CN104123319A - Method and device for analyzing query with map requirement - Google Patents
Method and device for analyzing query with map requirement Download PDFInfo
- Publication number
- CN104123319A CN104123319A CN201310156743.3A CN201310156743A CN104123319A CN 104123319 A CN104123319 A CN 104123319A CN 201310156743 A CN201310156743 A CN 201310156743A CN 104123319 A CN104123319 A CN 104123319A
- Authority
- CN
- China
- Prior art keywords
- word
- pattern
- tag
- map
- search
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
Abstract
The invention provides a method and device for analyzing a query with a map search requirement. The method comprises the steps that word segmentation processing is carried out on the query input by a user; tag mapping is carried out on natural language words in the query; according to similarities between the natural language words in the query and tags in a tag system, mapped tags are determined; the tags in the tag system are POI attributes in a map, and corresponding POIs can be marked; search keywords corresponding to the query are determined according to a tag mapping result, and a map search engine is used for searching for the determined search keywords. The method and device for analyzing the query with the map requirement can return a search result needed by the user for the natural language query, and does not need to depend on the coverage condition of an artificial word list.
Description
[technical field]
The present invention relates to the information search field in Computer Applied Technology, particularly a kind of method and apparatus of resolving thering is the search terms of map demand.
[background technology]
Along with developing rapidly of network technology, the information resources on network are enriched constantly, and information data amount is also in expansion at full speed.Search engine has become the important way of people's obtaining information gradually, and map search is wherein a kind of important search application, for people's trip is provided convenience.
In map search, user is in input frame after inputted search item (query), map search engine can provide cartographic information corresponding to this query to user, and for example, when user's input " KFC ", search engine represents to user after the positional information of KFC can being identified in map.In existing map search, conventionally query is left intact and directly carries out text matches, this query for user's input is named entities such as place name, building name, trade company's name, or some are during such as classifiers such as " quick hotel ", because its information description of ordering with map POI is consistent, the Search Results returning can be good at meeting consumers' demand.
But, the query statement of user's input in a lot of situations is more random, there is the feature of natural language, for example " it is joyful what Beijing has ", " neighbouring where have learn cook " etc., this Search Results that is difficult to find by traditional text matches mode, in map POI point, can there is not " joyful ", " learning cook " such description, in addition, even if the mode of mating by artificial vocabulary also can cause the infull problem of covering, can not solve the word of the natural language of not including.
[summary of the invention]
In view of this, the invention provides a kind of to thering is method and apparatus that the search terms of map demand resolves so that can return to the Search Results of user's request to the query of natural language.
Concrete technical scheme is as follows:
A method of resolving thering is the search terms query of map search demand, the method comprises:
S1, the query that user is inputted carry out word segmentation processing;
S2, the word of natural language in described query is carried out to tag mapping: according to the similarity between each tag in the word of natural language in described query and tag system, definite tag being mapped to; Tag in wherein said tag system is the point of interest POI attribute in map, can hit corresponding POI;
S3, determine according to tag mapping result the searched key word that described query is corresponding, map search engine is searched for definite searched key word.
One preferred implementation according to the present invention also comprises in described step S1: after removing participle, obtain the stop words in word.
One preferred implementation according to the present invention, at least one between described step S1 and step S2 in further comprising the steps of S11 and S12:
S11, based on attribute vocabulary, the word obtaining after participle is carried out to Attribute Recognition and determines attribute word;
S12, based on pattern expression formula table, the word obtaining after participle is carried out to map search pattern-recognition;
In described step S2 by described query unidentified for attribute word and unidentified go out the word of map search pattern be defined as the word of natural language.
One preferred implementation according to the present invention, the mode of setting up of described pattern expression formula table is:
The query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, and remaining word is defined as pattern word;
Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency;
Select the sequence of co-occurrence frequency to meet the pattern expression formula of the described known map search pattern of pattern word formation of preset requirement.
One preferred implementation according to the present invention, the similarity in described step S2 between word and the tag of natural language can embody by co-occurrence rate, and the higher similarity of co-occurrence rate is larger; Wherein the co-occurrence rate between word x and the tag y of natural language is determined in the following ways:
Add up described x and the described y co-occurrence times N 1 in one text or the same window in language material, add up all tags of described x respectively and including described y times N now altogether in one text or the same window in language material, determine that the co-occurrence rate between described x and described y is N1/N.
One preferred implementation according to the present invention, in described step S2, by and described query between the word of natural language similarity meet the tag that the tag of preset requirement is defined as being mapped to, wherein said preset requirement is: the highest or similarity of similarity reaches predetermined threshold value.
One preferred implementation according to the present invention, if identify attribute word, in described step S3, determine that according to tag mapping result the searched key word that described query is corresponding is:
The described tag being mapped to is formed to the searched key word that described query is corresponding with the attribute word identifying.
One preferred implementation according to the present invention, if identify map search pattern, in described step S3, map search engine is searched for definite searched key word and is: map search engine is searched for definite searched key word according to the map search pattern identifying;
Otherwise map search engine is searched for definite searched key word and is in described step S3: map search engine is searched for definite searched key word according to the map search pattern of acquiescence.
One preferred implementation according to the present invention, if described user inputs described query by common large search, if exist and identify attribute word, identify map search pattern and be mapped at least one in tag, determine that described query possesses map search demand, in the Search Results of common large search, embed the Search Results of described map search engine in described step S3, and the Search Results in described step S3 comes remarkable position by described map search engine in the Search Results of common large search.
A device of resolving thering is the search terms of map search demand, this device comprises:
Participle unit, for carrying out word segmentation processing to the query of user's input;
Map unit, for the word of described query natural language is carried out to tag mapping: the similarity between each tag in the word of natural language and tag system in the described query of foundation, determine the tag being mapped to; Tag in wherein said tag system is the POI attribute in map, can hit corresponding POI;
Search unit, determines for the tag mapping result according to described map unit the searched key word that described query is corresponding, and invocation map search engine is searched for definite searched key word.
One preferred implementation according to the present invention, described participle unit, also for removing the stop words of the word obtaining after participle.
One preferred implementation according to the present invention, this device also comprises at least one in Attribute Recognition unit and pattern recognition unit;
Described Attribute Recognition unit, for based on attribute vocabulary, carries out Attribute Recognition to the word obtaining after participle and determines attribute word;
Described pattern recognition unit, for based on pattern expression formula table, carries out map search pattern-recognition to the word obtaining after participle;
Described map unit by described query unidentified for attribute word and unidentified go out the word of map search pattern be defined as the word of natural language.
One preferred implementation according to the present invention, this device also comprises: Model Establishment unit, for setting up described pattern expression formula table, specifically carry out:
The query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, and remaining word is defined as pattern word;
Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency;
Select the sequence of co-occurrence frequency to meet the pattern expression formula of the described known map search pattern of pattern word formation of preset requirement.
One preferred implementation according to the present invention, the similarity between word and the tag of the natural language that described map unit adopts can embody by co-occurrence rate, and the higher similarity of co-occurrence rate is larger; Wherein the co-occurrence rate between word x and the tag y of natural language is determined in the following ways:
Add up described x and the described y co-occurrence times N 1 in one text or the same window in language material, add up all tags of described x respectively and including described y times N now altogether in one text or the same window in language material, determine that the co-occurrence rate between described x and described y is N1/N.
One preferred implementation according to the present invention, described map unit by and described query between the word of natural language similarity meet the tag that the tag of preset requirement is defined as being mapped to, wherein said preset requirement is: the highest or similarity of similarity reaches predetermined threshold value.
One preferred implementation according to the present invention, if described Attribute Recognition unit identifies attribute word, described search unit is in the time determining searched key word corresponding to described query according to tag mapping result, and the described tag being mapped to and the attribute word that identifies are formed to the searched key word that described query is corresponding.
One preferred implementation according to the present invention, if described pattern recognition unit identifies map search pattern, described search unit invocation map search engine is searched for definite searched key word according to the map search pattern identifying; Otherwise described search unit invocation map search engine is searched for definite searched key word according to the map search pattern of acquiescence.
One preferred implementation according to the present invention, if described user inputs described query by common large search, if existed, described Attribute Recognition unit identifies attribute word, pattern recognition unit identifies map search pattern and map unit is mapped at least one in tag, described search unit determines that described query possesses map search demand, in the Search Results of common large search, embed the Search Results of described map search engine, and in the Search Results of common large search, the Search Results of described map search engine is come to remarkable position.
As can be seen from the above technical solutions, the present invention is by carrying out tag mapping to the word of natural language in query, be mapped as the POI attribute in map, make the searched key word that utilizes the rear tag of mapping to form can hit POI corresponding in map, thereby make also can return to the Search Results of user's request for the query of natural language, and needn't depend on the covering situation of artificial vocabulary.
[brief description of the drawings]
The method flow diagram that Fig. 1 provides for the embodiment of the present invention one;
The Search Results instance graph of the map search engine that Fig. 2 provides for the embodiment of the present invention one;
The Search Results instance graph of the common large search that Fig. 3 provides for the embodiment of the present invention one;
The structure drawing of device that Fig. 4 provides for the embodiment of the present invention two.
[embodiment]
In order to make the object, technical solutions and advantages of the present invention clearer, describe the present invention below in conjunction with the drawings and specific embodiments.
Embodiment mono-,
The method flow diagram that Fig. 1 provides for the embodiment of the present invention, as shown in Figure 1, the method mainly comprises the following steps:
Step 101: the query to user's input carries out word segmentation processing.
Except being carried out, query word segmentation processing, can also remove the stop words in query., as example this query is carried out obtaining after word segmentation processing taking near query " where have learn cook's school " in embodiments of the present invention: " near ", " where having ", " ", " cook ", " ", " school ".Remove stop words wherein " ".
Step 102: based on attribute vocabulary, the word obtaining after participle is carried out to Attribute Recognition.
Here the attribute vocabulary adopting is set up in advance, wherein can include but not limited to: the named entities relevant to map such as place name, mechanism's name, building name, trade company's name, brand name, or, the classifiers relevant to map such as " quick hotel ", " school ", " bank ".
The mode of setting up of attribute vocabulary can adopt the mode of artificial interpolation or the mode of automatic mining, and being embodied as of this partial content is existing compared with proven technique, is not described in detail in this.
The upper example that continues, the word after word segmentation processing mates with attribute vocabulary respectively, and wherein " school " hits the classifier in attribute vocabulary, and other words are also miss.
Step 103: based on pattern expression formula table, the unidentified word for attribute word is carried out to map search pattern-recognition.
Pattern expression formula table is herein also that the map search pattern based on conventional is set up in advance, has expression formula corresponding to each map search pattern in this pattern expression formula table.When certain or some words mate with certain expression formula in pattern expression formula table, by this, certain or some words are identified as the map search pattern corresponding to expression formula of coupling.
Wherein, the map search pattern that pattern expression formula comprises can include but not limited to following several:
1) Perimeter pattern, whether its corresponding expression formula can be: " neighbouring where have * ", " where neighbouring * ", " neighbouring have * " etc.
2) route query pattern, its corresponding expression formula can be: " public transport is from * to * ", " subway * is to * ", " driving from * to * " etc.
3) place query pattern, its corresponding expression formula can be: " * is at which ", " * where ", " * position " etc.
Equally, the mode of setting up of pattern expression formula table can adopt the mode of artificial interpolation or the mode of automatic mining, while wherein adopting the mode of automatic mining, can each query of each map search pattern be carried out the statistics of word co-occurrence frequency and be obtained, particularly, can comprise following process: first the query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, remaining word is just pattern word; Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency; The pattern word of selecting the sequence of co-occurrence frequency to meet preset requirement forms the pattern expression formula of this known map search pattern, and the pattern word of for example selecting co-occurrence frequency to come front M forms the pattern expression formula of this known map search pattern, and M is default positive integer.In the time forming pattern expression formula, can further carry out handmarking.
Upper example continues, carry out after Attribute Recognition unidentified go out the word of attribute word be: " near ", " where having ", " ", " cook ", remaining word is mated in pattern expression formula table, wherein " near "+" where " hit near expression formula " where * ", near map search pattern corresponding to expression formula " where * " is Perimeter pattern, identify " near "+" where " corresponding Perimeter pattern.
It should be noted that, above-mentioned steps 102 and step 103 optionally one are carried out, also can all carry out as mentioned above, in the time all carrying out, can first perform step as mentioned above 102 rear execution steps 103, also can first perform step 103 rear execution steps 102, first based on pattern expression formula, the word after to word segmentation processing carries out map search pattern-recognition, and then based on attribute vocabulary to unidentified go out the word of map search pattern carry out Attribute Recognition.Even step 102 and step 103 can be carried out simultaneously, based on attribute vocabulary, the word after to word segmentation processing carries out the identification of attribute word, and, word based on pattern expression formula after to word segmentation processing carries out map search pattern-recognition, afterwards in step 104 again to not only unidentified for attribute word but also unidentified go out the word of map search pattern carry out tag mapping.
Step 104: according to the similarity between each tag in remaining word and tag system, determine the tag that remaining word is mapped to.
After above-mentioned processing before carry out this step, remaining word has been exactly the word of natural language conventionally, and said process also can be regarded the process of the word of determining natural language as.
Tag in tag system in the present invention can be the POI attribute in map, and the word classification, the character etc. of POI being described, can hit the POI in map by this tag.For example, tag " cook's training " is the attribute of the POI such as " Hai Hou cook training center ", " long-range cook vocational-technical training school ", " training exchange centre of Chinese cooking association ", can hit these POI; Tag snack food is the attribute of the POI such as " KFC ", " McDonald ", " Yonghe County's soya-bean milk ", " Burger King ", can hit these POI.
Similarity in remaining word and tag system between each tag can adopt the account form such as semantic similarity, preferably, can embody by co-occurrence rate in embodiments of the present invention, and the higher similarity of co-occurrence rate is larger.Co-occurrence rate between term1 and tag1 determines that method can be: statistics term1 and tag1 be the co-occurrence times N 1 in one text or the same window in language material, statistics term1 respectively with all tag times N now altogether in one text or the same window in language material, determine that the co-occurrence rate between term1 and tag1 is N1/N.
Then remaining word is mapped to and tag that between it, similarity meets preset requirement, wherein preset requirement can be that similarity is the highest, or similarity reaches predetermined threshold value etc.
Upper example continues: remaining word " learn cook " and tag " cook's training " the co-occurrence number of times in one text or the same window in language material is 200, with the number of occurrence altogether of other tag in tag system be 50, the co-occurrence rate between so remaining word " learn cook " and tag " cook's training " is 200/ (200+50)=0.8.Adopt identical method to calculate successively the co-occurrence rate between " learning cook " and other tag, finally determine that the co-occurrence rate between " learning cook " and tag " cook's training " is the highest, the similarity between " cook " and tag " cook's training " is the highest, will " learn cook " and be mapped to tag " cook's training ".
Similarity in remaining word and tag system between each tag can be calculated in real time, also can inquire about precalculated result, in advance the tag in the word in the query of some conventional natural languages and tag system is carried out to similarity calculating, carry out this step 104 in the process that the query of user's input is resolved time, directly inquire about precalculated similarity result of calculation.
Step 105: determine according to Attribute Recognition result and tag mapping result the searched key word that this query is corresponding, the map search pattern identifying according to step 103 is searched for definite searched key word.
In this step, can form new searched key word by hitting definite tag in the word (the attribute word identifying) of attribute vocabulary and tag mapping, upper example continues, the word that hits attribute vocabulary is " school ", in tag mapping, definite tag is " cook's training ", the new searched key word forming is " cook training school ", and the new search word certainly forming can be also " cook training school ", or " cook of school training " etc.Because the map search pattern identifying is Perimeter pattern, therefore according to Perimeter pattern, " cook training school " searched for.That is to say, the query " near the school that where has cook " of user's input, after above-mentioned resolving, is converted into: according to Perimeter pattern, " cook training school " searched for.Obvious this analysis result makes Search Results can more meet user's search need.Suppose that the current present position of user is in the Renmin University of China, the search of returning so can be the Search Results in Renmin University of China's Perimeter " cook training school ", as shown in Figure 2.
If the query of user's input does not identify map search pattern, can search for according to the map search pattern of acquiescence; If do not identify attribute word, the tag mapping being obtained is as searched key word.
A conventional application scenarios of above-mentioned analytic method is, user inputs query in the search box of map search, this application scenarios is actually have been given tacit consent to user and has had map search demand, that is to say, only have user to have map search demand and just can in the search box of map search, input query search for, the query of so now user's input is also that acquiescence has map search demand.
Also has a kind of application scenarios, user inputs query in the search box of common large search, now just directly default user has map search demand, under this application scenarios, also can carry out the flow process of above-described embodiment one, the Attribute Recognition of carrying out when step 102, as long as there is a kind of recognition result in the mapping of the tag in pattern-recognition and step 104 that step 103 is carried out, if exist word to be identified as attribute word, or the one in identified map search pattern, or be mapped on certain tag in tag system, can think that the query that user inputs has map search demand, can invocation map search engine according to searching for described in step 105, the Search Results of map search engine is embedded in the Search Results of large search, and can in the Search Results of large search, the Search Results of map search engine be come to remarkable position, for example rank the first, or highlight.
For example, user inputs near query " where have cook's school " in the search box of common large search, due in above-mentioned steps 102, in step 103 and step 104, be identified as attribute word by word, identified search pattern around and being mapped on tag " cook's training ", therefore determine that this query has map search demand, invocation map search engine is searched for according to the mode of step 105, the search engine of common large search also can be searched for and return to Search Results the query of user's input " near the school that where has cook " simultaneously, in this Search Results, comprise the Search Results of map search engine, and the Search Results of map search engine is ranked the first.As shown in Figure 3.
Be more than the detailed description that method provided by the present invention is carried out, below in conjunction with embodiment bis-, device provided by the invention be described in detail.
Embodiment bis-,
The structure drawing of device that Fig. 4 provides for the embodiment of the present invention two, this device arranges the server end with search engine, for carrying out alternately with browser or client, obtain the query of user's input, and return to browser or client the Search Results that this query is corresponding.As shown in Figure 4, this device mainly comprises: participle unit 01, map unit 02 and search unit 03.Further can comprise at least one in Attribute Recognition unit 04 and pattern recognition unit 05.
Get at the server end of search engine after the query of user's input that browser sends, if this query is the query that user inputs in the search box of map search, can directly think that this query has map search demand.Now, first participle unit 01 carries out word segmentation processing to the query of user's input, and further can remove to the word after word segmentation processing the processing of stop words.
The major function of map unit 02 is that the word of natural language in query is carried out to tag mapping, makes the word after mapping can directly hit map POI.Be specially: according to the similarity between each tag in the word of natural language in query and tag system, determine the tag being mapped to; Wherein the tag in tag system is the POI attribute in map, can hit corresponding POI.
Search unit 03 determines according to the tag mapping result of map unit the searched key word that query is corresponding afterwards, and invocation map search engine is searched for definite searched key word.
Above-mentioned Attribute Recognition unit 04 and pattern recognition unit 05 with select one or simultaneous mode be arranged between participle unit 01 and map unit 02.Attribute Recognition unit 04, for based on attribute vocabulary, carries out Attribute Recognition to the word obtaining after participle and determines attribute word.Pattern recognition unit 05, for based on pattern expression formula table, is carried out map search pattern-recognition to the word obtaining after participle.Map unit 02 by query unidentified for attribute word and unidentified go out the word of map search pattern be defined as the word of natural language.
If Attribute Recognition unit 04 and pattern recognition unit 05 exist simultaneously, serial exists in any order, and existence also can walk abreast.When serial, can be that Attribute Recognition unit 04 first carries out Attribute Recognition to the word after participle unit 01 word segmentation processing and determines attribute word, then pattern recognition unit 05 is carried out the map search pattern-recognition based on pattern expression formula table to the unidentified word for attribute word in query again, and shown in Fig. 4 is exactly this serial mode; Also can be pattern recognition unit 05 first based on pattern expression formula table, the word after to participle unit 01 word segmentation processing carries out map search pattern-recognition, then Attribute Recognition unit 04 again to pattern recognition unit 05 unidentified go out the word of map search pattern carry out the Attribute Recognition based on attribute vocabulary.
The mode of setting up of attribute vocabulary can adopt the mode of artificial interpolation or the mode of automatic mining, and being embodied as of this partial content is existing compared with proven technique, is not described in detail in this.Pattern expression formula table is also that the map search pattern based on conventional is set up in advance, has expression formula corresponding to each map search pattern in this pattern expression formula table.When certain or some words mate with certain expression formula in pattern expression formula table, by this, certain or some words are identified as the map search pattern corresponding to expression formula of coupling.Equally, the mode of setting up of pattern expression formula table can adopt the mode of artificial interpolation or the mode of automatic mining, while wherein adopting the mode of automatic mining, this device also comprises: Model Establishment unit 06, be used for setting up pattern expression formula table, the following process of concrete execution: the query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, and remaining word is defined as pattern word; Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency; The pattern word of selecting the sequence of co-occurrence frequency to meet preset requirement forms the pattern expression formula of known map search pattern.
The similarity of map unit 02 between the word and the tag that carry out the natural language adopting in the process of tag mapping can embody by co-occurrence rate, and the higher similarity of co-occurrence rate is larger; Wherein the co-occurrence rate between word term1 and the tag1 of natural language is determined in the following ways: statistics term1 and tag1 be the co-occurrence times N 1 in one text or the same window in language material, the all tags of statistics term1 respectively and including tag1 are the times N now altogether in one text or the same window in language material, determines that the co-occurrence rate between term1 and tag1 is N1/N.
The calculating of above-mentioned similarity can be real-time, also can be to obtain by inquiring about precalculated result, in advance the tag in the word in the query of some conventional natural languages and tag system is carried out to similarity calculating, in the process that the query of user's input is resolved, directly inquire about precalculated similarity result of calculation.
Afterwards, map unit 02 by and query between the word of natural language similarity meet the tag that the tag of preset requirement is defined as being mapped to, wherein preset requirement is: the highest or similarity of similarity reaches predetermined threshold value.
If Attribute Recognition unit 04 identifies attribute word, search unit 03, when determine searched key word corresponding to query according to tag mapping result, forms searched key word corresponding to query by the tag being mapped to and the attribute word that identifies.
If pattern recognition unit 05 identifies map search pattern, search unit 03 invocation map search engine is searched for definite searched key word according to the map search pattern identifying; Otherwise search unit 03 invocation map search engine is searched for definite searched key word according to the map search pattern of acquiescence.
Also there is so a kind of application scenarios, if user is by common large search input query, now just directly the query of default user input has map search demand, but need to whether there is map search demand to the query of user's input judge, if exist Attribute Recognition unit 04 to identify attribute word, pattern recognition unit 05 identifies at least one that map search pattern and map unit 02 are mapped in tag, search unit 03 just can determine that query possesses map search demand, now just can in the Search Results of common large search, embed the Search Results of map search engine, and in the Search Results of common large search, the Search Results of map search engine is come to remarkable position, for example rank the first, or highlight.
In several embodiment provided by the present invention, should be understood that disclosed apparatus and method can realize by another way.For example, device embodiment described above is only schematically, and for example, the division of described unit, is only that a kind of logic function is divided, and when actual realization, can have other dividing mode.In addition, the each functional unit in each embodiment of the present invention can be integrated in a processing unit, can be also that the independent physics of unit exists, and also can be integrated in a unit two or more unit.Above-mentioned integrated unit both can adopt the form of hardware to realize, and the form that also can adopt hardware to add SFU software functional unit realizes.
The integrated unit that the above-mentioned form with SFU software functional unit realizes, can be stored in a computer read/write memory medium.Above-mentioned SFU software functional unit is stored in a storage medium, comprise that some instructions (can be personal computers in order to make a computer equipment, server, or the network equipment etc.) or processor (processor) carry out the part steps of method described in each embodiment of the present invention.And aforesaid storage medium comprises: various media that can be program code stored such as USB flash disk, portable hard drive, ROM (read-only memory) (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disc or CDs.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any amendment of making, be equal to replacement, improvement etc., within all should being included in the scope of protection of the invention.
Claims (18)
1. a method of resolving having the search terms query of map search demand, is characterized in that, the method comprises:
S1, the query that user is inputted carry out word segmentation processing;
S2, the word of natural language in described query is carried out to tag mapping: according to the similarity between each tag in the word of natural language in described query and tag system, definite tag being mapped to; Tag in wherein said tag system is the point of interest POI attribute in map, can hit corresponding POI;
S3, determine according to tag mapping result the searched key word that described query is corresponding, map search engine is searched for definite searched key word.
2. method according to claim 1, is characterized in that, in described step S1, also comprises: after removing participle, obtain the stop words in word.
3. method according to claim 1, is characterized in that, at least one between described step S1 and step S2 in further comprising the steps of S11 and S12:
S11, based on attribute vocabulary, the word obtaining after participle is carried out to Attribute Recognition and determines attribute word;
S12, based on pattern expression formula table, the word obtaining after participle is carried out to map search pattern-recognition;
In described step S2 by described query unidentified for attribute word and unidentified go out the word of map search pattern be defined as the word of natural language.
4. method according to claim 3, is characterized in that, the mode of setting up of described pattern expression formula table is:
The query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, and remaining word is defined as pattern word;
Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency;
Select the sequence of co-occurrence frequency to meet the pattern expression formula of the described known map search pattern of pattern word formation of preset requirement.
5. method according to claim 1, is characterized in that, the similarity in described step S2 between word and the tag of natural language can embody by co-occurrence rate, and the higher similarity of co-occurrence rate is larger; Wherein the co-occurrence rate between word x and the tag y of natural language is determined in the following ways:
Add up described x and the described y co-occurrence times N 1 in one text or the same window in language material, add up all tags of described x respectively and including described y times N now altogether in one text or the same window in language material, determine that the co-occurrence rate between described x and described y is N1/N.
6. method according to claim 1 or 5, it is characterized in that, in described step S2, by and described query between the word of natural language similarity meet the tag that the tag of preset requirement is defined as being mapped to, wherein said preset requirement is: the highest or similarity of similarity reaches predetermined threshold value.
7. method according to claim 3, is characterized in that, if identify attribute word, in described step S3, determines that according to tag mapping result the searched key word that described query is corresponding is:
The described tag being mapped to is formed to the searched key word that described query is corresponding with the attribute word identifying.
8. method according to claim 3, it is characterized in that, if identify map search pattern, in described step S3, map search engine is searched for definite searched key word and is: map search engine is searched for definite searched key word according to the map search pattern identifying;
Otherwise map search engine is searched for definite searched key word and is in described step S3: map search engine is searched for definite searched key word according to the map search pattern of acquiescence.
9. according to the method described in claim 3,7 or 8, it is characterized in that, if described user inputs described query by common large search, if exist and identify attribute word, identify map search pattern and be mapped at least one in tag, determine that described query possesses map search demand, in the Search Results of common large search, embed the Search Results of described map search engine in described step S3, and the Search Results in described step S3 comes remarkable position by described map search engine in the Search Results of common large search.
10. a device of resolving having the search terms of map search demand, is characterized in that, this device comprises:
Participle unit, for carrying out word segmentation processing to the query of user's input;
Map unit, for the word of described query natural language is carried out to tag mapping: the similarity between each tag in the word of natural language and tag system in the described query of foundation, determine the tag being mapped to; Tag in wherein said tag system is the POI attribute in map, can hit corresponding POI;
Search unit, determines for the tag mapping result according to described map unit the searched key word that described query is corresponding, and invocation map search engine is searched for definite searched key word.
11. devices according to claim 10, is characterized in that, described participle unit, also for removing the stop words of the word obtaining after participle.
12. devices according to claim 10, is characterized in that, this device also comprises at least one in Attribute Recognition unit and pattern recognition unit;
Described Attribute Recognition unit, for based on attribute vocabulary, carries out Attribute Recognition to the word obtaining after participle and determines attribute word;
Described pattern recognition unit, for based on pattern expression formula table, carries out map search pattern-recognition to the word obtaining after participle;
Described map unit by described query unidentified for attribute word and unidentified go out the word of map search pattern be defined as the word of natural language.
13. devices according to claim 12, is characterized in that, this device also comprises: Model Establishment unit, for setting up described pattern expression formula table, specifically carry out:
The query of known map search pattern is carried out, after word segmentation processing, based on attribute vocabulary, the word that hits attribute vocabulary being filtered, and remaining word is defined as pattern word;
Pattern word is carried out to the statistics of co-occurrence frequency, and sort based on co-occurrence frequency;
Select the sequence of co-occurrence frequency to meet the pattern expression formula of the described known map search pattern of pattern word formation of preset requirement.
14. devices according to claim 10, is characterized in that, the similarity between word and the tag of the natural language that described map unit adopts can embody by co-occurrence rate, and the higher similarity of co-occurrence rate is larger; Wherein the co-occurrence rate between word x and the tag y of natural language is determined in the following ways:
Add up described x and the described y co-occurrence times N 1 in one text or the same window in language material, add up all tags of described x respectively and including described y times N now altogether in one text or the same window in language material, determine that the co-occurrence rate between described x and described y is N1/N.
15. according to the device described in claim 10 or 14, it is characterized in that, described map unit by and described query between the word of natural language similarity meet the tag that the tag of preset requirement is defined as being mapped to, wherein said preset requirement is: the highest or similarity of similarity reaches predetermined threshold value.
16. devices according to claim 12, it is characterized in that, if described Attribute Recognition unit identifies attribute word, described search unit is in the time determining searched key word corresponding to described query according to tag mapping result, and the described tag being mapped to and the attribute word that identifies are formed to the searched key word that described query is corresponding.
17. devices according to claim 12, it is characterized in that, if described pattern recognition unit identifies map search pattern, described search unit invocation map search engine is searched for definite searched key word according to the map search pattern identifying; Otherwise described search unit invocation map search engine is searched for definite searched key word according to the map search pattern of acquiescence.
18. according to claim 12, device described in 16 or 17, it is characterized in that, if described user inputs described query by common large search, if exist described Attribute Recognition unit to identify attribute word, pattern recognition unit identifies map search pattern and map unit is mapped at least one in tag, described search unit determines that described query possesses map search demand, in the Search Results of common large search, embed the Search Results of described map search engine, and in the Search Results of common large search, the Search Results of described map search engine is come to remarkable position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310156743.3A CN104123319B (en) | 2013-04-28 | 2013-04-28 | The method and apparatus that search terms with map demand are parsed |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310156743.3A CN104123319B (en) | 2013-04-28 | 2013-04-28 | The method and apparatus that search terms with map demand are parsed |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104123319A true CN104123319A (en) | 2014-10-29 |
| CN104123319B CN104123319B (en) | 2019-08-27 |
Family
ID=51768731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310156743.3A Active CN104123319B (en) | 2013-04-28 | 2013-04-28 | The method and apparatus that search terms with map demand are parsed |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104123319B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104462531A (en) * | 2014-12-23 | 2015-03-25 | 北京奇虎科技有限公司 | Method and system for determining whether search term invokes map interface |
| CN104537041A (en) * | 2014-12-23 | 2015-04-22 | 北京奇虎科技有限公司 | Method and system for determining whether map interface is called or not based on user search term |
| CN109783589A (en) * | 2018-12-13 | 2019-05-21 | 中国平安人寿保险股份有限公司 | The method, apparatus and storage medium of electronic map parsing address |
| CN110609880A (en) * | 2018-06-15 | 2019-12-24 | 北京搜狗科技发展有限公司 | Information query method and device and electronic equipment |
| JP2022532967A (en) * | 2020-04-24 | 2022-07-21 | バイドゥ オンライン ネットワーク テクノロジー（ペキン） カンパニー リミテッド | Similarity modeling methods, devices, electronic devices, storage media and programs for retrieving geographic locations |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010259A1 (en) * | 2006-07-10 | 2008-01-10 | Nec (China) Co., Ltd. | Natural language based location query system, keyword based location query system and a natural language and keyword based location query system |
| CN101339551A (en) * | 2007-07-05 | 2009-01-07 | 日电（中国）有限公司 | Natural language query demand extension equipment and its method |
| CN102855251A (en) * | 2011-06-30 | 2013-01-02 | 北京百度网讯科技有限公司 | Method and device for requirement identification |
| CN102880721A (en) * | 2012-10-15 | 2013-01-16 | 瑞庭网络技术（上海）有限公司 | Implementation method of vertical search engine |
-
2013
- 2013-04-28 CN CN201310156743.3A patent/CN104123319B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080010259A1 (en) * | 2006-07-10 | 2008-01-10 | Nec (China) Co., Ltd. | Natural language based location query system, keyword based location query system and a natural language and keyword based location query system |
| CN101339551A (en) * | 2007-07-05 | 2009-01-07 | 日电（中国）有限公司 | Natural language query demand extension equipment and its method |
| CN102855251A (en) * | 2011-06-30 | 2013-01-02 | 北京百度网讯科技有限公司 | Method and device for requirement identification |
| CN102880721A (en) * | 2012-10-15 | 2013-01-16 | 瑞庭网络技术（上海）有限公司 | Implementation method of vertical search engine |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104462531A (en) * | 2014-12-23 | 2015-03-25 | 北京奇虎科技有限公司 | Method and system for determining whether search term invokes map interface |
| CN104537041A (en) * | 2014-12-23 | 2015-04-22 | 北京奇虎科技有限公司 | Method and system for determining whether map interface is called or not based on user search term |
| CN104537041B (en) * | 2014-12-23 | 2018-05-04 | 北京奇虎科技有限公司 | A kind of definite user's query word whether the method and system of invocation map interface |
| CN110609880A (en) * | 2018-06-15 | 2019-12-24 | 北京搜狗科技发展有限公司 | Information query method and device and electronic equipment |
| CN109783589A (en) * | 2018-12-13 | 2019-05-21 | 中国平安人寿保险股份有限公司 | The method, apparatus and storage medium of electronic map parsing address |
| CN109783589B (en) * | 2018-12-13 | 2023-07-25 | 中国平安人寿保险股份有限公司 | Method, device and storage medium for resolving address of electronic map |
| JP2022532967A (en) * | 2020-04-24 | 2022-07-21 | バイドゥ オンライン ネットワーク テクノロジー（ペキン） カンパニー リミテッド | Similarity modeling methods, devices, electronic devices, storage media and programs for retrieving geographic locations |
| JP7203981B2 (en) | 2020-04-24 | 2023-01-13 | バイドゥ オンライン ネットワーク テクノロジー（ペキン） カンパニー リミテッド | Similarity model creation method, device, electronic device, storage medium and program for searching geographic location |
| US11836174B2 (en) | 2020-04-24 | 2023-12-05 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus of establishing similarity model for retrieving geographic location |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104123319B (en) | 2019-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101768521B1 (en) | Method and system providing informational data of object included in image | |
| CN109815308B (en) | Method and device for determining intention recognition model and method and device for searching intention recognition | |
| CN109189942B (en) | Construction method and device of patent data knowledge graph | |
| CN108121829B (en) | Software defect-oriented domain knowledge graph automatic construction method | |
| KR102422977B1 (en) | Natural language image search | |
| US9424524B2 (en) | Extracting facts from unstructured text | |
| WO2018050022A1 (en) | Application program recommendation method, and server | |
| CN111125343B (en) | Text analysis method and device suitable for person post matching recommendation system | |
| CN105760495B (en) | A kind of knowledge based map carries out exploratory searching method for bug problem | |
| CN106663117B (en) | Constructing graphs supporting providing exploratory suggestions | |
| TW202020691A (en) | Feature word determination method and device and server | |
| CN103699625A (en) | Method and device for retrieving based on keyword | |
| CN111832290B (en) | Model training method and device for determining text relevance, electronic equipment and readable storage medium | |
| CN107544988B (en) | Method and device for acquiring public opinion data | |
| CN111783518A (en) | Training sample generation method and device, electronic equipment and readable storage medium | |
| US10831820B2 (en) | Content based image management and selection | |
| US10740406B2 (en) | Matching of an input document to documents in a document collection | |
| CN112256845A (en) | Intention recognition method, device, electronic equipment and computer readable storage medium | |
| CN104123319A (en) | Method and device for analyzing query with map requirement | |
| CN104102639A (en) | Text classification based promotion triggering method and device | |
| CN111488468A (en) | Geographic information knowledge point extraction method and device, storage medium and computer equipment | |
| CN109271624B (en) | Target word determination method, device and storage medium | |
| KR101696499B1 (en) | Apparatus and method for interpreting korean keyword search phrase | |
| WO2020074017A1 (en) | Deep learning-based method and device for screening for keywords in medical document | |
| WO2015084757A1 (en) | Systems and methods for processing data stored in a database |
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 |