CN103488752B - A kind of search method of POI intelligent retrievals - Google Patents
A kind of search method of POI intelligent retrievals Download PDFInfo
- Publication number
- CN103488752B CN103488752B CN201310438035.9A CN201310438035A CN103488752B CN 103488752 B CN103488752 B CN 103488752B CN 201310438035 A CN201310438035 A CN 201310438035A CN 103488752 B CN103488752 B CN 103488752B
- Authority
- CN
- China
- Prior art keywords
- retrieval
- result
- user
- input
- searching step
- 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
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/36—Input/output arrangements for on-board computers
- G01C21/3679—Retrieval, searching and output of POI information, e.g. hotels, restaurants, shops, filling stations, parking facilities
-
- 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
- G06F16/907—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Automation & Control Theory (AREA)
- Library & Information Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of search method of POI intelligent retrievals, the beginning of intelligent retrieval, the input of user is waited first, the input of user can by input method hand-written, phonetic, letter, voice typing, audio file, map picture file, text, unprocessed user inputs information, it is necessary to be inputted by natural language for more than, these unprocessed information are changed into textual character, and unnatural language；When retrieval result has been provided, the self study process that there is a retrieving corrects the rule for retrieving built-in understanding, and thinking association tree etc. using the keyword of user search with the relation of the actual searching step performed；And back up retrieval fragment；So far, retrieving terminates, and the result retrieved is supplied into user.Advantages of the present invention：It is the embodiment of intelligent retrieval, the analysis process inputted to user is considered in this scheme, based on the understanding to user search purpose, to complete whole retrieving.
Description
Technical field
The present invention relates to in-vehicle navigation apparatus retrieval technique field, a kind of more particularly to retrieval side of POI intelligent retrievals
Method.
Background technology
Retrieval on POI currently exists the search function of several comparative maturities, and user can be according to different retrievals
Condition, selects different search functions to be retrieved, but for the first user using navigation software, can also manage unclear for the moment
Main threads, feels to have so many search functions, the place looked for which retrieval, is individual it is difficult to the problem of solving.
Search function is merged whether this problem can be solved by that, although have the upper merging of some functions, but such as
The combination that fruit is still rested on physical, will also tend to the effect not reached, because function becomes complicated, work(before
Can search condition be all different, only accomplish a really internal search function that permeates, really intelligence is got up,
This problem could fundamentally solved.
The content of the invention
The purpose of the present invention designs to solve the above problems, and intelligent retrieval allows user not go to be concerned about that search function has
What difference, is adapted to which type of retrieval scene, and user only needs to take notice of the purpose of oneself retrieval, if can pass through keyword
Clearly expressing just can be so that this use habit is more conform with our daily behavioral traits, and spy provides a kind of POI intelligence
The search method that can be retrieved.
The invention provides a kind of search method of POI intelligent retrievals, it is characterised in that：Described POI intelligent retrievals
Search method, it is contemplated that the analysis process inputted to user, based on the understanding to user search purpose, to complete entirely to retrieve
Journey, is shown in Fig. 1：
Intelligent retrieval starts, and waits the input of user first, the input of user can by input method hand-written, phonetic, word
Excessive limit is not taken in mother, voice typing, audio file, map picture file, text, the herein input to user
System, meets the use habit of user as far as possible；Unprocessed user inputs information, it is necessary to be inputted by natural language for more than,
These unprocessed information are changed into textual character, and unnatural language；For example：Voice signal to input is, it is necessary to voice
Identification changes into voice messaging, and picture file is also required to changing into the text message in picture into the side expressed in text or text
Tropism；
For the textual character after conversion, feature extraction is carried out according to default rule, the conversion of text, and word is carried out
With the replacement of symbol, some meaningless separation words are removed, selectively remove the not high word of some punctuation marks and discrimination；
Natural language understanding is based on the Statistical Probabilistic Models trained, in slave pattern rule base, to find approximate pattern, approximate mode
Can have multiple, each pattern can correspond to a retrieval thinking association tree, see Fig. 2, such as one user searches for " Shenyang station ",
Meeting searches corresponding thinking association tree according to the input at " Shenyang station ", if it is found, can be set according to thinking association, to retrieve knot
Really, Shenyang station can be retrieved first as railway station, and second step can retrieve the vehicles on Shenyang station periphery, can be preferential according to probability
The information of subway station is provided, the 3rd step can retrieve periphery fast food, and the 4th step can retrieve periphery lodging, and the 5th step can just be retrieved
Other include the POI of " railway station " title；If user's input is " company near the station of Shenyang ", company does not appear in inspection
In Suo Siwei associations tree, directly according to built-in retrieval model, it will be retrieved；
When multiple search modes occur, multiple searching steps can be generated, next can be to these searching steps
It is polymerize, and it is clearly irrational searching step to exclude, these processing procedures are the rules that define by some to enter
OK；After processing procedure as several wheels, some rough machined steps can be generated, these steps ensure it is to exist reasonably；
For these rough machined steps, also need plus preprocessing process, process etc. after processing, and to step comprehensive quantification
After each key element, sequence is optimized, the compiling optimization of whole searching step is just completed；
It is not the concrete operations for starting retrieval at once, but need first to inquire about and retrieve piece after the generation of these steps
Section whether there is, if it exists, just directly performing the fragment of retrieval；In retrieval fragment, a step searching step can be included
With a step retrieval result；Perform retrieval fragment branch, it is generally the case that retrieving only need to complete vacancy in searching step
Searching step, continued execution at appropriate " breakpoint "；It so can effectively accelerate the process of retrieval, perform retrieval piece
Duan Hou, can give retrieval result and arrange process；
If there is no retrieval fragment, searching step inventory can be generated, can initial interrogation and scheduling inspection in this process
The process of rope step, and opening space preserves contextual information between search argument and searching step etc.；Searching step is often held
Row one, all can subtract 1 in searching step inventory queue, when searching step inventory is reduced to 0, and retrieval has been worked it out
All retrieval results, can give retrieval result and arrange process afterwards；
Retrieval result arrange process the result retrieved can be scored and be classified, and to retrieval result according to scoring with
The sort result of classification, then the result to sequence merge, it is possible to return to use as the result of whole retrieving
Family；
After being performed according to searching step inventory before, occur retrieval result it is non-existent when, it is necessary to right before considering
The retrieval understanding content of user whether there is problem, therefore there is a feedback mechanism for understanding error here；Feedback machine
System can carry out re-organized to retrieval, and amplification search condition progressively phases out the not high word of those discriminations first, protect
The word that discrimination is high is stayed, until to the high word of discrimination is cancelled, retaining the not high word of discrimination；When there is result,
And when meeting the termination condition of retrieval, just do not continue to amplify search condition, processing procedure is to be terminated in advance, so
Give retrieval result again afterwards and arrange process, retrieval every time can all have tried to be supplied to user search result；
When retrieval result has been provided, the self study process that there is a retrieving utilizes the pass of user search
Key word corrects the rule for retrieving built-in understanding, and thinking association tree etc. with the relation of the actual searching step performed；And it is standby
Part retrieval fragment；So far, retrieving terminates, and the result retrieved is supplied into user.
The flow of intelligent retrieval is as follows：
Outside input：Input for receiving user, usually as directly there is provided a variety of defeated with the module of user mutual
Enter mode, meet the use habit of user, for example：User speech is inputted, handwriting input of user etc.；
Feature extraction：The behavior details of input to user, the content either inputted, or input, including symbol
Input, input of capital and small letter etc., or the still input to the keyword of input repeatedly, are identified as after useful feature, all
It can be recorded and extract as feature；
Text is changed：Need the content of feature extraction being further converted into content of text, some feature extractions go out source
From sound, some feature extractions are out derived from picture, and some feature extractions are out derived from string number, and these are required for these
Text implication of the Content Transformation into representative；When the process that text is changed, there is the situation of ambiguity if there is the explanation of text, need
These ambiguities are eliminated, the rule matched according to the dictionary of training and word carries out row's discrimination, and carries out participle to result,
The mark of part of speech role；
Semantic understanding：The result that this module can be changed to text, in pattern rules storehouse, carries out of pattern rules
Match somebody with somebody, the main execution step of retrieval can be generated；
Retrieve fragment：Core is a cache module, can be cached above and below the retrieval result of each step, each step execution
Text, also unsaved regular correction result and statistical information etc.；And the retrieval fragment of unified access interface and same type is provided
Classification storage and merging, swapping in and out strategy etc.；Allow a retrieving, in the presence of retrieval fragment, can save
Some steps are omitted, and there is the possibility for having secondary operation to each fragment；
Searching step is generated：In the case of the retrieval fragment asked is non-existent, it will perform complete searching step,
The step of searching step can be to semantic understanding, is compiled optimization processing, can consider after the key elements such as performance, internal memory, generation
One group of rational searching step, adds flow after the flow being connected between the flow of pretreatment, step and processing etc., will be most
The inventory of a searching step is produced eventually；At the same time, the initialization context variable memory headroom related to opening up can be also completed,
The processing of the step of for retrieval is prepared；
Searching step processing：This process, according to the service logic of retrieval, can travel through the data of retrieval according to the content of request
Storehouse content, obtains the retrieval result for meeting querying condition；Number is more than at zero, searching step the step of searching step inventory is defined
Reason will be called repeatedly；The end of each step, can all preserve the context of retrieval, retrieval and record retrieval piece for next step
Section is used；
Retrieval result processing：The result of retrieval can be classified, sorted, merged etc. by retrieval result processing to be operated, this
The result of step can just be given to external output module, for being exported to external device；
Error feedback processing：To after the generation of primary retrieval step, searching step processing, retrieval result is not present, intelligence
Energy searching system can judge that to the understanding that user inputs be there is error, it is necessary to change querying condition, error feedback processing meeting
Searching step is regenerated, new retrieval is carried out, after the condition terminated is met, can just stop retrieval；
Adaptive learning：This is a study module, result that can be according to retrieval and the input of user, is carried out adaptively
Study, to reach the purpose for constantly adapting to user's use habit；According to the multiple retrieval of user, the continuous update the system acquiescence of meeting
Rule, can influence sequence of result and retrieval result of retrieval etc.；
External output：It is exactly user interface there is provided the result to user search, the retrieval result asked user is carried out
Response.
Advantages of the present invention：
The search method of POI intelligent retrievals of the present invention, the embodiment of intelligent retrieval considers in this scheme
The analysis process inputted to user, based on the understanding to user search purpose, to complete whole retrieving.
Brief description of the drawings
Below in conjunction with the accompanying drawings and embodiment the present invention is further detailed explanation：
Fig. 1 is intelligent retrieval process schematic；
Fig. 2 is thinking association tree schematic diagram；
Fig. 3 is intelligent retrieval process chart.
Embodiment
Embodiment 1
The invention provides a kind of search method of POI intelligent retrievals, it is characterised in that：Described POI intelligent retrievals
Search method, it is contemplated that the analysis process inputted to user, based on the understanding to user search purpose, to complete entirely to retrieve
Journey, is shown in Fig. 1：
Intelligent retrieval starts, and waits the input of user first, the input of user can by input method hand-written, phonetic, word
Excessive limit is not taken in mother, voice typing, audio file, map picture file, text, the herein input to user
System, meets the use habit of user as far as possible；Unprocessed user inputs information, it is necessary to be inputted by natural language for more than,
These unprocessed information are changed into textual character, and unnatural language；For example：Voice signal to input is, it is necessary to voice
Identification changes into voice messaging, and picture file is also required to changing into the text message in picture into the side expressed in text or text
Tropism；
For the textual character after conversion, feature extraction is carried out according to default rule, the conversion of text, and word is carried out
With the replacement of symbol, some meaningless separation words are removed, selectively remove the not high word of some punctuation marks and discrimination；
Natural language understanding is based on the Statistical Probabilistic Models trained, in slave pattern rule base, to find approximate pattern, approximate mode
Can have multiple, each pattern can correspond to a retrieval thinking association tree, see Fig. 2, such as one user searches for " Shenyang station ",
Meeting searches corresponding thinking association tree according to the input at " Shenyang station ", if it is found, can be set according to thinking association, to retrieve knot
Really, Shenyang station can be retrieved first as railway station, and second step can retrieve the vehicles on Shenyang station periphery, can be preferential according to probability
The information of subway station is provided, the 3rd step can retrieve periphery fast food, and the 4th step can retrieve periphery lodging, and the 5th step can just be retrieved
Other include the POI of " railway station " title；If user's input is " company near the station of Shenyang ", company does not appear in inspection
In Suo Siwei associations tree, directly according to built-in retrieval model, it will be retrieved；
When multiple search modes occur, multiple searching steps can be generated, next can be to these searching steps
It is polymerize, and it is clearly irrational searching step to exclude, these processing procedures are the rules that define by some to enter
OK；After processing procedure as several wheels, some rough machined steps can be generated, these steps ensure it is to exist reasonably；
For these rough machined steps, also need plus preprocessing process, process etc. after processing, and to step comprehensive quantification
After each key element, sequence is optimized, the compiling optimization of whole searching step is just completed；
It is not the concrete operations for starting retrieval at once, but need first to inquire about and retrieve piece after the generation of these steps
Section whether there is, if it exists, just directly performing the fragment of retrieval；In retrieval fragment, a step searching step can be included
With a step retrieval result；Perform retrieval fragment branch, it is generally the case that retrieving only need to complete vacancy in searching step
Searching step, continued execution at appropriate " breakpoint "；It so can effectively accelerate the process of retrieval, perform retrieval piece
Duan Hou, can give retrieval result and arrange process；
If there is no retrieval fragment, searching step inventory can be generated, can initial interrogation and scheduling inspection in this process
The process of rope step, and opening space preserves contextual information between search argument and searching step etc.；Searching step is often held
Row one, all can subtract 1 in searching step inventory queue, when searching step inventory is reduced to 0, and retrieval has been worked it out
All retrieval results, can give retrieval result and arrange process afterwards；
Retrieval result arrange process the result retrieved can be scored and be classified, and to retrieval result according to scoring with
The sort result of classification, then the result to sequence merge, it is possible to return to use as the result of whole retrieving
Family；
After being performed according to searching step inventory before, occur retrieval result it is non-existent when, it is necessary to right before considering
The retrieval understanding content of user whether there is problem, therefore there is a feedback mechanism for understanding error here；Feedback machine
System can carry out re-organized to retrieval, and amplification search condition progressively phases out the not high word of those discriminations first, protect
The word that discrimination is high is stayed, until to the high word of discrimination is cancelled, retaining the not high word of discrimination；When there is result,
And when meeting the termination condition of retrieval, just do not continue to amplify search condition, processing procedure is to be terminated in advance, so
Give retrieval result again afterwards and arrange process, retrieval every time can all have tried to be supplied to user search result；
When retrieval result has been provided, the self study process that there is a retrieving utilizes the pass of user search
Key word corrects the rule for retrieving built-in understanding, and thinking association tree etc. with the relation of the actual searching step performed；And it is standby
Part retrieval fragment；So far, retrieving terminates, and the result retrieved is supplied into user.
The flow of intelligent retrieval is as follows：
Outside input：Input for receiving user, usually as directly there is provided a variety of defeated with the module of user mutual
Enter mode, meet the use habit of user, for example：User speech is inputted, handwriting input of user etc.；
Feature extraction：The behavior details of input to user, the content either inputted, or input, including symbol
Input, input of capital and small letter etc., or the still input to the keyword of input repeatedly, are identified as after useful feature, all
It can be recorded and extract as feature；
Text is changed：Need the content of feature extraction being further converted into content of text, some feature extractions go out source
From sound, some feature extractions are out derived from picture, and some feature extractions are out derived from string number, and these are required for these
Text implication of the Content Transformation into representative；When the process that text is changed, there is the situation of ambiguity if there is the explanation of text, need
These ambiguities are eliminated, the rule matched according to the dictionary of training and word carries out row's discrimination, and carries out participle to result,
The mark of part of speech role；
Semantic understanding：The result that this module can be changed to text, in pattern rules storehouse, carries out of pattern rules
Match somebody with somebody, the main execution step of retrieval can be generated；
Retrieve fragment：Core is a cache module, can be cached above and below the retrieval result of each step, each step execution
Text, also unsaved regular correction result and statistical information etc.；And the retrieval fragment of unified access interface and same type is provided
Classification storage and merging, swapping in and out strategy etc.；Allow a retrieving, in the presence of retrieval fragment, can save
Some steps are omitted, and there is the possibility for having secondary operation to each fragment；
Searching step is generated：In the case of the retrieval fragment asked is non-existent, it will perform complete searching step,
The step of searching step can be to semantic understanding, is compiled optimization processing, can consider after the key elements such as performance, internal memory, generation
One group of rational searching step, adds flow after the flow being connected between the flow of pretreatment, step and processing etc., will be most
The inventory of a searching step is produced eventually；At the same time, the initialization context variable memory headroom related to opening up can be also completed,
The processing of the step of for retrieval is prepared；
Searching step processing：This process, according to the service logic of retrieval, can travel through the data of retrieval according to the content of request
Storehouse content, obtains the retrieval result for meeting querying condition；Number is more than at zero, searching step the step of searching step inventory is defined
Reason will be called repeatedly；The end of each step, can all preserve the context of retrieval, retrieval and record retrieval piece for next step
Section is used；
Retrieval result processing：The result of retrieval can be classified, sorted, merged etc. by retrieval result processing to be operated, this
The result of step can just be given to external output module, for being exported to external device；
Error feedback processing：To after the generation of primary retrieval step, searching step processing, retrieval result is not present, intelligence
Energy searching system can judge that to the understanding that user inputs be there is error, it is necessary to change querying condition, error feedback processing meeting
Searching step is regenerated, new retrieval is carried out, after the condition terminated is met, can just stop retrieval；
Adaptive learning：This is a study module, result that can be according to retrieval and the input of user, is carried out adaptively
Study, to reach the purpose for constantly adapting to user's use habit；According to the multiple retrieval of user, the continuous update the system acquiescence of meeting
Rule, can influence sequence of result and retrieval result of retrieval etc.；
External output：It is exactly user interface there is provided the result to user search, the retrieval result asked user is carried out
Response.
Claims (2)
1. a kind of search method of POI intelligent retrievals, it is characterised in that：The search method of described POI intelligent retrievals, intelligence inspection
The beginning of rope, waits the input of user first, the input of user is by the way that input method is hand-written, phonetic, letter, voice typing, audio
File, map picture file, text, unprocessed user inputs information, it is necessary to be inputted by natural language for more than,
These unprocessed information are changed into textual character, and unnatural language；
For the textual character after conversion, feature extraction is carried out according to default rule, the conversion of text, and word and symbol is carried out
Number replacement, remove some meaningless separation words, selectively remove the not high word of some punctuation marks and discrimination；It is natural
Language understanding is based on the Statistical Probabilistic Models trained, in slave pattern rule base, to find approximate pattern, approximate mode has many
It is individual, the corresponding retrieval thinking association tree of each pattern；
When multiple search modes occur, multiple searching steps are generated, next these searching steps are polymerize,
And it is clearly irrational searching step to exclude, these processing procedures are the rules that define by some to carry out；By several
After the such processing procedure of wheel, some rough machined steps are generated, it is ensured that these steps are reasonable；
For these rough machined steps, also need plus preprocessing process, process after processing, and to each key element of step comprehensive quantification
Afterwards, sequence is optimized, the compiling optimization of whole searching step is just completed；
It is not the concrete operations for starting retrieval at once after the generation of these steps, but needing first to inquire about retrieval fragment is
It is no to exist, if it exists, just directly performing the fragment of retrieval；In retrieval fragment, a step searching step and a step are included
Retrieval result；Retrieval fragment branch is performed, retrieving in searching step, need to only complete the searching step of vacancy；So have
The process of the acceleration retrieval of effect, is performed after retrieval fragment, is given retrieval result and is arranged process；
If retrieval fragment is not present in whole searching steps, generate searching step inventory, in this process, initial interrogation and
The process of searching step is dispatched, and opening space preserves the contextual information between search argument and searching step；Searching step
One is often performed, all subtracts 1 in searching step inventory queue, when searching step inventory is reduced to 0, retrieval has been made
Carry out all retrieval results, retrieval result is given afterwards and arranges process；
Retrieval result arrange process the result retrieved is scored and classified, and to retrieval result according to scoring and classify
Sort result, then the result to sequence merge, return to user as the result of whole retrieving；
After being performed according to searching step inventory before, occur retrieval result it is non-existent when, it is necessary to user before considering
Retrieval understanding content whether there is problem, here exist a feedback mechanism for understanding error；Feedback mechanism is to retrieval
Re-organized is carried out, amplification search condition progressively phases out the not high word of those discriminations, retains discrimination high first
Word, until to the high word of discrimination is cancelled, retaining the not high word of discrimination；When there is result, and meet inspection
During the termination condition of rope, just do not continue to amplify search condition, processing procedure is to be terminated in advance, and retrieval is then given again
As a result process is arranged, retrieval every time all has tried to be supplied to user search result；
When retrieval result has been provided, the self study process that there is a retrieving utilizes the keyword of user search
The rule for retrieving built-in understanding, and thinking association tree are corrected with the relation of the actual searching step performed；And back up retrieval
Fragment；So far, retrieving terminates, and the result retrieved is supplied into user.
2. according to the search method of the POI intelligent retrievals described in claim 1, it is characterised in that：The flow of intelligent retrieval is as follows：
Outside input：Input for receiving user, directly there is provided a variety of inputs for meeting user's use habit with user mutual
Mode；
Feature extraction：Input to user, includes the content of input, and the behavior details of input, the content of the input includes symbol
Number input, the input of capital and small letter, the behavior details of the input includes input keyword repeatedly, is identified as useful feature
Afterwards, all it is recorded and extracts as feature；
Text is changed：Need the content of feature extraction being further converted into content of text, the source of the feature extraction includes
Sound, picture, string number, the text implication that content of text is converted into representing；The process changed in text, if there is text
There is the situation of ambiguity in this explanation, and these ambiguities are eliminated, and is carried out according to the rule that the dictionary of training and word are matched
Discrimination is arranged, and participle, the mark of part of speech role are carried out to result；
Semantic understanding：In pattern rules storehouse, the matching of pattern rules is carried out to the result that text is changed, it is main that generation is retrieved
Perform step；
Retrieve fragment：Cache the retrieval result of each step, the context that each step is performed, also unsaved regular correction result with
Statistical information；And the storage of retrieval segment classification and merging, the swapping in and out strategy of unified access interface and same type are provided；Permit
Perhaps a retrieving, in the presence of retrieval fragment, dispenses some steps, and presence has secondary add to each fragment
The possibility of work；
Searching step is generated：In the case of asked retrieval fragment is not present in whole searching steps, perform completely
Searching step, is compiled optimization processing the step of to semantic understanding, considers after performance, internal memory key element, and generation one is combined
The searching step of reason, adds the flow after the flow being connected between the flow of pretreatment, step and processing, final to produce an inspection
The inventory of rope step；At the same time, the initialization context variable memory headroom related to opening up is completed, is the processing of searching step
Prepare；
Searching step processing：This process, according to the service logic of retrieval, is traveled through in the database retrieved according to the content of request
Hold, obtain the retrieval result for meeting querying condition；Number is more than zero the step of searching step inventory is defined, and searching step processing is just
It is called multiple；The end of each step, all preserves the context of retrieval, and the retrieval and record for next step retrieve fragment to make
With；
Retrieval result processing：The result of retrieval is classified, sorted by retrieval result processing, union operation, by the knot of this step
Fruit is given to external output module, for being exported to external device；Wherein, external output module：Be exactly user interface there is provided
To the result of user search, the retrieval result asked user carries out response；
Error feedback processing：To after the generation of primary retrieval step, searching step processing, if retrieval result is not present, sentencing
It is fixed to there is error to the understanding that user inputs, querying condition is changed, error feedback processing regenerates searching step, carries out newly
Retrieval, after the condition terminated is met, stops retrieval；
Adaptive learning：According to the input of the result of retrieval and user, adaptive study is carried out, user is constantly adapted to reach
The purpose of use habit；According to the multiple retrieval of user, the rule of continuous update the system acquiescence influences result and the retrieval of retrieval
As a result sequence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310438035.9A CN103488752B (en) | 2013-09-24 | 2013-09-24 | A kind of search method of POI intelligent retrievals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310438035.9A CN103488752B (en) | 2013-09-24 | 2013-09-24 | A kind of search method of POI intelligent retrievals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103488752A CN103488752A (en) | 2014-01-01 |
| CN103488752B true CN103488752B (en) | 2017-08-25 |
Family
ID=49828978
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310438035.9A Active CN103488752B (en) | 2013-09-24 | 2013-09-24 | A kind of search method of POI intelligent retrievals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103488752B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12026593B2 (en) | 2020-10-15 | 2024-07-02 | Google Llc | Action suggestions for user-selected content |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6193179B2 (en) * | 2014-05-27 | 2017-09-06 | アイシン・エィ・ダブリュ株式会社 | Facility output system, facility output method, and facility output program |
| US9798801B2 (en) | 2014-07-16 | 2017-10-24 | Microsoft Technology Licensing, Llc | Observation-based query interpretation model modification |
| US10970646B2 (en) * | 2015-10-01 | 2021-04-06 | Google Llc | Action suggestions for user-selected content |
| CN105468468B (en) * | 2015-12-02 | 2018-07-27 | 北京光年无限科技有限公司 | Data error-correcting method towards question answering system and device |
| CN107292302B (en) * | 2016-03-31 | 2021-05-14 | 阿里巴巴（中国）有限公司 | Method and system for detecting interest points in picture |
| CN106057200A (en) * | 2016-06-23 | 2016-10-26 | 广州亿程交通信息有限公司 | Semantic-based interaction system and interaction method |
| CN106776981B (en) * | 2016-12-06 | 2020-12-15 | 广州同构科技有限公司 | Intelligent retrieval method based on empirical knowledge |
| CN109344342B (en) * | 2018-12-17 | 2021-04-09 | 北京百度网讯科技有限公司 | Map data retrieval method, map data retrieval device, map data retrieval server and map data retrieval system |
| CN110516094A (en) * | 2019-08-29 | 2019-11-29 | 百度在线网络技术（北京）有限公司 | De-weight method, device, electronic equipment and the storage medium of class interest point data |
| CN112784088A (en) * | 2019-11-04 | 2021-05-11 | 北京旷视科技有限公司 | Personnel retrieval method, device, electronic equipment and readable storage medium |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101916294A (en) * | 2010-08-27 | 2010-12-15 | 黄斌 | Method for realizing exact search by utilizing semantic analysis |
| CN102831177A (en) * | 2012-07-31 | 2012-12-19 | 聚熵信息技术（上海）有限公司 | Statement error correction method and system |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1335574A (en) * | 2001-09-05 | 2002-02-13 | 罗笑南 | Intelligent semantic searching method |
| CN101685448A (en) * | 2008-09-28 | 2010-03-31 | 国际商业机器公司 | Method and device for establishing association between query operation of user and search result |
| CN102436448A (en) * | 2010-09-29 | 2012-05-02 | 腾讯科技（深圳）有限公司 | Search method and search system |
| CN102890689B (en) * | 2011-07-22 | 2017-06-06 | 北京百度网讯科技有限公司 | The method for building up and system of a kind of user interest model |
| CN102543082B (en) * | 2012-01-19 | 2014-01-15 | 北京赛德斯汽车信息技术有限公司 | Voice operation method for in-vehicle information service system adopting natural language and voice operation system |
-
2013
- 2013-09-24 CN CN201310438035.9A patent/CN103488752B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101916294A (en) * | 2010-08-27 | 2010-12-15 | 黄斌 | Method for realizing exact search by utilizing semantic analysis |
| CN102831177A (en) * | 2012-07-31 | 2012-12-19 | 聚熵信息技术（上海）有限公司 | Statement error correction method and system |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12026593B2 (en) | 2020-10-15 | 2024-07-02 | Google Llc | Action suggestions for user-selected content |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103488752A (en) | 2014-01-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103488752B (en) | A kind of search method of POI intelligent retrievals | |
| US10896222B1 (en) | Subject-specific data set for named entity resolution | |
| CN105917327A (en) | System and method for inputting text into electronic devices | |
| US10997223B1 (en) | Subject-specific data set for named entity resolution | |
| CN105608218A (en) | Intelligent question answering knowledge base establishment method, establishment device and establishment system | |
| CN106776564B (en) | Semantic recognition method and system based on knowledge graph | |
| US10942977B2 (en) | Systems and methods for targeting, reviewing, and presenting online social network data by evidence context | |
| US20060129396A1 (en) | Method and apparatus for automatic grammar generation from data entries | |
| CN103678684A (en) | Chinese word segmentation method based on navigation information retrieval | |
| CN111159385A (en) | Template-free universal intelligent question-answering method based on dynamic knowledge graph | |
| US20180173694A1 (en) | Methods and computer systems for named entity verification, named entity verification model training, and phrase expansion | |
| CN106095778A (en) | The Chinese search word automatic error correction method of search engine | |
| CN106057200A (en) | Semantic-based interaction system and interaction method | |
| CN101561725B (en) | Method and system of fast handwriting input | |
| CN102693279A (en) | Method, device and system for fast calculating comment similarity | |
| CN113901079A (en) | System and method for content-based medical macro classification and search system | |
| CN109522396B (en) | Knowledge processing method and system for national defense science and technology field | |
| EP4147142A1 (en) | Creating and interacting with data records having semantic vectors and natural language expressions produced by a machine-trained model | |
| CN112699232A (en) | Text label extraction method, device, equipment and storage medium | |
| CN114860942B (en) | Text intention classification method, device, equipment and storage medium | |
| CN116450834A (en) | Archive knowledge graph construction method based on multi-mode semantic features | |
| CN115269834A (en) | High-precision text classification method and device based on BERT | |
| WO2023246719A1 (en) | Method and apparatus for processing meeting record, and device and storage medium | |
| CN117454898A (en) | Method and device for realizing legal entity standardized output according to input text | |
| Loglisci et al. | Toward geographic information harvesting: Extraction of spatial relational facts from Web documents |
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 | ||
| CP03 | Change of name, title or address |
Address after: 110000 No. 861-6, shangshengou village, Hunnan District, Shenyang City, Liaoning Province Patentee after: Shenyang Meihang Technology Co.,Ltd. Address before: 110179 13th floor, Debao building, No.1 Jinhui street, Hunnan New District, Shenyang City, Liaoning Province Patentee before: SHENYANG MXNAVI Co.,Ltd. |
|
| CP03 | Change of name, title or address |