CN103136226B - A kind of method and apparatus for searching for user - Google Patents
A kind of method and apparatus for searching for user Download PDFInfo
- Publication number
- CN103136226B CN103136226B CN201110380325.3A CN201110380325A CN103136226B CN 103136226 B CN103136226 B CN 103136226B CN 201110380325 A CN201110380325 A CN 201110380325A CN 103136226 B CN103136226 B CN 103136226B
- Authority
- CN
- China
- Prior art keywords
- user
- information
- index
- word
- 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.)
- Active
Links
Abstract
The invention discloses a kind of method and device for searching for user.This method includes：Each user profile in community is obtained, and the user profile of acquisition is handled, generates index information；The search key of user's input is received, inquires about index information and user profile corresponding to the search key in the user profile of storage and index information respectively, and obtained result output will be inquired about., can be with expanded search means, lifting search efficiency using the present invention.
Description
Technical field
The present invention relates to computer communication technology, more particularly to a kind of method and apparatus for searching for user.
Background technology
With the development of computer networking technology, by point-to-point technology come realize the instant communication on internet i.e.
Using more and more common, IM is received Shi Tongxin (IM, Instant Messaging) Software tool by most user
And pass through IM softwares as a part essential in user's daily life and work, user, it is possible to achieve with friend, together
Exchanges and communication between thing and classmate.
The part that miniature blog is applied as IM instruments, be an Information Sharing based on customer relationship, propagate and
Platform is obtained, user can be expressed by terminal devices such as personal PC, mobile phones by a word of 140 words or so
Oneself emotion, share with paying close attention to the user of oneself, so as to realize personal information share immediately, between user it is interactive, be to develop at present
Internet service the swiftest and the most violent.
Due to the popularity of miniature blog applications, user can carry out user profile in community conveniently by IM clients
Consulting and search, for example, by realizing user information search with the method for similar database search, for example, user logs in
Community, the search key such as user's pet name or user name, search engine (search server) root are inputted in the search column of community
User in the user profile to match according to complete matching algorithm (for example, string matching algorithm) inquiry with user's pet name is close
Claim, and user's nickname information is shown to user.
From above-mentioned, the method for existing search user, by search keys such as user's pet name or user names in community
Interior progress user's search, search means are single, it is difficult to meet the growing information search request of search user；Further,
Using the searching method matched completely with search key so that search result may omit search user's information needed, search
It is less efficient, reduce the search experience of user；Moreover, displaying user profile in user's pet name or user name content compared with
It is few, lack the complementary description information to the user.
The content of the invention
In view of this, the present invention proposes a kind of method for searching for user, can lift search efficiency.
The present invention also proposes a kind of device for searching for user, can lift search efficiency.
To reach above-mentioned purpose, the invention provides a kind of method for searching for user, this method includes：
Obtain each user profile in community；
Word segmentation processing is carried out to the user profile, obtains candidate index word；
Word filtering policy according to pre-setting filters to candidate index word, obtains indexing word, the mistake
Filter includes removing nonsense words from the candidate index word；
Processing is indexed to index word, obtains index information, the index information, which is used to map, carries out before processing
User profile；
The search key of user's input is received, is looked into respectively in the user profile of storage and the index information
Ask and obtain index information and user profile corresponding to the search key, and obtained result output will be inquired about.
Further comprise：
User corresponding to obtaining index information, inquires about the information that the user delivers；
The information delivered based on each user, obtain user behavior characteristic information；
User Priority fraction is calculated according to the user behavior characteristic information of acquisition and sorted, the use of output sequence top N
Index information result corresponding to family, wherein, N is natural number.
Further comprise：
The user of sequence top N is obtained, is matched with performing the friend information of user of search, is adjusted according to matching result
Entire row sequence, index information result output corresponding to the user after sorting will be adjusted.
Further comprise：
The information delivered each user carries out word segmentation processing, obtains candidate index word.
Further comprise：
To in user profile user's pet name carry out multi-characters combined treatment, and using obtained multi-characters combination as
Index information.
The user behavior characteristic information includes：User quality characteristic attribute, text quality's characteristic attribute, propagate temperature spy
Levy one or its any combination in attribute and time latitude characteristic attribute.
The calculating User Priority fraction includes：
Obtain user quality characteristic attribute value, text quality's characteristic attribute value respectively, propagate temperature characteristic attribute value and
Time latitude characteristic attribute value；
By the weight coefficient pre-set the user quality characteristic attribute value with acquisition, text quality's characteristic attribute respectively
Value, propagate temperature characteristic attribute value and the multiplication of time latitude characteristic attribute value；
The sum of each value obtained after being multiplied is calculated, as the User Priority fraction.
A kind of device for searching for user, the device include：Information excavating module and search inquiry module, wherein,
Information excavating module, for obtaining each user profile in community；
Cutting processing unit, for carrying out word segmentation processing to the user profile, obtain candidate index word；
Filter element, for being filtered according to the word filtering policy pre-set to candidate index word, obtain rope
Draw word, the filtering includes removing nonsense words from the candidate index word；
Index process unit, for being indexed processing to index word, index information is obtained, the index information is used for
Mapping carries out the user profile of before processing；
Search inquiry module, for receive user input search key, respectively storage the user profile with
And index information and user profile corresponding to the search key are inquired about in the index information, and the result that inquiry is obtained is defeated
Go out, and obtained result output will be inquired about.
Further comprise：User behavior characteristic information acquisition module, priority score calculates and order module, wherein,
User behavior characteristic information acquisition module, for obtaining user corresponding to index information, inquire about what the user delivered
Information, the information delivered based on each user, obtain user behavior characteristic information；
Priority score calculates and order module, for calculating User Priority according to the user behavior characteristic information of acquisition
Index information result corresponding to the user of output sequence top N after fraction and sequence, wherein, N is natural number.
Further comprise：
Matching module, for obtaining the user of sequence top N, matched with performing the friend information of user of search,
Adjusted and sorted according to matching result, index information result output corresponding to the user after sorting will be adjusted.
The cutting processing unit is further used for obtaining the information that each user delivers in community, and information is segmented
Processing, obtains candidate index word.
Described information is excavated module and further comprised：
Character combination indexing units, for carrying out multi-characters combined treatment to user's pet name in user profile, and will
Obtained multi-characters combine the index information as remarks.
As seen from the above technical solutions, a kind of method and device for searching for user provided in an embodiment of the present invention, is obtained
Each user profile in community, and the user profile of acquisition is handled, generate index information；The search for receiving user's input is closed
Keyword, inquiry obtains index information and use corresponding to the search key in the user profile of storage and index information respectively
Family information, and obtained result output will be inquired about.So, by being indexed processing to user so that perform the user of search
Any index information in index information is simply entered as search key, the user profile and setting can be obtained
Index information corresponding with the user, search means are not only extended, and the related letter of the user as much as possible can be obtained
Breath, so as to lift search efficiency and user's search experience.
Brief description of the drawings
Fig. 1 is the method flow schematic diagram of the search user of the embodiment of the present invention.
Fig. 2 is the method idiographic flow schematic diagram of the search user of the embodiment of the present invention.
Fig. 3 is the apparatus structure schematic diagram of the search user of the embodiment of the present invention.
Embodiment
It is right below in conjunction with the accompanying drawings and the specific embodiments to make the object, technical solutions and advantages of the present invention clearer
The present invention is described in further detail.
It is existing that the method for user is searched for, it is necessary to which user inputs accurately using the complete matching algorithm of character string by user's pet name
User's pet name or user name, if the user's pet name or user name of input are inaccurate, search can be caused to fail so that search
Means are single, it is difficult to meet to search for the growing information search request of user, search efficiency is relatively low.In the embodiment of the present invention,
Consider to carry out cutting word and word segmentation processing to user profile, and processing is indexed to the user profile of cutting word and word segmentation processing
To provide the complementary description information of user, so as to extend the index to match with the search key of user's input, showing
While user profile, index information corresponding to the user is shown in the lump, lifts the search experience of user.
Fig. 1 is the method flow schematic diagram of the search user of the embodiment of the present invention.Referring to Fig. 1, the flow includes：
Step 101, each user profile in community is obtained, and the user profile of acquisition is handled, generates index information；
In this step, user profile refers to personal information information of the user in business, including：User name, user's mark
(ID), user's pet name, personal explanation etc., see the relevant technical literature for details for its data structure, will not be repeated here.Wherein,
Individual's explanation includes：Illustrate that text and remarks describe text.
The user profile of acquisition is handled, generation index information includes：
A11, word segmentation processing is carried out to user profile, obtain candidate index word；
In this step, text, explanation text can be described according to the user name in user profile, user's pet name, remarks, is entered
Row word segmentation processing, obtain each candidate index word.Wherein, remarks, which describe text, includes the learning experiences of user, career etc..
On carrying out the flow of word segmentation processing, belong to prior art, see the relevant technical literature for details, no longer superfluous herein
State.
Further, the information that can also be delivered each user, based on data mining and it is indexed processing.Perform step
Rapid A11 '.
A11 ', the information delivered each user carry out word segmentation processing, obtain candidate index word；
The word filtering policy that A12, basis are pre-set filters to candidate index word, obtains indexing word；
In this step, in each word obtained after word segmentation processing, including some stop words and unrelated part of speech word, this stops
Word and unrelated part of speech word are redundancy for the search of user, without practical significance, it is then desired to be filtered.
Wherein, stop words refers to insignificant common word or some symbols, for example, " ", " energy ", " ", " and ", " he " etc., it is unrelated
Part of speech word includes：Conjunction, descriptive word, tone function word, adjective, pronoun etc., for example, pronoun has " you ", " I ", " he " etc., even
Word have " and ", "AND" etc..
A13, processing is indexed to index word, as index information.
In this step, after being indexed processing to index word, i.e., index information is mapped as before carrying out word segmentation processing
Corresponding user information, so,, can be with according to the index word if the search key of user's input includes index word
The user profile of index word mapping is searched, so as to extend search means.
It is preferred that to be filtrated to get index word in noun, name (especially famous person), mechanism name, professional noun,
Place name is indexed, as the index information of remarks, in this way it can be ensured that user by search key (for example, with
The family pet name) when scanning for, while the index information set in user profile corresponding to the search key can be obtained, go forward side by side
Row displaying, with expanded search means, lift search efficiency.
In the embodiment of the present invention, in user profile each the text field carry out cutting, index after, can also to
Name in an account book or user's pet name carry out polynary character combination, and the index information using the combination of obtained multi-characters as remarks.I.e.
Further comprise：
A14, multi-characters combined treatment is carried out to user's pet name in user profile, and obtained multi-characters are combined
Index information as remarks.
In this step, by carrying out multi-characters combined treatment to user's pet name, the search with user's input can be improved
The matching probability of keyword, reduce the accuracy requirement of the search key to user's input.For example, if user's pet name is
" ABCD ", then changed by capital and small letter, the conversion of full half-angle, the processing such as either traditional and simplified characters conversion, and carry out multi-characters combination, can distinguish
Establish the index information of remarks：A, B, C, D, ABC, BCD, wherein, A, B, C, D are one-gram word, and ABC, BCD segment for ternary,
So, when can make it that the search key that user inputs includes any one index terms in the index of remarks, can search for
To user's nickname information corresponding to the index terms.Certainly, in practical application, binary character can also be used to combine and established respectively
The index information of remarks, for example, AB, AC, AD, BC, BD, CD.
Described above is the participle mode for establishing index, in practical application, for the retrieval request of user, what it was segmented
Mode can be slightly different with establishing the mode of index, be cut for example, carrying out unitary for search key of the length no more than 2
Word, such as " AB ", can directly be cut into A and B；And it is more than or equal to 3 search key for length, such as " ABCD ", can
In the form of being cut into ternary participle, i.e.,：ABC and BCD.So, be ABCD for user's pet name, search user can by with
Under several search keys, search for (recalling) user situation：
1. a or b or c or d is inputted, then asking to recall by one-gram word；
2. input ab or bc or cd, then ask, by participle one-gram word, can also be called back；
3. inputting abc or bcd, direct ternary participle is recalled；
4. inputting abcd, it can be segmented and recalled by ternary.
Step 102, the search key of user's input is received, is looked into respectively in the user profile of storage and index information
Index information and user profile corresponding to the search key are ask, and obtained result output will be inquired about.
In this step, receive user input search key after, from provide search service user profile and
Scanned in the index information of generation, i.e., the user profile to be matched according to the inquiry of complete matching algorithm with search key,
And the index information that inquiry matches with search key in the index information of generation, and obtained index letter will be searched for
Breath or, index information and user profile return to user.So, if inquiry is less than with searching in the user profile of storage
The user profile that rope keyword matches, the index to match with search key can also be inquired about from the index information of storage
Information simultaneously shows user, not only extends search means, also lifts search efficiency, furthermore, it is possible to obtain as much as possible be somebody's turn to do
User related information, improve user's search experience.
In practical application, search obtain index information corresponding to search key or, index information and user profile
Afterwards, can further include：
Step 103, user corresponding to acquisition index information, inquires about the information that the user delivers；
In this step, the subscribing relationship based on user, the social circle where user is obtained, wherein, subscribing relationship is user
Between relationship expression, for example, user A booking reader B formed subscribing relationship after, then all information that user B is sent can all push away
User A is sent to, equally, all information that user A is sent can all be pushed to user B, and social circle refers to user in each product society
The union of good friend in area.
From the social circle where user, the information that user delivers in social circle is inquired about.
Step 104, the information delivered based on each user, user behavior characteristic information is obtained；
In this step, user behavior characteristic information includes：User quality characteristic attribute, text quality's characteristic attribute, propagate
One in temperature characteristic attribute and time latitude characteristic attribute or its any combination.Wherein,
User quality characteristic attribute, to identify the information quality that user delivers, generally characterized with the quantity delivered
And use normalized so that the interval range of the user quality characteristic attribute value after normalized is [0,1], and 1 represents
A most active user, that is, user quality characteristic attribute value corresponding to the most user of the information content delivered, 0 represents do not have
Deliver user quality characteristic attribute value corresponding to the user of information.
Text quality's characteristic attribute, to identify the noun quantity included in the information that user delivers, normalizing can be used
Change is handled, and the interval range of text quality's characteristic attribute value is [0,1], and the noun quantity included in 1 information for representing to deliver is most
A more users, that is, text quality's characteristic attribute value corresponding to the most user of the noun quantity that includes in the information delivered, 0
Text quality's characteristic attribute value corresponding to the user of information is not delivered in expression., can be with obtaining the noun that includes in information
By foregoing word segmentation processing and counted to obtain.
Temperature characteristic attribute is propagated, to identify the relay number for each information that user delivers, comment number and number of visits,
Can according to pre-set strategy to relay number, comment number and number of visits carry out composite computing, and by normalization at
Reason obtains a propagation temperature characteristic attribute value；
Time latitude characteristic attribute, to identify the time for each information that user delivers, the letter that the user searched delivers
Breath is more late, shows the ageing stronger of the information, and normalized time latitude characteristic attribute value corresponding to it is also bigger.For example,
Stabbed using the earliest information time that search server stores to current time as time latitude, and be divided into and pre-set quantity
Period, for example, before the earliest information time stamp information of search server storage is 10 hours, then time dimension
(1,2,3,4,5,6,7,8,9,10) is could be arranged to, each scale represents a past chronomere (hour), i.e. 1 table
Show that information time stamp information is before 1 hour, and time latitude characteristic attribute value corresponding to each scale is set, for example, 1 correspondence
Time latitude characteristic attribute value be 1, time latitude characteristic attribute value corresponding to 10 be 0 etc..
Step 105, User Priority fraction is calculated according to the user behavior characteristic information of acquisition and sorted.
In this step, calculating User Priority fractional formula is：
V=ξ1×userank+ξ2×textrank+ξ3×hotrank+ξ4×timerank
In formula,
V is User Priority fraction；
Userank is user quality characteristic attribute value；
Textrank is text quality's characteristic attribute value；
Hotrank is propagation temperature characteristic attribute value；
Timerank is time latitude characteristic attribute value；
ξ1、ξ2、ξ3、ξ4For weight coefficient, can be determined according to being actually needed, it is preferred that 0≤ξi≤ 1,
After User Priority fraction is obtained, it is ranked up according to priority score height, and according to the plan pre-set
Slightly, index information result corresponding to the user of output sequence top N, wherein, N is natural number.So, user's matter based on information
Measure feature attribute, text quality's characteristic attribute, temperature characteristic attribute and time latitude characteristic attribute are propagated, using comprehensive related
The sortord of property, meets information search request of the user in community, this is that traditional search engines are difficult to, Ke Yijin
One step provides the index information result of enhancing, index information result is more met the demand of search user, so as to further be lifted
Search efficiency and the search experience for searching for user.
It is preferred that for the user for the top N that sorts, further can also be carried out with the friend information in the friend circle of user
Matching, adjusted and sorted according to matching result, indexed results corresponding to the user after sorting will be adjusted to search user's displaying, that is, entered
One step includes step 106.
Step 106, obtain the user of sequence top N, matched with the friend information of user for performing search, according to
Adjust and sort with result, index information result output corresponding to the user after sorting will be adjusted.
In this step, for example, be ranked up according to the offline level attribute fraction of user, take preceding 5 users, be sequentially
1~user of user 5, then, matched with the friend information in the friend circle of user, wherein, user 2, user 3 and user
5 is search for the good friend of user, i.e., the match is successful with the friend information in the friend circle of user, then user's difference after adjustment sequence
For：User 2, user 3, user 5, user 1 and user 4, that is to say, that matched into the friend information in the friend circle of user
The User Priority highest of work(, its corresponding index information is preferentially to search user's displaying.So, by by the top N of sorting
User is matched with searching for the good friend of user again, preferential displaying and the search user good friend user that the match is successful, further
Improve the search experience of search user.
From above-mentioned, the method for the search user of the embodiment of the present invention, by obtaining each user profile in community, and base
Increase the index information of remarks in the user profile of acquisition in data mining；The search key inputted according to user, in society
Inquiry obtains index information and user profile corresponding to the search key and exported in area.So, by carrying out rope to user
Draw processing so that include multiple index informations in the user, thus, search user simply enters any rope in index information
Fuse breath is used as search key, can obtain the index information corresponding with the user of the user profile and setting, from
And complementary description information is provided for search user, search means are extended, the growing information of search user is met and searches
Rope demand, search user is obtained the user related information as much as possible, body is searched for so as to lift search efficiency and user
Test.
It is below IT civilian worker using the entitled Zhang San of user, pet name xiaoz19889898, personal description, poor little Bai
Exemplified by the user of neck, the searching method of the embodiment of the present invention is illustrated.
Fig. 2 is the method idiographic flow schematic diagram of the search user of the embodiment of the present invention.Referring to Fig. 2, the flow includes：
Step 201, user profile is obtained；
In this step, search server obtains user profile, and extracts text message therein, retains the master of user profile
It is dry：Zhang San/xiaoz/IT civilian worker poverty young white collar worker.
Step 202, word segmentation processing is carried out, obtains candidate index word；
In this step, according to existing word segmentation processing algorithm, Zhang San/xiaoz/IT civilian worker poverty young white collar worker is segmented
Processing, and polynary character combination is carried out to user's pet name, the result after cutting and combination is：
, three, Zhang San；
x、i、a、o、z、xia、iao、aoz；
IT, civilian worker, poverty, small, white collar.
Step 203, candidate index word is filtered, and the candidate index word to being filtrated to get is indexed place
Reason, the index information as remarks；
In this step, nonsense words are removed from candidate index word：IT, poverty, small etc., retain civilian worker, white collar, and
Index is established for the candidate index word after filtering, i.e., by the index word of foundation, it is corresponding that the index word can be indexed
User, by the user, the index information that other in the user profile are set can be obtained.
Step 204, the index information of remarks is loaded into search server；
In this step, the index information of remarks is loaded into search server, see the relevant technical literature for details, herein
Repeat no more.
Step 205, the search key of user's input is received, is entered with the user profile in search server and index information
Row matching, if hit, return to search result；Otherwise, return and search for issueless prompt message.
In this step, if the search key of user's input is " Zhang San ", pass through and the index in search server
Matched, the indexed results of matching are returned to search user：Zhang San：User's pet name is xiaoz19889898, civilian worker's white collar.
If the search key of user's input is " three ", by being matched with the index in search system, returned to search user
Returning the indexed results matched and removing includes Zhang San：User's pet name is xiaoz19889898, outside civilian worker's white collar, in addition to " three "
The index information for the other users matched somebody with somebody, for example, the index information on Zhao three stored in search server.
Fig. 3 is the apparatus structure schematic diagram of the search user of the embodiment of the present invention.Referring to Fig. 3, the device includes：Information is dug
Module and search inquiry module are dug, wherein,
Information excavating module, handle, generate for obtaining each user profile in community, and to the user profile of acquisition
Index information；
Search inquiry module, for receiving the search key of user's input, respectively in the user profile and rope of storage
Index information and user profile corresponding to the search key are inquired about in fuse breath, and obtained result output will be inquired about.
Wherein, information excavating module includes：Cutting processing unit, filter element and index process unit (do not show in figure
Go out), wherein,
Cutting processing unit, for carrying out word segmentation processing to user profile, obtain candidate index word；
In the embodiment of the present invention, text, explanation text are described according to the user name of user, user's pet name, remarks, is divided
Word processing, obtains each candidate index word.
Filter element, for being filtered according to the word filtering policy pre-set to candidate index word, obtain rope
Draw word；
Index process unit, for being indexed processing to index word, the index information as remarks.
In the embodiment of the present invention, it is preferred that to be filtrated to get index word in noun, name (especially famous person),
Mechanism name, professional noun, place name are indexed, the index information as remarks.
Cutting processing unit is further used for obtaining the information that each user delivers in community, and information is carried out at participle
Reason, obtains candidate index word.
Information excavating module further comprises：
Character combination indexing units, for carrying out multi-characters combined treatment to user's pet name in user profile, and will
Obtained multi-characters combine the index information as remarks.
The device further comprises：User behavior characteristic information acquisition module, priority score calculates and order module, its
In,
User behavior characteristic information acquisition module, for obtaining user corresponding to index information, inquire about what the user delivered
Information, the information delivered based on each user, obtain user behavior characteristic information；
In the embodiment of the present invention, user behavior characteristic information includes：User quality characteristic attribute, text quality's feature category
Property, propagate temperature characteristic attribute and time latitude characteristic attribute.Wherein,
Priority score calculates and order module, for calculating User Priority according to the user behavior characteristic information of acquisition
Index information result corresponding to the user of output sequence top N after fraction and sequence, wherein, N is natural number.
It is preferred that the device further comprises：
Matching module, for obtaining the user of sequence top N, matched with performing the friend information of user of search,
Adjusted and sorted according to matching result, index information result output corresponding to the user after sorting will be adjusted.
In practical application, the device of search user can be search server.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvement for being made etc., the guarantor of the present invention should be included in
Within the scope of shield.
Claims (10)
- A kind of 1. method for searching for user, it is characterised in that this method includes：Obtain each user profile in community；Word segmentation processing is carried out to the user profile, obtains candidate index word；The candidate index word is filtered according to the word filtering policy pre-set, obtains indexing word, the mistake Filter includes removing nonsense words from the candidate index word；Processing is indexed to the index word, obtains index information, the index information, which is used to map, carries out before processing User profile；The search key of user's input is received, inquiry should in the user profile of storage and the index information respectively Index information corresponding to search key and user profile, and obtained result output will be inquired about；Methods described also includes：Multi-characters combined treatment is carried out to user's pet name in user profile, and obtained multi-characters are combined as remarks Index information.
- 2. the method as described in claim 1, it is characterised in that further comprise：User corresponding to obtaining index information, inquires about the information that the user delivers；The information delivered based on each user, obtain user behavior characteristic information；User Priority fraction is calculated according to the user behavior characteristic information of acquisition and sorted, the user couple of output sequence top N The index information result answered, wherein, N is natural number.
- 3. method as claimed in claim 2, it is characterised in that further comprise：The user of sequence top N is obtained, is matched with performing the friend information of user of search, is adjusted and arranged according to matching result Sequence, index information result output corresponding to the user after sorting will be adjusted.
- 4. the method as described in claim 1, it is characterised in that further comprise：The information delivered each user carries out word segmentation processing, obtains candidate index word.
- 5. method as claimed in claim 2 or claim 3, it is characterised in that the user behavior characteristic information includes：User quality is special Levy attribute, text quality's characteristic attribute, propagate one in temperature characteristic attribute and time latitude characteristic attribute or its is any Combination.
- 6. method as claimed in claim 5, it is characterised in that the calculating User Priority fraction includes：User quality characteristic attribute value, text quality's characteristic attribute value are obtained respectively, propagate temperature characteristic attribute value and time Latitude characteristic attribute value；By the weight coefficient pre-set respectively the user quality characteristic attribute value with acquisition, text quality's characteristic attribute value, pass Broadcast temperature characteristic attribute value and time latitude characteristic attribute value is multiplied；The sum of each value obtained after being multiplied is calculated, as the User Priority fraction.
- 7. a kind of device for searching for user, it is characterised in that the device includes：Information excavating module and search inquiry module, Wherein,Information excavating module, for obtaining each user profile in community；Cutting processing unit, for carrying out word segmentation processing to the user profile, obtain candidate index word；Filter element, for being filtered according to the word filtering policy pre-set to candidate index word, obtain index terms Language, the filtering include removing nonsense words from the candidate index word；Index process unit, for being indexed processing to index word, index information is obtained, the index information is used to map Carry out the user profile of before processing；Search inquiry module, for receiving the search key of user's input, the respectively user profile in storage and institute State and index information and user profile corresponding to the search key are inquired about in index information, and obtained result will be inquired about；Described information is excavated module and further comprised：Character combination indexing units, for carrying out multi-characters combined treatment to user's pet name in user profile, and it will obtain Multi-characters combine index information as remarks.
- 8. device as claimed in claim 7, it is characterised in that further comprise：It is user behavior characteristic information acquisition module, excellent First level fraction calculates and order module, wherein,User behavior characteristic information acquisition module, for obtaining user corresponding to index information, the information that the user delivers is inquired about, The information delivered based on each user, obtain user behavior characteristic information；Priority score calculates and order module, for calculating User Priority fraction according to the user behavior characteristic information of acquisition And index information result corresponding to the user of output sequence top N after sorting, wherein, N is natural number.
- 9. device as claimed in claim 8, it is characterised in that further comprise：Matching module, for obtaining the user of sequence top N, matched with performing the friend information of user of search, according to Matching result adjustment sequence, index information result output corresponding to the user after sorting will be adjusted.
- 10. device as claimed in claim 7, it is characterised in that the cutting processing unit is further used for obtaining in community The information that each user delivers, and word segmentation processing is carried out to information, obtain candidate index word.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110380325.3A CN103136226B (en) | 2011-11-25 | 2011-11-25 | A kind of method and apparatus for searching for user |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110380325.3A CN103136226B (en) | 2011-11-25 | 2011-11-25 | A kind of method and apparatus for searching for user |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103136226A CN103136226A (en) | 2013-06-05 |
| CN103136226B true CN103136226B (en) | 2017-11-17 |
Family
ID=48496062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110380325.3A Active CN103136226B (en) | 2011-11-25 | 2011-11-25 | A kind of method and apparatus for searching for user |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103136226B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103559243A (en) * | 2013-10-28 | 2014-02-05 | 陶睿 | Method and system for searching users in mobile devices on basis of labels |
| CN103699675B (en) * | 2013-12-30 | 2017-07-04 | 语联网（武汉）信息技术有限公司 | A kind of method of interpreter's hierarchical index |
| CN105159899B (en) * | 2014-06-12 | 2020-09-29 | 北京搜狗科技发展有限公司 | Searching method and device |
| CN105913322A (en) * | 2016-04-07 | 2016-08-31 | 合肥学院 | Subscriber social circle-based information dissemination assessment method and device |
| US10911382B2 (en) | 2017-01-30 | 2021-02-02 | Futurewei Technologies, Inc. | Personalized message priority classification |
| CN107092639A (en) * | 2017-02-23 | 2017-08-25 | 武汉智寻天下科技有限公司 | A kind of search engine system |
| CN110019190B (en) * | 2017-09-21 | 2023-05-30 | 阿里云计算有限公司 | Method and device for creating index |
| CN111611372A (en) * | 2019-02-25 | 2020-09-01 | 北京嘀嘀无限科技发展有限公司 | Search result sorting method and device and music searching method and device |
| CN110197389A (en) * | 2019-03-04 | 2019-09-03 | 腾讯科技（深圳）有限公司 | A kind of user identification method and device |
| CN110941765A (en) * | 2019-12-04 | 2020-03-31 | 青梧桐有限责任公司 | Search intention identification method, information search method and device and electronic equipment |
| CN111294275B (en) * | 2020-02-26 | 2022-10-14 | 上海云鱼智能科技有限公司 | User information indexing method, device, server and storage medium of IM tool |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8572094B2 (en) * | 2007-08-17 | 2013-10-29 | Google Inc. | Ranking social network objects |
| US20090171691A1 (en) * | 2007-12-28 | 2009-07-02 | Humanbook, Inc. | System and method for a web-based social networking database |
| CN101374148A (en) * | 2008-09-18 | 2009-02-25 | 腾讯科技（深圳）有限公司 | Method and system for managing blog intimate authority |
| CN101446961A (en) * | 2008-12-24 | 2009-06-03 | 腾讯科技（深圳）有限公司 | Method and system for carrying out association on users and friends thereof in network community |
-
2011
- 2011-11-25 CN CN201110380325.3A patent/CN103136226B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN103136226A (en) | 2013-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103136226B (en) | A kind of method and apparatus for searching for user | |
| CN103761242B (en) | Search method, searching system and natural language understanding system | |
| CN107301213A (en) | Intelligent answer method and device | |
| CN102866990B (en) | A kind of theme dialogue method and device | |
| CN102088419B (en) | Method and system for searching information of good friends in social network | |
| CN108304437A (en) | A kind of automatic question-answering method, device and storage medium | |
| US20150302056A1 (en) | Method, system, and storage medium for information search | |
| US20140280242A1 (en) | Method and apparatus for acquiring hot topics | |
| CN100578506C (en) | Network information searching method and searching system thereof | |
| CN113704451A (en) | Power user appeal screening method and system, electronic device and storage medium | |
| CN110134756A (en) | Minutes generation method, electronic device and storage medium | |
| WO2016196004A1 (en) | Joining semantically-related data using big table corpora | |
| US11836169B2 (en) | Methods, systems and techniques for providing search query suggestions based on non-personal data and user personal data according to availability of user personal data | |
| CN107515932A (en) | Artificial intelligence law consulting system based on typical problem storehouse | |
| KR20150041592A (en) | Method for updating contact information in callee electronic device, and the electronic device | |
| Wang et al. | Who are in the darknet? Measurement and analysis of darknet person attributes | |
| KR20110086233A (en) | Apparatus and method for searching user of common interest | |
| JP4369104B2 (en) | Community formation support system, its terminal, server and program | |
| CN109636682A (en) | A kind of teaching resource auto-collection system | |
| Wirawan et al. | Balinese historian chatbot using full-text search and artificial intelligence markup language method | |
| US20200401582A1 (en) | Low-latency typeahead with improved recall | |
| WO2013084065A1 (en) | A method and system for interacting with a web site | |
| JP2006350779A (en) | Communication system | |
| Shin | Socio-technical design of algorithms: Fairness, accountability, and transparency | |
| TW200947241A (en) | Database indexing algorithm and method and system for database searching using the same |
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 |