invention, characterized parameter corresponding to related landmark data (is relevant to objective layer in addition to above-mentioned popularity Grade classification and the address reference number counted on the internet), can also further there be the matching degree of related landmark data (for example, matching degree of text) and apart from score (for example, distance degree of terrestrial reference and user).However, at this In another embodiment of invention, characterized parameter corresponding to landmark data can be popularity, matching degree and in score One of person, it is without restriction herein.
In step S740, at least landmark data to be sorted is shown in electronic device 101.At this point, user can By electronic device 101 come in the landmark data after multiple sequences for being relevant to inputted place name keyword, inquiry most has The landmark data of benefit.
Fig. 8 is the flow chart according to the display methods of the landmark data of another embodiment of the present invention.As shown in figure 8, root Display methods according to the landmark data of the embodiment of the present invention includes step S710, S720, S731~S733 and S740.It is following by needle To explaining with above-mentioned difference.Under when being listed in the landmark data for the hunted out correspondence place name keyword of sorting, can more into One step executes step S731~S733.
In step S731, according to the respective landmark names of an at least landmark data and place name keyword, to calculate The respective matching degree of an at least landmark data.In other words, i.e., according to the landmark data hunted out to the symbol of place name keyword Conjunction degree calculates its matching degree.For example, if the place name keyword of user's input is " South Beauty ", " South Beauty meal The industry of drink ", " South Beauty Art Design " and " South Beauty " matching degree, can all be higher than the matching degree of " pretty woman's medium ".
In step S732, according at least position of a landmark data respective positions and electronic device 101, to calculate At least a landmark data is respective apart from score out.In other words, i.e., according to the landmark data hunted out to electronic device 101 Relative position calculate it apart from score.For example, if user is located at Beijing and inputs place name to electronic device 101 Keyword " South Beauty ", then positioned at Beijing " South Beauty food and drink " and " South Beauty Art Design " apart from score, all can be high In " the South Beauty industry " for being located at Suzhou City apart from score.
In step S733, according to popularity, matching degree and apart from score and its corresponding weighted value, to sort at least One landmark data.For example, electronic device 101 can define popularity, matching degree and according to different requirements, apart from score Corresponding weighted value is used and determines popularity, matching degree and the influence degree apart from score respectively for ranking results.
Thereby, the landmark data after characterizing can allow user when inquiring specific landmark, the search result of display It is to sort according to the general sense organ or cognition habit of user, to save the query time of user.
In conclusion the present invention is according to corresponding to the address reference number of multiple landmark datas and objective hierarchical category Conversion value calculates its popularity with weighted value, after receiving place name keyword, according to the related landmark data hunted out Its matching degree is calculated the matching degree of place name keyword, according to the related landmark data hunted out to the phase of electronic device It is calculated apart from score to position, and is searched relatively by popularity, matching degree and apart from score to sort Data are marked, the related landmark data after sequence is shown in electronic device.
It is following will be from another technological standpoint, to describe the running between electronic device 101 and servomechanism 103.Such as Fig. 1 Shown, in another embodiment of the invention, electronic device 101 and servomechanism 103 also can be used for executing the ground of data file Field mark method.In addition, the act of electronic device 101 will be represented in the narration of the present embodiment with region labelling apparatus 900 Example.In other words, electronic device 101 and region labelling apparatus 900 can be to be substantially equal same and interchangeable device.
User can be communicated by network with servomechanism 103 by region labelling apparatus 900, use acquirement data file Or the reference information to flag data file.In embodiments of the present invention, data file can be Internet news.Citing comes It says, firstly, carrying out regional classification to Internet news, then after Internet news editor obtains a large amount of Internet news Regional label can be carried out to Internet news.In embodiments of the present invention, Internet news editor can fill by ground field mark 900 are set to obtain the reference information of regional title from servomechanism 103 to build up specific tree, is used as analysis net The contents attribute of network news and marking to it is used.In another embodiment of the invention, Internet news editor also can be by The specific tree that construction is completed directly is obtained from servomechanism 103 by region labelling apparatus 900.In this tree Each node represents each specific regional title, and passes through this tree it can be seen that each specific regional title all ranks thereon The administrative area title of layer.Regional title may include administrative area title and significant title, and section belonging to significant title Point can be most end stratum in tree, wherein significant title can be place name or sight name or any have Regional name, public organization's title or other titles etc., it is without restriction herein.Thereby, for example, can learn In tree area belonging to any sight spot or public organization why (i.e. its each father node).Then, region labelling apparatus 900 Can analyze whether each Internet news has regional content (keyword of such as place name), if this regional content may conform to it is tree-shaped Any node in structure, then region labelling apparatus 900 can be by the node that is met, its Internet news to be marked. In other words, Internet news editor can be such that the Internet news with regional content has by region labelling apparatus 900 Corresponding regional Characteristics, to complete the label or regional classification to each Internet news, for example, some Internet news is to belong to Classification in which area.It is following this to be further illustrated.
Fig. 9 is the block diagram according to the region labelling apparatus of the data file of one embodiment of the invention.As shown in figure 9, Region labelling apparatus 900 includes taxon 910, acquisition unit 930, comparing unit 950, marking unit 970 and storage number According to library 990.Comparing unit 950 is coupled to acquisition unit 930, and marking unit 970 is coupled to comparing unit 950, stores database 990 are coupled to taxon 910, acquisition unit 930, comparing unit 950 and marking unit 970.Taxon 910 captures list Member 930, comparing unit 950, marking unit 970 can be various forms of functional modules or microprocessor, store database 990 It can be various forms of storage medium.Figure 10 is the region labeling method according to the data file of one embodiment of the invention Flow chart.As shown in Figure 10, region labeling method according to an embodiment of the present invention includes step S1010~S1040.Figure 11 is root According to the schematic diagram of the tree of one embodiment of the invention.It is following referring to Fig. 9, Figure 10 and Figure 11.
In step S1010, taxon 910 can obtain tree by network.In embodiments of the present invention, this sets Shape structure can have multiple nodes, this multiple node may include multiple administrative area titles and significant title, and this multiple row There can be social strata relation between administrative division title and significant title, in addition, taxon 910 can store up tree obtained It is stored in storage database 990.For example, as shown in figure 11, each node of this tree may include each stratum of China Administrative area title and the sight name in it, and can be corresponded between its administrative area or sight spot between the upper lower node of tree Social strata relation, for example, may include its each province or municipality directly under the Central Government (such as Shanghai, Jiangsu Province and Anhui Province under the node " China " Deng) child node, may include the son of its each subregion (such as Pudong New District, Huangpu District and Jing'an District) under the node " Shanghai " Node, may include its each sight spot (such as Century Park, Oriental Pearl and Jin Mao Tower) under the node " Pudong New District " Child node.In addition, as described above, significant title can be with regional name and public organization's title, such as Figure 11 institute Show, node " Shanghai " can further include the child node of its professional basketball team " Shanghai shark " and the son of its mayor " Yang Xiong " Node.
In step S1020, acquisition unit 930 can receive data file by network, and capture at least from data file One keyword.For example, acquisition unit 930 can receive a large amount of Internet news from servomechanism 103 by network and store To storage database 990, and may include various regional keywords in the content of received Internet news, such as " Jiangsu Save " or " Oriental Pearl " etc., acquisition unit 930 can analyze its content to capture this keyword.
In step S1030, comparing unit 950 can compare an at least keyword and multiple nodes, to find out and at least one The first node of keyword match.It for example, has included the row of known each stratum of China in above-mentioned tree Administrative division title and place name or sight name in it, if capturing keyword " Oriental Pearl " from the content of Internet news, Then tree can then be searched, and have found and be all the first node of " Oriental Pearl ".This represents this keyword " " affiliated Internet news is to have regional feature and be classifiable for this tree for Oriental Pearl.At this In inventive embodiments, comparing unit 950 can find the first node met by the algorithm of various trees, herein not It limits.
In step S1040, marking unit 970 can mark a first node at least father node relevant to first node in Data file.For example, if can search from tree and meet the of the keyword in Internet news " Oriental Pearl " One node, then its relevant father node is " Pudong New District ", " Shanghai " and " China ".Therefore, belonging to keyword " Oriental Pearl " Internet news other than it can be marked with first node " Oriental Pearl ", can also be marked with each stratum on " Oriental Pearl " Each father node " Pudong New District ", " Shanghai " and " China " on administrative area, that is, first node " Oriental Pearl ".
Figure 12 is the flow chart according to the region labeling method of the data file of another embodiment of the present invention.Such as Figure 12 institute Show, region labeling method according to an embodiment of the present invention include step S1011~S1013, S1021~S1022, S1031~ S1033 and S1041~S1042.Figure 13 A~Figure 13 D is the construction process according to the tree of one embodiment of the invention Schematic diagram.
In step S1011, taxon 910 obtains regional title.For example, taxon 910 can pass through net Network obtains the reference information at each stratum administrative area in relation to China and the sight spot in it from servomechanism 103, this reference information can It is without restriction herein being presented with various taxons 910 cognizable formats.Taxon 910 can be according to this ginseng Information is examined to obtain each regional title one by one, as shown in FIG. 13A, when having node " China " in tree, grouping sheet Member 910 achieves administrative area title " Shanghai ".
In step S1012, taxon 910 judges the second section whether regional title is subordinated in tree Point.If taxon 910 is judged as YES, step S1013 can be then executed.Second node can be acquired regional name Claim the node in the affiliated area of minimum stratum in tree.For example, as shown in FIG. 13A, when having had in tree When having node " China ", taxon 910 achieves administrative area title " Shanghai ", and taxon 910 can determine whether administrative area at this time The second node of title " Shanghai " institute subordinate can be the dotted line node 1301 in Figure 13 A.
In step S1013, tree is added in regional title by taxon 910.For example, such as Figure 13 A institute Show, then taxon 910 can correspond to acquired administrative area title by construction egress " Shanghai " in tree.
In step S1014, taxon 910 judges tree whether complete by construction.If tree construction is completed, Then follow the steps S1021.If the non-construction of tree is completed, above-mentioned steps S1011~step S1013 can be recycled constantly It executes.For example, taxon 910 can determine whether out second node belonging to administrative area title " Pudong New District " in Figure 13 B It for dotted line node 1302 and is added into, taxon 910 can determine whether out belonging to sight name " Century Park " in Figure 13 C Second node is dotted line node 1303 and to be added into, and taxon 910 can determine whether out sight name " east is bright in Figure 13 D " affiliated second node is dotted line node 1304 and is added into that the above process will be repeated continuously, until taxon to pearl 910, by the information in each stratum administrative area obtained in relation to China and the sight spot in it, are seriatim construed as tree In each node, as shown in figure 11.The construction process of above-mentioned tree can come by the related algorithm of various trees Reach, it is without restriction herein.As described above, as shown in figure 11, after tree construction is completed, each section of tree The administrative area title and the sight name in it that point includes each stratum of China, and can between the upper lower node of tree Social strata relation between its corresponding administrative area or sight spot.
In step S1021, acquisition unit 930 captures an at least keyword by the title or text of data file.Citing For, since the content of Internet news may include the content of title and the content of text, acquisition unit 930 can be from title Content and the content of text capture keyword, use the foundation as the regional Characteristics for differentiating its Internet news.
In step S1022, acquisition unit 930 by data file a source capturing at least keyword.Of the invention real It applies in example, the source of data file may include the site of data file relevant scene and data file provider. For example, it since the content of the title of Internet news and text may not include any regional keyword, picks Take unit 930 further can capture keyword from the relevant scene of Internet news, for example, if Internet news is hair It is distributed in the field of " Huangpu District local items " of particular web portal, then acquisition unit 930, which can be used, captures keyword " yellow Pu area ", either, acquisition unit 930 can further capture keyword, example from the site of Internet news supplier Such as, if Internet news is issued by newspaper office's " Shanghai Daily ", acquisition unit 930, which can be used, captures keyword " Shanghai ", Either, if Internet news is to be issued by newspaper office " Xinmin Evening News ", and the location of newspaper office " Xinmin Evening News " is in Shanghai, therefore Acquisition unit 930 still can be by tabling look-up or according to relevant information, to obtain keyword " Shanghai ".
In step S1031, comparing unit 950 is respectively to the corresponding weight of an at least keyword definition.This weight can generation Influence degree of its correspondence keyword of table for the regional Characteristics of affiliated data file.In other words, corresponding to keyword Weight it is higher, then comparing unit 950 with this keyword come as differentiate belonging to data file regional Characteristics a possibility that It is higher.For example, as described above, keyword corresponding to Internet news can from the text of Internet news, title or It is obtained according to the site of its relevant scene and its supplier, and keyword acquired by different sources can correspond to Different weight, for example, if the weight of the keyword captured according to the scene of Internet news is A, from Internet news The weight of keyword that is captured of title be B, the weight of the keyword captured from the text of Internet news is C, root The weight of the keyword captured according to the site of the supplier of Internet news is D, then its relativeness can be A>B>C> D.However, the relativeness of above-mentioned weight can have other arrangements and change, it is without restriction herein.
In step S1032, comparing unit 950 searches tree, has first to compare to whether there is in tree Node, and administrative area title or significant title included by its first node and at least one of keyword is identical. If comparing unit 950 judges that there are first nodes in tree, then execute step S1033.Such as in step S1031 It is described, by calculated weight can represent its and correspond to influence of the keyword for the regional Characteristics of affiliated data file Degree, thus in embodiments of the present invention, comparing unit 950 can come according still further to the corresponding weight of keyword as than Cis-position reference to keyword and node.For example, as described above, consolidated network news can have simultaneously according to Internet news The scene keyword captured and the keyword captured from the text of Internet news, at this point, due to basis The weight for the keyword that the scene of Internet news is captured can be greater than the key captured from the text of Internet news The weight of word, therefore, comparing unit 950 will preferentially use the keyword captured according to the scene of Internet news, by To search tree.The person of connecing, the first node that comparing unit 950 can be found by the search algorithm method of tree, and The keyword that administrative area title or significant title included by this first node are searched with preferentially using is identical.In this hair In bright embodiment, the search algorithm method of above-mentioned tree can be reached by various applications, without restriction herein.
In step S1033, comparing unit 950 finds out an at least father node relevant to first node in tree. Its each father can be looked for from each upper stratum of first node by stratum's characteristic, comparing unit 950 possessed by tree itself Node.For example, as shown in figure 11, if first node is " Oriental Pearl ", relevant father node is that " Pudong is new Area ", " Shanghai " and " China ".
In step S1041, marking unit 970 is built according to a first node at least father node relevant to first node Found corresponding multiple labels.For example, when comparing unit 950 have found include keyword " Oriental Pearl " Internet news First node, then marking unit 970 in addition to can by " Oriental Pearl " set thus multiple labels of Internet news wherein it Outside one, can a more step " Pudong New District ", " Shanghai " and " China " is also set to the label of this Internet news.Establish the side of label Formula can be records its correspond to first node and related father node title, or capture its correspondence first node to it is related The connection of father node, it is without restriction herein.
In step S1042, marking unit 970 links multiple labels and data file to complete to mark and store data text Part is in storage database 990.For example, when " east is bright for each label for the Internet news for including keyword " Oriental Pearl " Pearl ", " Pudong New District ", " Shanghai " and " " after the completion of all establishing, these labels are then linked to corresponding by marking unit 970 for China Internet news.The mode of connection label can be increases its title for corresponding to first node and related father node in Internet news Content, or add in Internet news its connection for corresponding to first node and related father node, it is without restriction herein.
In conclusion the present invention is built up by the second node of corresponding regional title is added one by one with multiple sections The tree of point, so as to there is social strata relation between administrative area title and significant title included by multiple nodes, and According to the title content of data file, body matter, related scene and the site of data file supplier, to obtain Regional keyword is obtained, and is defining weight corresponding to each keyword as the cis-position for comparing keyword and tree With reference to later, finding out, matched first node is marked on corresponding data file with its father node, so that data file has There are corresponding regional Characteristics.
It is following will be from another technological standpoint, to describe the running between electronic device 101 and servomechanism 103.Such as Fig. 1 Shown, in another embodiment of the invention, electronic device 101 and servomechanism 103 also can be used for executing the row of data file Sequence method.
It in embodiments of the present invention, can will be unknown when the data file in electronic device 101 with unknown current ranking The data file of ranking is uploaded to servomechanism 103 to do the analysis of content to it at present, then generates by sequence algorithm It the prediction ranking of the data file of unknown current ranking and sorts, finally again sends back its result in electronic device 101 out.? In the embodiment of the present invention, if the prediction ranking of the data file of unknown current ranking is before 100, this data file is attached most importance to It wants, if the prediction ranking of the data file of unknown current ranking, after 100, this data file is inessential.In this hair In bright embodiment, before the data file that servomechanism 103 receives unknown current ranking, servomechanism 103 can be by multiple known The data file of ranking generates sequence algorithm at present.In another embodiment of the invention, electronic device 101 itself can be by Sequence algorithm is generated by the data file of multiple known current rankings, therefore, electronic device 101 does not need to pass through servomechanism 103 can obtain the prediction ranking of the data file of unknown current ranking.It is following to will be described generation sequence algorithm and generation The details of the prediction ranking of data file.
Figure 14 is the flow chart according to the sort method of the data file of one embodiment of the invention.As shown in figure 14, root According to the data file of the embodiment of the present invention sort method the step of include S1410~S1450.Table 2 is to be implemented according to the present invention The schematic diagram of the data file of the known current ranking of example.Table 3 is the data according to the unknown current ranking of the embodiment of the present invention The schematic diagram of file.In embodiments of the present invention, data file can be news file.As shown in table 2 and table 3, data file Content may include title content and body matter.It is following referring to Figure 14, table 2 and table 3.
[table 2]
[table 3]
In step S1410, multiple keywords are captured from the content of multiple data files.It for example, can be in data Respective keyword is captured in the content of file 1~4.For example, fechtable goes out key in the title content of data file 1 Word " two Conferences ", Yi Ji, in the body matter of data file 1 fechtable go out keyword " National People's Congress ", " CPPCC ", " Xi Jinping ", " Hu Jintao " and " two sides ".
In step S1420, keyword ranking corresponding to multiple keywords is retrieved by Search engine.For example, By Search engine data file 1 keyword " two Conferences ", " National People's Congress ", " CPPCC ", " Xi Jinping ", " Hu Jintao " and " two sides " Corresponding keyword ranking may respectively be " 152 ", " 96 ", " 135 ", " 33 ", " 47 " and " 95 ".In embodiments of the present invention, Keyword ranking can be the same day inquired by google Search engine, when week or of that month keyword ranking, herein It is without restriction.
In step S1430, keyword categories corresponding to multiple keywords are searched.For example, the pass of data file 1 The corresponding keyword categories searched of keyword " two Conferences ", " National People's Congress ", " CPPCC ", " Xi Jinping ", " Hu Jintao " and " two sides " can Respectively " political meeting ", " political meeting ", " political meeting ", " politician ", " politician " and " international relations ".At this In inventive embodiments, it can be searched by encyclopaedia database (for example, wikipedia) or other databases with classification mechanism Its corresponding keyword categories is sought, it is without restriction herein.
In step S1440, according to multiple keywords, the respective keyword ranking of multiple keywords and keyword categories with And multiple respective current rankings of data file, to generate sequence algorithm.For example, by the data file 1 in table 2~ 4 respective possessed keywords, the keyword ranking of its keyword and the current row of keyword categories and data file 1~4 Name (25,38,67 and 184), can be to predict the sequence algorithm of the prediction ranking of another data file to generate.In the present invention In embodiment, multiple keywords, the respective keyword ranking of multiple keywords and keyword categories can be set as to sequence calculation The input of method, and by the respective current ranking setting of multiple data files be sort algorithm output, with generate sequence calculation Method.It is related as having to the current ranking of the keyword ranking of keyword possessed by data file and data file itself Property, it, can there are sequence algorithms to correspond to its relationship therefore with sufficient amount of data file.This Outside, keyword categories can correspond to the weighted value of its keyword, in other words, can be by the keyword categories of keyword, to determine Influence degree of the keyword for the current ranking of data file out.In embodiments of the present invention, when generating sequence algorithm, The class weight parameter of keyword categories and the ranking weight parameter of keyword ranking can be pre-defined out, and is largely being tested As a result in, class weight parameter and ranking weight parameter are therefrom adjusted and change, until the input value and output of the algorithm that sorts Until the result of value is in accuracy permissible range.In another embodiment of the invention, when generate sort algorithm when, can be by Inquired by curve-fitting method by or approximately through finite sequence data point (for example, sequence algorithm input value and Output valve) analog function (for example, analytical function), and curve-fitting method can be least square method, not limited herein System.
In step S1450, it is used to calculate the prediction ranking of another data file by sequence algorithm.For example, Due to data file 5 current ranking be it is unknown, when by above-mentioned data file 1~4 acquire sequence algorithm after, can be then It captures the keyword of data file 5 and inquires the keyword ranking of the keyword of file 5 and keyword type and input this Sort algorithm, and the prediction ranking that can calculate data file 5 is 360, and can thereby sorting data file 5.
As described above, in embodiments of the present invention, electronic device 101 can transmit the data file 5 of unknown current ranking To servomechanism 103, step S1410~S1440 is executed by servomechanism 103 to generate sequence algorithm and execute step Prediction ranking of the S1450 to generate data file 5 is finally again sent back its result in electronic device 101 with sorting to it.And In another embodiment of the invention, step S1410~S1450 all can be performed in electronic device 101, not limited herein System.
Figure 15 is the flow chart according to the sort method of the data file of one embodiment of the invention.As shown in figure 3, according to The step of sort method of the data file of the embodiment of the present invention includes S1410~S1430, S1441, S1442 and S1450.Under Column will be for explaining with above-mentioned difference.In embodiments of the present invention, step S1441, S1442 can further be executed Generate sequence algorithm.
In step S1441, the respective keyword categories of multiple keywords are converted into multiple keyword categories conversion values. For example, can be by the mode tabled look-up or specific formulation calculates, " the political meeting by the keyword categories of data file 1 View ", " politician " and ", international relations " was converted to keyword categories conversion value 10,20 and 30, by the keyword of data file 2 Classification " smartphone ", " scientific & technical corporation ", " scientific and technological personage " and " country " are converted to keyword categories conversion value 40,50,60 And 70, the keyword categories " program " of data file 3 and " singer " are converted into keyword categories conversion value 80 and 90, by data The keyword categories " team ", " sportsman " and " city " of file 4 are converted to keyword categories conversion value 100,110 and 120.It is above-mentioned The keyword categories conversion value lifted is the purposes as explanation, without restriction herein.
In step S1442, the respective keyword ranking of multiple keywords and keyword categories conversion value are set as The current ranking setting of multiple data files is the codomain of analog function, uses generation to hold by the domain of analog function The analog function of row sequence algorithm.For example, if the corresponding parameter of keyword ranking is x0、x1、x2、x3、x4And x5, crucial The corresponding parameter of word class is y0、y1、y2、y3、y4And y5And analog function is f (x0,x1,x2,x3,x4,x5,y0,y1,y2, y3,y4,y5), then the data file in the table of comparisons 1, for analog function f (x0,x1,x2,x3,x4,x5,y0,y1,y2,y3,y4,y5) For, x0Domain respectively includes 152,21,17 and 139, x1Domain respectively includes 96,57,53 and 87, x2Domain difference Including 135,42,66 and 106, x3Domain respectively includes 33,108,0 and 127, x4Domain respectively includes 47,317,0 and 0, x5Domain respectively includes 95,96,0 and 0, y0Domain respectively includes 10,40,80 and 100, y1Domain respectively includes 10, 50,90 and 110, y2Domain respectively includes 10,60,90 and 120, y3Domain respectively includes 20,60,0 and 120, y4Domain Respectively include 20,70,0 and 0, y5Domain respectively includes 30,50,0 and 0, analog function f (x0,x1,x2,x3,x4,x5,y0,y1, y2,y3,y4,y5) codomain respectively include 25,38,67 and 184, then can generate analog function by a large amount of test result f(x0,x1,x2,x3,x4,x5,y0,y1,y2,y3,y4,y5), or analog function f (x can be inquired by curve-fitting method0, x1,x2,x3,x4,x5,y0,y1,y2,y3,y4,y5).In embodiments of the present invention, analog function can be linear function and non-thread One of property function.
In step S1450, it is used to calculate the prediction ranking of another data file by sequence algorithm.For example, As described above, the prediction for calculating data file 5 can be used after producing the analog function to execute sequence algorithm Ranking.For example, the keyword ranking for first capturing the keyword of data file 5 is respectively 262,396,137 and 192 (its difference Corresponding x0、x1、x2And x3, and x4=x5=0), then to search the keyword categories of the keyword of data file 5 be respectively " scientific and technological people Object ", " Venture Capital "tibco software, inc." "TIBCO Software, " scientific & technical corporation " and " scientific & technical corporation ", and its keyword categories respective value respectively can for 60,130,50 and 50 (it respectively corresponds y0、y1、y2And y3, and y4=y5=0), it is inputted above-mentioned obtained analog function f (x0,x1,x2,x3, x4,x5,y0,y1,y2,y3,y4,y5) after, can acquire data file 5 prediction ranking be f (262,396,137,192,0,0, It 60,130,50,50,0,0)=360, and can thereby sorting data file 5.
As described above, in embodiments of the present invention, electronic device 101 can transmit the data file 5 of unknown current ranking To servomechanism 103, execute step S1410~S1430, S1441 and S1442 by servomechanism 103 generate sequence algorithm with And prediction ranking of the step S1450 to generate data file 5 is executed to sort to it, its result is finally sent back into electronics dress again It sets in 101.And in another embodiment of the invention, step S1410~S1430, S1441~S1442 and S1450 are all executable It is without restriction herein in electronic device 101.
In conclusion the present invention is by capturing the keyword in multiple data files, and by keyword, the pass retrieved Conversion value corresponding to keyword ranking and the keyword categories hunted out is set as the domain of analog function, by multiple numbers According to ranking setting current known to file be analog function codomain after, then can by definition class weight parameter and Ranking weight parameter generates analog function in a large amount of test results, or simulation letter can be inquired by curve-fitting method Number finally recycles analog function to execute sequence algorithm to calculate the prediction ranking of another data file, with another number that sorts According to file.
Although the present invention is disclosed as above with embodiment, however, it is not to limit the invention, and those skilled in the art exist Under the premise of not departing from the spirit and scope of the present invention, can make some changes and embellishment, thus protection scope of the present invention be with Subject to claim of the invention.

Claims (12)

1. a kind of sort method of data file, is suitable for electronic device, which includes：
Multiple keywords are captured respectively from the title content and body matter of multiple data files；
Keyword ranking corresponding to these keywords is retrieved by Search engine；
Search keyword categories corresponding to these keywords；And
According to respective keyword ranking corresponding to these keywords, these keywords and corresponding respective crucial part of speech The other and respective current ranking of these data files, to generate the analog function to execute sequence algorithm；
Wherein, which calculates the prediction of another data file to multiple keywords according to another data file Ranking, with another data file that sorts,
The step of wherein generating the sequence algorithm further include：
The respective keyword ranking of these keywords and the keyword categories are set as to multiple rankings of sequence algorithm Input varible and multiple classification input varibles, and be sequence calculation by the respective current ranking setting of these data files The output of method, to generate the sequence algorithm,
Wherein when the quantity of the keyword of these data files is defeated less than the ranking input varible of the sequence algorithm and classification When entering the quantity of parameter, extra ranking input varible and classification input varible are set to zero.
2. sort method as described in claim 1, wherein the sequence algorithm be by being executed using analog function, and should Analog function is one of linear function and nonlinear function.
3. sort method as claimed in claim 2, wherein the step of generating the sequence algorithm, further includes：
The respective keyword ranking of these keywords and the keyword categories are set as to the domain of the analog function, it will The current ranking setting of these data files is the codomain of the analog function, uses and generates the analog function.
4. sort method as described in claim 1, wherein the sequence algorithm further includes ranking weight parameter and classification power Weight parameter, to respectively correspond the respective keyword ranking of these keywords and the keyword categories.
5. sort method as described in claim 1, wherein the step of generating the sequence algorithm, further includes：
The respective keyword categories of these keywords are converted into multiple keyword categories conversion values.
6. a kind of sort method of data file is suitable for electronic device, including：
At least one first keyword is captured from the title content and body matter of the first data file；
Keyword ranking corresponding at least one first keyword is retrieved by Search engine；
Search keyword categories corresponding at least one first keyword；And
The respective keyword ranking of at least one first keyword and the keyword categories are respectively set as sequence algorithm Multiple ranking input varibles and multiple classification input varibles, to export the prediction ranking of first data file, with sequence First data file,
Wherein, which is to be performed by analog function, and analog function is according to multiple second data files Content and the respective current ranking of these second data files and generate,
Wherein, when ranking input varible and classification input of the quantity of at least one first keyword less than the sequence algorithm When the quantity of parameter, extra ranking input varible and classification input varible are set to zero.
7. sort method as claimed in claim 6, wherein the sequence algorithm is also generated according to the following steps：
Multiple second keywords are captured from the content of these the second data files；
Keyword ranking corresponding to these second keywords is retrieved by the Search engine；
Search keyword categories corresponding to these second keywords；And
According to these second keywords, the respective keyword ranking of these second keywords and the keyword categories and these The respective current ranking of second data file, to generate the sequence algorithm.
8. sort method as claimed in claim 7, wherein the sequence algorithm is also generated according to the following steps：
The respective keyword ranking of these second keywords and the keyword categories are set as to the input of sequence algorithm, And by the respective current ranking setting of these second data files be the sequence algorithm output, with generate the sequence calculation Method.
9. sort method as claimed in claim 8, wherein the sequence algorithm be by being executed using analog function, and should Analog function is one of linear function and nonlinear function.
10. sort method as claimed in claim 9, wherein the sequence algorithm is also generated according to the following steps：
The respective keyword ranking of these second keywords and the keyword categories are set as to the definition of the analog function The current ranking setting of these the second data files is the codomain of the analog function, uses and generate the analog function by domain.
11. sort method as claimed in claim 8, wherein the sequence algorithm further includes ranking weight parameter and classification power Weight parameter, to respectively correspond the respective keyword ranking of at least one first keyword and the keyword categories.
12. sort method as claimed in claim 8, wherein the sequence algorithm is also generated according to the following steps：
The respective keyword categories of these second keywords are converted into multiple keyword categories conversion values.
CN201310273231.5A 2013-07-02 2013-07-02 The sort method of data file Active CN104281577B (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN201310273231.5A CN104281577B (en) 2013-07-02 2013-07-02 The sort method of data file
TW102125770A TWI610257B (en) 2013-07-02 2013-07-18 Sorting method of data documents and display method for sorting landmark data
US14/271,458 US9558262B2 (en) 2013-07-02 2014-05-07 Sorting method of data documents and display method for sorting landmark data
US15/361,015 US10083241B2 (en) 2013-07-02 2016-11-24 Sorting method of data documents and display method for sorting landmark data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310273231.5A CN104281577B (en) 2013-07-02 2013-07-02 The sort method of data file

Publications (2)

Publication Number Publication Date
CN104281577A CN104281577A (en) 2015-01-14
CN104281577B true CN104281577B (en) 2018-11-16

Family

ID=52256461

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310273231.5A Active CN104281577B (en) 2013-07-02 2013-07-02 The sort method of data file

Country Status (2)

Country Link
CN (1) CN104281577B (en)
TW (1) TWI610257B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104281576B (en) * 2013-07-02 2018-08-31 威盛电子股份有限公司 The display methods of landmark data
TWI682286B (en) * 2018-08-31 2020-01-11 愛酷智能科技股份有限公司 System for document searching using results of text analysis and natural language input
CN113054739B (en) * 2021-03-12 2022-04-01 上海华翌电气有限公司 Emergency inverter power supply
CN113923209B (en) * 2021-09-29 2023-07-14 北京轻舟智航科技有限公司 Processing method for downloading batch data based on LevelDB

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101350154A (en) * 2008-09-16 2009-01-21 北京搜狗科技发展有限公司 Method and apparatus for ordering electronic map data
CN103034718A (en) * 2012-12-12 2013-04-10 北京博雅立方科技有限公司 Target data sequencing method and target data sequencing device

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090024598A1 (en) * 2006-12-20 2009-01-22 Ying Xie System, method, and computer program product for information sorting and retrieval using a language-modeling kernel function
US9105039B2 (en) * 2006-01-30 2015-08-11 Groupon, Inc. System and method for providing mobile alerts to members of a social network
GB2458388A (en) * 2008-03-21 2009-09-23 Dressbot Inc A collaborative online shopping environment, virtual mall, store, etc. in which payments may be shared, products recommended and users modelled.
TW201030540A (en) * 2009-02-11 2010-08-16 Intumit Inc L System for conducting a geographic-oriented keyword advertisement recommendation and method of the same
TWI468956B (en) * 2011-05-20 2015-01-11 104 Corp Method and system for personalizedly sorting searched information
CN103077190A (en) * 2012-12-20 2013-05-01 人民搜索网络股份公司 Hot event ranking method based on order learning technology

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101350154A (en) * 2008-09-16 2009-01-21 北京搜狗科技发展有限公司 Method and apparatus for ordering electronic map data
CN103034718A (en) * 2012-12-12 2013-04-10 北京博雅立方科技有限公司 Target data sequencing method and target data sequencing device

Also Published As

Publication number Publication date
CN104281577A (en) 2015-01-14
TW201503016A (en) 2015-01-16
TWI610257B (en) 2018-01-01

Similar Documents

Publication Publication Date Title
CN104281578B (en) The region labeling method and device of data file
WO2021139701A1 (en) Application recommendation method and apparatus, storage medium and electronic device
JP7182585B2 (en) program
CN109271574A (en) A kind of hot word recommended method and device
CN109783798A (en) Method, apparatus, terminal and the storage medium of text information addition picture
CN108984731A (en) Sing single recommended method, device and storage medium
CN109829104A (en) Pseudo-linear filter model information search method and system based on semantic similarity
US9916396B2 (en) Methods and systems for content-based search
US11216499B2 (en) Information retrieval apparatus, information retrieval system, and information retrieval method
CN104281577B (en) The sort method of data file
CN102930048B (en) Use the data rich found automatically with reference to the semanteme with vision data
CN106663113A (en) Saving and retrieving locations of objects
CN108228720A (en) Identify method, system, device, terminal and the storage medium of target text content and artwork correlation
CN104283904B (en) message notification method and device
CN104281576B (en) The display methods of landmark data
US9965529B2 (en) Maintaining search context
CN102737091A (en) Playlist creation apparatus, playlist creation method and playlist creating program
US20170075911A1 (en) Sorting method of data documents and display method for sorting landmark data
CN104615620A (en) Map search type identification method and device and map search method and system
JP2011018152A (en) Information presentation device, information presentation method, and program
CN108846103A (en) A kind of data query method and device
CN107251010B (en) Unstructured UI
KR102442224B1 (en) Method and system for providing information on search terms whose popularity increase rapidly
CN107463311A (en) Intelligent list is read
CN114168789A (en) Song tag expansion method and device, equipment, medium and product thereof

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant