CN106156250A - A kind of search focus recommendation method and system - Google Patents
A kind of search focus recommendation method and system Download PDFInfo
- Publication number
- CN106156250A CN106156250A CN201510208472.0A CN201510208472A CN106156250A CN 106156250 A CN106156250 A CN 106156250A CN 201510208472 A CN201510208472 A CN 201510208472A CN 106156250 A CN106156250 A CN 106156250A
- Authority
- CN
- China
- Prior art keywords
- user
- search
- information
- recommendation
- text
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of search focus recommendation method and system, described method includes: obtain the search characteristics information of user；Described search characteristics information includes that user uses the custom characteristic information of search engine and user to search for content custom characteristic information；According to described search characteristics information, described user is divided into some groups；Recommend to search for accordingly recommendation information according to described group user in group.The scheme of the embodiment of the present invention, it is possible to be accustomed to according to the search of user, recommend, for user, the hot information that search is relevant, accurately obtains the search need of user thus recommends search recommendation information accurately, greatly improve user experience.
Description
Technical field
The present invention relates to Internet technical field, search for focus recommendation method and system particularly to one.
Background technology
Along with the development of Internet technology, become alternately is more and more important.Online is mutual, has become as day
The communication model that benefit is important.The needs mutual in order to meet users' information, various interactive softwares or ditch
Logical software arises at the historic moment.
Wechat is the free application journey providing instant messaging to service for intelligent terminal that company of Tengxun releases
Sequence, wechat support across common carrier, spanning operation system platform by network quickly send freely (need consumption
A small amount of network traffics) voice SMS, video, picture and word, it is also possible to use by sharing stream
The service plugs such as the data of media content and location-based social plug-in unit.Wechat provides public platform, friend
The functions such as circle, message propelling movement, user can pass through " shaking ", " searching number ", " neighbouring people ", sweep
Quick Response Code mode is added good friend and pays close attention to public platform, and content is shared with good friend and by user by wechat simultaneously
The splendid contents seen shares wechat circle of friends.
The application account that wechat public's account is developer or businessman applies on wechat public platform, account
With QQ account intercommunication, by public's account, businessman can realize on wechat platform and special group word,
Picture, voice, the comprehensive communication of video, interaction.The on-line off-line wechat defining a kind of main flow is mutual
Dynamic marketing mode.As the micro-clan of representative of on-line off-line wechat interactive marketing, the industry of the proposition standard that takes the lead in
The wechat platform development theory of common template and deep customization combines.Define on-line off-line wechat interactive marketing
Open applications platform.
User can carry out interaction by equipment of itself and public's account.Disappeared to the transmission of public's account by mobile phone
Breath, it is generally required to following steps:
First the wechat public's account of oneself is paid close attention to by your handset Wechat.Then the wechat public of oneself is logged in
Public account assistant clicked on by platform.The second step arranged public account assistant selects input, and you are to be bound micro-
Signal code.Then click on transmission wechat identifying code.Your handset Wechat can receive an information, then by number
Word is input to identifying code input frame and completes to verify and just complete binding.Subsequently into address list, search for the public
Account assistant.Then finding public's account assistant this number of this contact person is the account of official of Tengxun
Mphelper (assistant's account of Tengxun's public platform needs to carry out message cluster transmition by it).Just pay close attention to this account
Can be by sending pocket transmission news to it.Click through chat interface, be then sent to message, then
See whether the concern user of your public's account can receive this message.
Search engine (Search Engine) refers to according to certain strategy, uses specific computer program
From the Internet, collect information, after information is organized and processed, provide the user retrieval service, will
The information that user search is relevant shows the system of user.Search engine include full-text index, directory index,
META Search Engine, vertical search engine, aggregation type search engine, door search engine and free lists of links
Deng.
One search engine is made up of searcher, index, searcher and four parts of user interface.
The function of searcher is to roam in the Internet, finds and collection information.The function of index is to understand to search
The information that rope device is searched for, therefrom extracts index entry, for representing document and generating the rope of document library
Draw table.The function of searcher is the Rapid Detection document in index database of the inquiry according to user, carry out document with
The covariance mapping of inquiry, is ranked up the result that will export, and realizes certain End-user relevance feedback
Mechanism.The effect of user interface is input user's inquiry, display Query Result, provides End-user relevance feedback
Mechanism.
In prior art, usual user, after entering the wechat public number page, can launch at the page provided
Function of search.This function of search can be the search in the range of wechat public number, it is also possible to be searching of the whole network
Rope.Generally between search content and the dependency of result, existing search engine can provide certain
Result recommend, recommend Search Results or recommend other relevant to Search Results the most accurately for user
Information.But, existing result is recommended the most intelligent, it is recommended that result often with the actual demand of user
Differ greatly.Thus, need badly and want a kind of new result suggested design, to provide a user with heat the most accurately
Dot information is recommended, and improves user experience.
Summary of the invention
The present invention provides a kind of search focus recommendation method and system, in order to solve Search Results in prior art
The inaccurate problem of content recommendation.
The present invention provides a kind of search focus recommendation method, including:
Obtain the search characteristics information of user；Described search characteristics information includes that user uses the habit of search engine
Used characteristic information and user search for content custom characteristic information；
According to described search characteristics information, described user is divided into some groups；
Recommend to search for accordingly recommendation information according to described group user in group.
Described method also includes:
According to the keyword of described search recommendation information, described search recommendation information is classified；
According to described classification, described search recommendation information is recommended corresponding user.
Described method also includes:
Search characteristics information according to described user, uses cluster analysis, described user is divided into some little
Group.
Described method also includes:
The corresponding some users of each described group；Same described user can belong to several described groups.
Described method also includes:
Described search recommendation information is showed described user with picture and text tabular form；
Described user browses concrete exhibition information by the link of described picture and text list.
Described method also includes:
Described user is by paying close attention to corresponding wechat public number, and the page provided in described wechat public number enters
Line search operates；
Described wechat public number obtains the search characteristics information of user, and is that user pushes away at the wechat public number page
Recommend described search recommendation information.
A kind of search focus recommendation system, including:
Characteristic acquisition unit, for obtaining the search characteristics information of user；Described search characteristics information bag
Including user uses the custom characteristic information of search engine and user to search for content custom characteristic information；
User grouping unit, for being divided into some groups according to described search characteristics information by described user；
Search information recommendation unit, for recommending to search for accordingly to push away according to described group user in group
Recommend information.
Described search information recommendation unit, also includes:
Classification subelement, for the keyword according to described search recommendation information, by described search recommendation information
Classification；
Recommend subelement, for according to described classification, described search recommendation information is recommended corresponding user.
Described user grouping unit, also includes:
Cluster subelement, for according to described search characteristics information, uses cluster analysis to obtain the packet of user
Information；
Packet subelement, for the grouping information according to described user, is divided into some groups by described user.
Described search information recommendation unit, also includes:
Text exhibition subelement, for showing described use by described search recommendation information with picture and text tabular form
Family；
Link subelement, for arranging the link of described picture and text list, described user is by described picture and text list
Link browse concrete exhibition information.
In the embodiment of the present invention, by obtaining the search characteristics information of user；Described search characteristics information includes
User uses the custom characteristic information of search engine and user to search for content custom characteristic information；Search according to described
Described user is divided into some groups by rope characteristic information；Recommend accordingly according to described group user in group
Search recommendation information.The scheme of the embodiment of the present invention, it is possible to be accustomed to according to the search of user, push away for user
Recommend the hot information that search is relevant, accurately obtain the search need of user thus recommend search recommendation accurately
Breath, greatly improves user experience.
Other features and advantages of the present invention will illustrate in the following description, and, partly from explanation
Book becomes apparent, or understands by implementing the present invention.The purpose of the present invention and other advantages can
Realize by structure specifically noted in the description write, claims and accompanying drawing and obtain
?.
Below by drawings and Examples, technical scheme is described in further detail.
Accompanying drawing explanation
Accompanying drawing is for providing a further understanding of the present invention, and constitutes a part for description, with this
Bright embodiment is used for explaining the present invention together, is not intended that limitation of the present invention.In the accompanying drawings:
A kind of search focus recommendation Method And Principle flow chart that Fig. 1 provides for the embodiment of the present invention 1；
A kind of search focus recommendation system structure schematic diagram that Fig. 2 provides for the embodiment of the present invention 2；
Search information recommendation unit 23 structural representation that Fig. 3 provides for the embodiment of the present invention 2；
User grouping unit 22 structural representation that Fig. 4 provides for the embodiment of the present invention 2.
Detailed description of the invention
Below in conjunction with accompanying drawing, the preferred embodiments of the present invention are illustrated, it will be appreciated that described herein
Preferred embodiment is merely to illustrate and explains the present invention, is not intended to limit the present invention.
As it is shown in figure 1, a kind of search focus recommendation Method And Principle flow chart provided for the embodiment of the present invention 1,
Wherein,
Step 11, extracts the search characteristics information of user.
Search characteristics information includes that user uses the custom characteristic information of search engine and user to search for content and practise
Used characteristic information, namely include the information of user self, user use search engine behavioural information,
User searches for content custom, the self-defined information etc. of user's self-defining.These information can be used by acquisition
The search history data at family and obtain, it is also possible to the search custom of real time record user thus obtain.User's
Search characteristics information records all search information relevant to search engine of user.
Firstly the need of obtaining these information.Generally, the method for acquisition user profile includes and requires that user is voluntarily
Upload, during user uses, extract the modes such as search characteristics information, or, use application user
In log information, obtain user's characteristic information by the method for text analyzing.
It is necessary to use text feature and carry if carrying out Text Feature Extraction if based on the daily record data of user behavior
Take.Text mining is an intercrossing subject, relates to data mining, machine learning, pattern recognition, artificial intelligence
Multiple fields such as energy, statistics, Computational Linguistics, computer networking technology, informatics.Text mining
Being exactly to find tacit knowledge and a kind of Method and kit for of pattern from substantial amounts of document, it is sent out from data mining
Exhibition, but have again many different from traditional data mining.Text mining to as if magnanimity, isomery, point
The document (web) of cloth；Document content is the natural language that the mankind are used, and lacks the intelligible semanteme of computer.
Data handled by traditional data mining are structurized, and document (web) is all half structure or structureless.
So, the matter of utmost importance that text mining faces is the most reasonably to represent text, is allowed to wrap
Containing enough information to reflect the feature of text, it is unlikely to again excessively complexity and makes learning algorithm to process.Great
In the huge and voluminous network information, the information of 80% is deposited in a text form, and WEB text mining is
A kind of important form that WEB content is excavated.
Choosing of the expression of text and characteristic item thereof is a basic problem of text mining, information retrieval, it
The Feature Words extracted from text is carried out quantization to represent text message.They are structureless from one
Urtext is converted into structurized computer and i.e. text can be carried out science take out with the information of identifying processing
As, set up its mathematical model, in order to describe and to replace text.Enable a computer to by this model
Calculating and operation realize the identification to text.Owing to text is non-structured data, want from substantial amounts of
Text excavates useful information be necessary for first converting the text to accessible structured form.People at present
Generally use vector space model to describe text vector, but if directly unite by segmentation methods and word frequency
The characteristic item that meter method obtains is to represent each dimension in text vector, and so the dimension of this vector is by right and wrong
Normal is big.This undressed text vector not only brings huge computing cost to follow-up work, makes whole
The efficiency of processing procedure is very low, and can damage the accuracy of classification, clustering algorithm, so that obtained
Result be difficult to satisfactory.It is therefore necessary to text vector to be done further purified treatment, ensureing that original text contains
On the basis of justice, find out text feature the most representational to text feature classification.In order to solve this problem,
Maximally effective way is through feature selection and carrys out dimensionality reduction.
Research about text representation at present focuses primarily upon selection and the Feature Words selection of text representation model
Choosing of algorithm.For representing that the ultimate unit of text is commonly referred to feature or the characteristic item of text.Feature
Item must possess certain characteristic: 1) characteristic item is wanted can really identify content of text；2) characteristic item has mesh
The ability that mark text is distinguished mutually with other texts；3) number of characteristic item can not be too many；4) characteristic item separate than
It is easier to realize.Word, word or phrase can be used in Chinese text as the characteristic item representing text.Phase
Comparatively, word has higher ability to express than word, and word is compared with phrase, and the cutting difficulty ratio of word is short
The cutting difficulty of language is much smaller.Therefore, current most of Chinese Text Classification System all uses word as feature
, referred to as Feature Words.These Feature Words as the intermediate representation of document, be used for realizing document and document,
Similarity Measure between document and ownership goal.If using all of word all as characteristic item, then feature
The dimension of vector will be the hugest, thus cause amount of calculation too big, in this case, and text to be completed
Classification is nearly impossible.The major function of feature extraction is in the case of not damaging text core information
Reduce word number to be processed as far as possible, reduce dimension of a vector space with this, thus simplify calculating, improve literary composition
The speed of present treatment and efficiency.Text feature selection to the filtration of content of text and classification, clustering processing, from
The research of the parties concerned such as dynamic summary and user interest pattern discovery, Knowledge Discovery has very important shadow
Ring.Calculate the score value of each feature generally according to certain feature evaluation function, then press score value to these
Feature is ranked up, choose several score values the highest as Feature Words, here it is feature extraction (Feature
Selection)。
The mode of Feature Selection has 4 kinds: (I) is transformed to primitive character less by the method mapped or convert
New feature；(2) from primitive character, pick out some the most representational features；(3) according to the knowledge of expert
Select the most influential feature；(4) choose by the method for mathematics, find out the feature of information of most classifying,
This method is a kind of more accurate method, and the interference of anthropic factor is less, is particularly suitable for text automatic
Classified excavation systematic difference.
It practice, during extracting user's search characteristics, need user profile and behavioural information complete
Surface analysis, and set reasonably extraction feature quantity and particular content, more fully to react user's reality
Behavior characteristics, thus process for follow-up classification and lay the foundation.
Step 12, is divided into some groups according to search characteristics information by user.
On the basis of extracting user's search characteristics, according to the contact between user's search characteristics, user is divided
For several groups.The corresponding one or more user's search characteristics information of each group, thus by user's root
It is assigned to different groups according to the search characteristics information of self.With the corresponding multiple users of a small group, same user
Search characteristics information according to self can corresponding multiple groups.
As a rule, according to user's search characteristics by the scheme of user grouping, need to use clustering algorithm.Poly-
Alanysis belongs to the data analysing method of exploration.Generally, we to utilize cluster analysis would look like unordered right
As carrying out being grouped, sort out, to reach to be more fully understood that the purpose of object of study.In cluster result requirement group right
As similarity is higher, between group, objects similarity is relatively low.In user study, a lot of problems can be by cluster
Analyze and solve, such as, the information classification problem of website, the click behavior relatedness problem of webpage and use
Family classification problem etc..Wherein, user's classification is modal situation.
Cluster analysis computational methods mainly have the most several:
1, division methods (partitioning methods)
A given data set having N number of tuple or record, disintegrating method will construct K packet, each
Individual packet just represents a cluster, K < N.And this K packet meets following condition: (1) each
Packet is including at least a data recording；(2) each data recording belongs to and only belongs to a packet (note
Meaning: this requirement can be relaxed in some fuzzy clustering algorithm)；For given K, first algorithm is given
Going out an initial group technology, the method by iterating changes packet later so that improve each time
Packet scheme afterwards is the best, and so-called good standard is exactly: the record in same packet is the nearest
The best, and the record in different grouping is the most remote more good.The algorithm using this basic thought has: K-MEANS
Algorithm, K-MEDOIDS algorithm, CLARANS algorithm；
Major part division methods is based on distance.Given number of partitions k to be built, first division methods is created
Build one and initialize division.Then, it use a kind of iteration re-positioning technology, by object from one
Group moves to another group and divides.The general preparation of one good division is: right in same bunch
As the most close to each other or relevant, and the object in different bunches far as possible from or different.Also has many
Pass judgment on other criterions dividing quality.Traditional division methods can expand to subspace clustering rather than search
The whole data space of rope.When there is a lot of attribute and Sparse, this is useful.In order to reach complete
Office is optimum, may need exhaustive all possible division based on the cluster divided, and amount of calculation is very big.It practice,
Great majority application all have employed popular heuristic, such as k-average and k-CENTER ALGORITHM, asymptotic raising
Clustering result quality, approaches locally optimal solution.These heuristic clustering methods are well suited for finding the data of middle and small scale
In the data base of storehouse middle and small scale spherical bunch.In order to find to have complicated shape bunch and to ultra-large type data
Collection clusters, and needs to further expand the method based on dividing.
2, hierarchical method (hierarchical methods)
This method is decomposed as given data set carries out level, until certain condition meets.Tool
Body can be divided into again " bottom-up " and " top-down " two schemes.Such as in " bottom-up " scheme, just
During the beginning, each data recording forms one single group, and in ensuing iteration, it is mutual those
Neighbouring group is merged into a group, until all of record one packet of composition or certain condition meet is
Only.Represent algorithm to have: BIRCH algorithm, CURE algorithm, CHAMELEON algorithm etc..
Hierarchy clustering method can be based on distance or based on density or connectedness.Hierarchy clustering method
Some extensions have also contemplated that subspace clustering.The defect of hierarchical method is, once step (merge or
Division) complete, it cannot be revoked.These strict regulations are useful, because not worrying different choosing
The combined number selected, it will produce less computing cost.But the decision that this technology can not be righted the wrong.
Have been proposed for some methods improving hierarchical clustering quality.
3, method based on density (density-based methods)
Method based on density with other method fundamental difference is: it be not based on various away from
From, but based on density.Algorithm based on distance thus can be overcome can only to find the poly-of " similar round "
The shortcoming of class.The guiding theory of this method is exactly, as long as the density of the point in a region is greater than certain valve
Value, is just added in the most close cluster.Represent algorithm to have: DBSCAN algorithm, OPTICS
Algorithm, DENCLUE algorithm etc..
4, method based on grid (grid-based methods)
First data space is divided into the network of limited unit (cell) by this method, all of
Processing is all with single unit as object.The prominent advantage so processed be exactly processing speed very
Hurry up, generally this is unrelated with the number of record in target database, and it is only divided into how many with data space
Individual unit is relevant.Represent algorithm to have: STING algorithm, CLIQUE algorithm, WAVE-CLUSTER calculate
Method.
A lot of Spatial Data Mining problems, using grid is the most all a kind of effective method.Therefore, based on
The method of grid can be integrated with other clustering methods.
5, method based on model (model-based methods)
Method based on model supposes a model to each cluster, then looks for can be good at meeting
The data set of this model.Such a model be probably data point density fonction in space or
Other.An its potential supposition is exactly: target data set is determined by a series of probability distribution.
Generally have two kinds attempt directions: the scheme of statistics and the scheme of neutral net.
Certainly clustering method also has: Transitive Closure Method, Boolean matrix method, Direct Cluster Analysis, correlation analysis
Cluster, clustering method etc. based on statistics.
Existing cluster the most successfully solves the clustering problem of low-dimensional data.But owing to reality should
By the complexity of middle data, when processing many problems, existing algorithm often lost efficacy, especially for height
Dimension data and the situation of large data.Because traditional clustering method is when high dimensional data is concentrated and clustered, main
Run into two problems.One is that high dimensional data concentrates the most unrelated attribute of existence to make to deposit in all dimensions
Bunch probability almost nil；Another is that Data In High-dimensional Spaces is dilute compared with data distribution in lower dimensional space
Dredge, wherein data pitch from almost equal be universal phenomenon, and traditional clustering method is to gather based on distance
Class, therefore cannot build bunch based on distance in higher dimensional space.
High Dimensional Clustering Analysis analysis has become an important research direction of cluster analysis.High dimensional data clusters also simultaneously
It it is the difficult point of clustering technique.Along with the progress of technology makes data collection become increasingly easier, cause data
Storehouse scale is increasing, complexity is more and more higher, as various types of trade transaction datas, Web document,
Gene expression datas etc., their dimension (attribute) can generally achieve hundreds and thousands of dimension, the highest.
But, being affected by " dimensionality effect ", many shows good clustering method and is used in low-dimensional data space
The Clustering Effect that often cannot obtain on higher dimensional space.High dimensional data cluster analysis is in cluster analysis one
Very active field, it is also a challenging job simultaneously.High dimensional data cluster analysis is in city
The aspects such as field analysis, information security, finance, amusement, anti-terrorism have and are widely applied very much.
In the present embodiment, concrete restriction be there is no for clustering method, as long as user's search characteristics can be believed
Cease according to necessary condition stub, and according to the result of classification, user is grouped.
Step 13, the user in group recommends to search for accordingly recommendation information.
After completing the packet of user, the user in group is needed to recommend to search for recommendation information.Here search
Recommendation information is associated with user's search characteristics information.In principle only by search user-dependent in group
Recommendation information recommends the user in group.
Search recommendation information needs the keyword according to search recommendation information to classify, according to classification by correspondence
Search recommendation information be sent to the user in corresponding group.Such as, user's search characteristics information in group
Including football, then can extracting the keyword of search recommendation information, if there being this keyword of football, then will
This search recommendation information is sent to the user in this group.
In the present embodiment, described search recommendation information is showed described user with picture and text tabular form；Described
User browses concrete exhibition information by the link of described picture and text list.
Generally, described user by pay close attention to corresponding wechat public number, and described wechat public number provide
The page scans for operation；Described wechat public number obtains the search characteristics information of user, and the wechat public
Number page is that user recommends described search recommendation information.
It practice, the scheme of the present embodiment is not limited in the analysis to user's search characteristics, in the same way,
The present embodiment can be understood as recommended engine, is actively discovered that user is current or potential demand, and active push letter
Cease to the information network of user.The hobby of digging user and demand, actively recommend it interested to user or
The object needed.It not passive lookup, but active push；Bu Shi independent media, but media network；
It not search mechanism, but Active Learning.
The present embodiment utilizes based on content, based on user behavior, based on multiple methods such as social networks networks,
Its commodity liked or content is recommended for user.
Content-based recommendation is " gene " analyzing the content that user is browsing, and selection has with Current Content
The object recommendation of similar " gene " is to user.The most also " gene " of the most browsed content of user is analyzed,
Thus obtain its preference, then by the object recommendation with user preference to user.Such as, user is browsing one
The when of money bag, recommend the similar bag of other profiles for it.
Recommendation based on user behavior is then to utilize group intelligence algorithm, analyzes the group behavior of user, comprehensively
Analyze the similarity between user and user, user's individual demand to minority's commodity, thus improve simultaneously
Accuracy, multiformity and the novelty recommended.
Recommendation based on social networks network is the social networks network by analyzing user place, finds it
The user that can have influence on, or it is best able to have influence on the user of this user, then the individual character of comprehensive every user
Change preference to recommend.
As in figure 2 it is shown, a kind of search focus recommendation system structure schematic diagram provided for the embodiment of the present invention 2,
Wherein,
Characteristic acquisition unit 21, for obtaining the search characteristics information of user；Described search characteristics information
The custom characteristic information of search engine and user is used to search for content custom characteristic information including user；
User grouping unit 22, for being divided into some groups according to described search characteristics information by described user；
Search information recommendation unit 23, for recommending to search for accordingly according to described group user in group
Recommendation information.
As it is shown on figure 3, above-mentioned search information recommendation unit 23, also include:
Classification subelement 231, for the keyword according to described search recommendation information, recommends described search
Information classification；
Recommend subelement 232, for according to described classification, described search recommendation information is recommended accordingly
User.
Further, above-mentioned search information recommendation unit, also include:
Text exhibition subelement 233, for showing institute by described search recommendation information with picture and text tabular form
State user；
Link subelement 234, for arranging the link of described picture and text list, described user passes through described picture and text
The link of list browses concrete exhibition information.
As shown in Figure 4, above-mentioned user grouping unit 22, also include:
Cluster subelement 221, for according to described search characteristics information, uses cluster analysis to obtain user's
Grouping information；
Packet subelement 222, for the grouping information according to described user, is divided into some little by described user
Group.
In sum, in the embodiment of the present invention, by obtaining the search characteristics information of user；Described search spy
Reference breath includes that user uses the custom characteristic information of search engine and user to search for content custom characteristic information；
According to described search characteristics information, described user is divided into some groups；According to described group use in group
Family recommends to search for recommendation information accordingly.The scheme of the embodiment of the present invention, it is possible to be accustomed to according to the search of user,
Recommend, for user, the hot information that search is relevant, accurately obtain the search need of user thus recommend accurately to search
Rope recommendation information, greatly improves user experience.
Those skilled in the art are it should be appreciated that embodiments of the invention can be provided as method, system or meter
Calculation machine program product.Therefore, the present invention can use complete hardware embodiment, complete software implementation or knot
The form of the embodiment in terms of conjunction software and hardware.And, the present invention can use and wherein wrap one or more
Computer-usable storage medium containing computer usable program code (include but not limited to disk memory and
Optical memory etc.) form of the upper computer program implemented.
The present invention is with reference to method, equipment (system) and computer program product according to embodiments of the present invention
The flow chart of product and/or block diagram describe.It should be understood that can by computer program instructions flowchart and
/ or block diagram in each flow process and/or flow process in square frame and flow chart and/or block diagram and/
Or the combination of square frame.These computer program instructions can be provided to general purpose computer, special-purpose computer, embedding
The processor of formula datatron or other programmable data processing device is to produce a machine so that by calculating
The instruction that the processor of machine or other programmable data processing device performs produces for realizing at flow chart one
The device of the function specified in individual flow process or multiple flow process and/or one square frame of block diagram or multiple square frame.
These computer program instructions may be alternatively stored in and computer or the process of other programmable datas can be guided to set
In the standby computer-readable memory worked in a specific way so that be stored in this computer-readable memory
Instruction produce and include the manufacture of command device, this command device realizes in one flow process or multiple of flow chart
The function specified in flow process and/or one square frame of block diagram or multiple square frame.
These computer program instructions also can be loaded in computer or other programmable data processing device, makes
Sequence of operations step must be performed to produce computer implemented place on computer or other programmable devices
Reason, thus the instruction performed on computer or other programmable devices provides for realizing flow chart one
The step of the function specified in flow process or multiple flow process and/or one square frame of block diagram or multiple square frame.
Obviously, those skilled in the art can carry out various change and modification without deviating from this to the present invention
The spirit and scope of invention.So, if these amendments of the present invention and modification belong to the claims in the present invention
And within the scope of equivalent technologies, then the present invention is also intended to comprise these change and modification.
Claims (10)
1. a search focus recommendation method, it is characterised in that including:
Obtain the search characteristics information of user；Described search characteristics information includes that user uses the habit of search engine
Used characteristic information and user search for content custom characteristic information；
According to described search characteristics information, described user is divided into some groups；
Recommend to search for accordingly recommendation information according to described group user in group.
2. the method for claim 1, it is characterised in that described method also includes:
According to the keyword of described search recommendation information, described search recommendation information is classified；
According to described classification, described search recommendation information is recommended corresponding user.
3. the method for claim 1, it is characterised in that described method also includes:
Search characteristics information according to described user, uses cluster analysis, described user is divided into some little
Group.
4. method as claimed in claim 3, it is characterised in that described method also includes:
The corresponding some users of each described group；Same described user can belong to several described groups.
5. the method for claim 1, it is characterised in that described method also includes:
Described search recommendation information is showed described user with picture and text tabular form；
Described user browses concrete exhibition information by the link of described picture and text list.
6. the method for claim 1, it is characterised in that described method also includes:
Described user is by paying close attention to corresponding wechat public number, and the page provided in described wechat public number enters
Line search operates；
Described wechat public number obtains the search characteristics information of user, and is that user pushes away at the wechat public number page
Recommend described search recommendation information.
7. a search focus recommendation system, it is characterised in that including:
Characteristic acquisition unit, for obtaining the search characteristics information of user；Described search characteristics information bag
Including user uses the custom characteristic information of search engine and user to search for content custom characteristic information；
User grouping unit, for being divided into some groups according to described search characteristics information by described user；
Search information recommendation unit, for recommending to search for accordingly to push away according to described group user in group
Recommend information.
8. system as claimed in claim 7, it is characterised in that described search information recommendation unit, also
Including:
Classification subelement, for the keyword according to described search recommendation information, by described search recommendation information
Classification；
Recommend subelement, for according to described classification, described search recommendation information is recommended corresponding user.
9. system as claimed in claim 7, it is characterised in that described user grouping unit, also includes:
Cluster subelement, for according to described search characteristics information, uses cluster analysis to obtain the packet of user
Information；
Packet subelement, for the grouping information according to described user, is divided into some groups by described user.
10. system as claimed in claim 7, it is characterised in that described search information recommendation unit, also
Including:
Text exhibition subelement, for showing described use by described search recommendation information with picture and text tabular form
Family；
Link subelement, for arranging the link of described picture and text list, described user is by described picture and text list
Link browse concrete exhibition information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510208472.0A CN106156250A (en) | 2015-04-28 | 2015-04-28 | A kind of search focus recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510208472.0A CN106156250A (en) | 2015-04-28 | 2015-04-28 | A kind of search focus recommendation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106156250A true CN106156250A (en) | 2016-11-23 |
Family
ID=57347576
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510208472.0A Pending CN106156250A (en) | 2015-04-28 | 2015-04-28 | A kind of search focus recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106156250A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111400577A (en) * | 2018-12-14 | 2020-07-10 | 阿里巴巴集团控股有限公司 | Search recall method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110010363A1 (en) * | 2009-07-08 | 2011-01-13 | Sony Corporation | Information processing apparatus, information processing method, and program |
| CN102984049A (en) * | 2012-11-26 | 2013-03-20 | 北京奇虎科技有限公司 | Client and method for user group partition and information transfer according to subjects |
| CN103678672A (en) * | 2013-12-25 | 2014-03-26 | 北京中兴通软件科技股份有限公司 | Method for recommending information |
| CN103686236A (en) * | 2013-11-19 | 2014-03-26 | 乐视致新电子科技（天津）有限公司 | Method and system for recommending video resource |
| CN104239450A (en) * | 2014-09-01 | 2014-12-24 | 百度在线网络技术（北京）有限公司 | Search recommending method and device |
-
2015
- 2015-04-28 CN CN201510208472.0A patent/CN106156250A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110010363A1 (en) * | 2009-07-08 | 2011-01-13 | Sony Corporation | Information processing apparatus, information processing method, and program |
| CN102984049A (en) * | 2012-11-26 | 2013-03-20 | 北京奇虎科技有限公司 | Client and method for user group partition and information transfer according to subjects |
| CN103686236A (en) * | 2013-11-19 | 2014-03-26 | 乐视致新电子科技（天津）有限公司 | Method and system for recommending video resource |
| CN103678672A (en) * | 2013-12-25 | 2014-03-26 | 北京中兴通软件科技股份有限公司 | Method for recommending information |
| CN104239450A (en) * | 2014-09-01 | 2014-12-24 | 百度在线网络技术（北京）有限公司 | Search recommending method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111400577A (en) * | 2018-12-14 | 2020-07-10 | 阿里巴巴集团控股有限公司 | Search recall method and device |
| CN111400577B (en) * | 2018-12-14 | 2023-06-30 | 阿里巴巴集团控股有限公司 | Search recall method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sun et al. | Multivis: Content-based social network exploration through multi-way visual analysis | |
| Rajawat et al. | Suspicious big text data analysis for prediction—on darkweb user activity using computational intelligence model | |
| JP5092165B2 (en) | Data construction method and system | |
| Gong et al. | Novel heuristic density-based method for community detection in networks | |
| Pan et al. | Clustering of designers based on building information modeling event logs | |
| CN108763496A (en) | A kind of sound state data fusion client segmentation algorithm based on grid and density | |
| CN103778206A (en) | Method for providing network service resources | |
| Valero-Mas et al. | On the suitability of Prototype Selection methods for kNN classification with distributed data | |
| Huang et al. | Technology–function matrix based network analysis of cloud computing | |
| Suthar et al. | A survey of web usage mining techniques | |
| CN106649380A (en) | Hot spot recommendation method and system based on tag | |
| Sun et al. | Graph force learning | |
| CN106649262B (en) | Method for protecting sensitive information of enterprise hardware facilities in social media | |
| Mehrotra et al. | Comparative analysis of K-Means with other clustering algorithms to improve search result | |
| Bai et al. | A rumor detection model incorporating propagation path contextual semantics and user information | |
| Gamgne Domgue et al. | Community structure extraction in directed network using triads | |
| Dhoot et al. | Efficient Dimensionality Reduction for Big Data Using Clustering Technique | |
| Sundari et al. | A study of various text mining techniques | |
| Gupta et al. | Feature selection: an overview | |
| Rashid et al. | Unlocking the Power of Social Networks with Community Detection Techniques for Isolated and Overlapped Communities: A Review | |
| CN106156259A (en) | A kind of user behavior information displaying method and system | |
| Ye et al. | An interpretable mechanism for personalized recommendation based on cross feature | |
| CN109062551A (en) | Development Framework based on big data exploitation command set | |
| CN106156250A (en) | A kind of search focus recommendation method and system | |
| Meng et al. | Adaptive resonance theory (ART) for social media analytics |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161123 |
|
| RJ01 | Rejection of invention patent application after publication |