CN107180111A - A kind of information recommendation method, electronic equipment, storage medium and system - Google Patents
A kind of information recommendation method, electronic equipment, storage medium and system Download PDFInfo
- Publication number
- CN107180111A CN107180111A CN201710444314.4A CN201710444314A CN107180111A CN 107180111 A CN107180111 A CN 107180111A CN 201710444314 A CN201710444314 A CN 201710444314A CN 107180111 A CN107180111 A CN 107180111A
- Authority
- CN
- China
- Prior art keywords
- keyword
- information
- keywords
- user
- recommendation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Abstract
The invention discloses a kind of information recommendation method, electronic equipment, storage medium and system, this method includes:Criterion material is obtained, and extracts some keyword generation keywords pair, the associated ratings and weight of keyword pair is obtained, obtains keyword to Classifying Sum to keyword to storehouse;The keyword of user's request is obtained, is matched according to user's keyword in keyword is to storehouse and obtains new keywords set, recommendation information is obtained according to new keywords index of set related information and user terminal is pushed to.New keywords are matched in keyword is to storehouse by the keyword according to user's request, obtain new keywords set, recommendation information is obtained according to new keywords index of set related information and user terminal is pushed to, user is set to obtain the recommendation information of the keyword associated with the keyword of demand, making the relevant information of search keyword does not have limitation, makes the scattered key word information related to the keyword of user's request throughout be easier to be found by user.
Description
Technical field
The present invention relates to technical field of Internet information, more particularly to a kind of information recommendation method, electronic equipment, storage Jie
Matter and system.
Background technology
With the development of information technology and internet, information is changed into overload, the producers and consumers of information from scarcity
All it is difficult to find the information required for oneself from substantial amounts of information.User wants to obtain information, it is necessary to actively in a search engine
Suitable keyword is inputted, the information of oneself needs is found with this.
But, when the keyword of all relevant informations can not be depicted in user, user wants to obtain by search keyword
Scattered relevant information throughout is just helpless.Particularly, towards areas of expertise, user lacks corresponding knowledge storage
It is standby, it is impossible to judge to whether there is correlation between information, also can not just input suitable keyword.Therefore conventional search keyword
Relevant information there is certain limitation, it is impossible to solve the problem of scattered relevant information throughout is not easy to be found.
The content of the invention
In order to overcome the deficiencies in the prior art, an object of the present invention is to provide a kind of information recommendation method, its energy
Solving the relevant information of conventional search keyword has certain limitation, it is impossible to solves scattered relevant information throughout and does not allow
The problem of being easily found.
The second object of the present invention is offer a kind of electronic equipment, and it can solve the relevant information of conventional search keyword
With certain limitation, it is impossible to solve the problem of scattered relevant information throughout is not easy to be found.
The third object of the present invention is to provide a kind of computer-readable recording medium, and it can solve conventional search keyword
Relevant information there is certain limitation, it is impossible to solve the problem of scattered relevant information throughout is not easy to be found.
The fourth object of the present invention is to provide a kind of information recommendation system, and it can solve the correlation of conventional search keyword
Information has certain limitation, it is impossible to solve the problem of scattered relevant information throughout is not easy to be found.
An object of the present invention is realized using following technical scheme:
A kind of information recommendation method, this method includes:
Obtain criterion material;
Some keywords are extracted in the criterion material, keyword pair is generated according to some keywords, obtains institute
The associated ratings and weight of keyword pair are stated, according to the associated ratings and the weight by the keyword to Classifying Sum,
Keyword is obtained to storehouse;
Obtain the keyword of user's request, i.e. user's keyword;
Some new keywords are matched in the keyword is to storehouse according to user's keyword, new keywords collection is obtained
Close, be integrated into according to the new keywords in default information bank and index related information, obtain recommendation information, and recommend described
Information pushes to user terminal.
Further, it is described that new keywords set is matched including according to institute in keyword is to storehouse according to user's keyword
State user's keyword and carry out Keywords matching first, obtain preliminary keyword set, by user's keyword and described preliminary
Keyword set is incorporated as known keyword set, carries out non-Keywords matching first, obtains pre-output keyword set, sentence
Whether the pre-output keyword set of breaking is empty set, if so, then using the known keyword set as new keywords set,
If it is not, the known keyword set is merged with the pre-output keyword set, new known keyword set is obtained, by institute
New known keyword is stated as known keyword set, repeats the non-Keywords matching first;
The Keywords matching first is that keyword first is matched in keyword is to storehouse according to user's keyword
Set, user's keyword is removed in the keyword set first, preliminary keyword set is obtained;
The non-Keywords matching first is to be integrated into keyword to matching non-head in storehouse according to the known keyword
Secondary keyword set, removes the known keyword set in the non-keyword set first, obtains pre-output keyword
Set.
Further, it is described and by recommendation information push to user terminal include according to the associated ratings of the new keywords and
The sequence of weight is sorted to the recommendation information, and the recommendation information is pushed into user terminal according to the sequence.
The second object of the present invention is realized using following technical scheme:
A kind of electronic equipment, the equipment includes:Processor;
Memory;And program, wherein described program is stored in the memory, and is configured to by processor
Perform, described program includes being used to perform the information recommendation method in the present invention.
The third object of the present invention is realized using following technical scheme:
A kind of computer-readable recording medium, is stored thereon with computer program, and the computer program is held by processor
Information recommendation method in the row present invention.
The fourth object of the present invention is realized using following technical scheme:
A kind of information recommendation system, the system includes:Criterion material acquisition module, keyword are to extraction module, correlation etc.
Level and Weight Acquisition module, keyword are pushed away to Classifying Sum module, customer requirement retrieval module, Keywords matching module, result
Recommend module;
The criterion material acquisition module obtains criterion material, and the criterion material acquisition module sends out the criterion material
The keyword is delivered to extraction module, the keyword extracts some keywords to extraction module in the criterion material,
And keyword pair is generated according to some keywords, the associated ratings and Weight Acquisition module obtain the keyword pair
Associated ratings and weight, the keyword to Classifying Sum module according to the associated ratings and weight by the keyword to point
Class collects, and obtains keyword to storehouse;
The customer requirement retrieval module obtains the keyword of user's request, i.e. user's keyword;The Keywords matching
Module matches new keywords according to user's keyword in the keyword is to storehouse, obtains new keywords set, described
As a result recommending module is integrated into default information bank according to the new keywords and indexes related information, obtains recommendation information, and
The recommendation information is pushed into user terminal.
Further, the associated ratings and Weight Acquisition module include associated ratings acquiring unit and Weight Acquisition list
Member, the associated ratings acquiring unit is used for the associated ratings for obtaining the keyword pair, and the Weight Acquisition unit is used to obtain
Take the weight of the keyword pair.
Further, the result recommending module includes information index unit and push unit, described information indexing units
Recommendation information, the push unit are obtained for being integrated into index related information in default information bank according to the new keywords
For the recommendation information to be pushed into user terminal.
Compared with prior art, the beneficial effects of the present invention are:If by obtaining in criterion material extraction criterion material
Dry keyword generates keyword pair, the associated ratings and weight of keyword pair is obtained, according to associated ratings and weight to keyword
Keyword is obtained to storehouse to carrying out Classifying Sum, and some new passes are matched in keyword is to storehouse according to the keyword of user's request
Keyword, obtains new keywords set, and being integrated into index related information in default information bank according to new keywords obtains recommendation
Breath, pushes to user terminal by recommendation information, user is only inputted the keyword of known demand, just can obtain the pass with demand
The recommendation information of the associated keyword of keyword, making the relevant information of search keyword does not have limitation, makes to disperse throughout
The key word information related to the keyword of user's request is easier to be found by user.
Brief description of the drawings
Fig. 1 is a kind of flow chart one of information recommendation method of the present invention;
Fig. 2 is a kind of flowchart 2 of information recommendation method of the present invention;
Fig. 3 is a kind of block architecture diagram of information recommendation system of the present invention;
Fig. 4 is a kind of module connection figure of information recommendation system of the present invention.
Embodiment
Below, with reference to accompanying drawing and embodiment, the present invention is described further, it is necessary to which explanation is, not
Under the premise of afoul, new implementation can be formed between various embodiments described below or between each technical characteristic in any combination
Example:
A kind of information recommendation method of the present patent application, as shown in Figure 1-2, comprises the following steps:
Step S10:Criterion material is obtained, criterion material can be paper document, patent application and other professional electronics
Library, obtains criterion material quantity n0.
Step S20:Some keywords are extracted in criterion material, keyword pair, such as step are generated according to some keywords
When criterion material quantity in S10 is n0, keyword is extracted from a parts of criterion materials, keyword quantity n1 is obtaineda, a=1,
2 ..., n0, meets n1a≥2;Obtain b-th of keyword k0 of a parts of criterion materials simultaneouslyab, a=1,2 ..., n0, b=1,
2 ..., n1a.To all keyword k0 extracted in a parts of criterion materialsabMatch two-by-two, obtain keyword to quantityC-th of keyword is obtained in a parts of criterion materials simultaneously to p0ac:
p0ac={ k0ad, k0ae,
A=1,2 ..., n0, c=1,2 ..., n2a, d=1,2 ..., n1a-1, e=2,3 ..., n1a, d < e.
Step S30:The associated ratings and weight of keyword pair are obtained, to correlation between the keyword of keyword centering etc.
Level assigns associated ratings according to degree of correlation, associated ratings quantity n3 is obtained, while obtaining f-th of associated ratings lf, f=1,
2 ..., n3, remembers lfCollection be combined into L, to c-th of keyword of a parts of criterion materials to assign associated ratings r0acWith weight w0ac,
New keyword is obtained to p1ac:
p1ac={ k0ad, k0ae, r0ac, w0ac,
r0ac∈ L, a=1,2 ..., n0, c=1,2 ..., n2a, d=1,2 ..., n1a-1, e=2,3 ..., n1a, d
< e.
Step S40:According to associated ratings and weight by keyword to Classifying Sum, keyword is obtained to storehouse, such as step S30
In all keywords to p1ac, classify according to associated ratings, obtain associated ratings quantity n4.To all keywords to p1acAccording to
Associated ratings and keyword are obtained g-th to (the keyword location swap of keyword centering still for same keyword to) classification
The keyword of associated ratings is to classification quantity n5g, g=1,2 ..., n4 further, obtain g-th associated ratings and believed for h-th
Keyword is ceased to p2gh={ k1gi, k1gj, r1g, w1gh, r1g∈ L, g=1,2 ..., n4, h=1,2 ..., n5g.Wherein, exist
G-th of associated ratings, h-th of keyword is in classification, obtaining keyword to quantity n6Gh, g=1,2 ..., n4, h=1,
2 ..., n5g, while obtaining g-th of associated ratings, h-th of keyword to m-th of keyword in classification to weight w2ghm, g=1,
2 ..., n4, h=1,2 ..., n5g, m=1,2 ..., n6gh, and then obtain
Step S50:Obtain the keyword of user's request, i.e. user's keyword;The Given information demand of user's input is obtained,
Obtain keyword, i.e. user's keyword.Obtained keyword quantity is n7.Obtain user's request keyword K2 simultaneouslyq, q=
12 ..., n7.Remember K2qCollection be combined into K2.
Step S60:Some new keywords are matched in keyword is to storehouse according to user's keyword, new keywords collection is obtained
Close, step S60 specifically includes step S61-66:
Step S61:Keywords matching first is carried out according to user's keyword, preliminary keyword set is obtained, it is crucial first
Word is matched:Keyword set first is matched in keyword is to storehouse according to user's keyword, in keyword set first
User's keyword is removed, the preliminary keyword set of non-duplicate removal is obtained, duplicate key is removed in the preliminary keyword set of non-duplicate removal
Word, retains the maximum keyword of respective weights in identical keyword, obtains preliminary keyword set;
Step S62:User's keyword and preliminary keyword set are incorporated as known keyword set;
Step S63:Non- Keywords matching first is carried out, pre-output keyword set, the non-tool of Keywords matching first is obtained
Body is:Keyword is integrated into matching non-keyword set first in storehouse according to known keyword in step S62, it is non-first
Known keyword set is removed in keyword set, non-duplicate removal pre-output keyword set is obtained, it is crucial in non-duplicate removal pre-output
Duplicate key word is removed in set of words, retains the maximum keyword of respective weights in same keyword, obtains pre-output keyword
Set;
Step S64:Judge that pre-output keyword set is no for empty set, if so, continuing step S66, if it is not, continuing step
S65。
Step S65:Known keyword set is merged with pre-output keyword set, new known keyword set is obtained,
Using new known keyword as the known keyword set in step S62, return to step S62 repeats non-keyword first
Matching, so as to continue the judgement in step S64;
Step S66:Then using known keyword set as new keywords set, new keywords set is obtained.
Step S61-66 can be specially according to data in step S50:Using user keyword set K2, r1 is specified in matchingg
All keywords in associated ratings are to p2gh, when obtaining the keyword pair comprising user's keyword set K2, extract keyword pair
In another keyword, obtain keyword set K4 first0.Then from keyword set K4 first0Middle removal user's keyword set
K2 is closed, the preliminary keyword set of non-duplicate removal is obtained, duplicate key word is removed in the preliminary keyword set of non-duplicate removal, is retained identical
Keyword in the maximum keyword of respective weights, obtain preliminary keyword set K50, by user's keyword set K2 and tentatively
Keyword set K50It is incorporated as known keyword set K30.Non- Keywords matching first is carried out, there is known keyword set
K31=K30, the specified r1 of matchinggAll keywords in associated ratings are to p2gh, obtain including known keyword set K31Pass
Keyword pair, extracts another keyword of keyword centering, obtains non-keyword set K4 first1.Then from non-keyword first
Set K41Middle removal known keyword set K31, non-duplicate removal pre-output keyword set is obtained, in non-duplicate removal pre-output keyword
Duplicate key word is removed in set, retains the maximum keyword of respective weights in same keyword, obtains pre-output keyword set
Close K51, whenWhen, obtain new keywords set K6=K31.Otherwise, proceed non-Keywords matching first, have
MS keyword K3s=K3s-1∪K5s-1, s >=2, the specified r1 of matchinggAll keywords in associated ratings are to p2gh, comprising
Known keyword set K3sKeyword pair, extract keyword centering another keyword, obtain non-keyword set first
K4s.Then from non-keyword set K4 firstsMiddle removal known keyword set K3s, obtain non-duplicate removal pre-output keyword set
Close.Duplicate key word is removed in non-duplicate removal pre-output keyword set, retains the maximum pass of respective weights in same keyword
Keyword, obtains pre-output keyword set K5s, untilWhen, stop non-Keywords matching first, obtain new keywords
Set K6=K3s。
Step S70:Index related information in default information bank is integrated into according to new keywords in step S60 to be recommended
Information, and recommendation information is pushed into user terminal, recommendation information is pushed into user terminal is included according to correlation of new keywords etc.
The sequence of level and weight is sorted to the corresponding recommendation information of new keywords, and recommendation information is pushed into user terminal according to sequence.It is logical
The default information bank containing lot of documents and scientific achievement is crossed, is indexed out in default information bank new crucial in step S60
The corresponding information with relevance of set of words K6, user terminal is pushed information to according to the sequence of associated ratings and weight.
A kind of information recommendation method in the present patent application is applied to scientific research information in the first embodiment to be recommended, second
Health and fitness information is applied in embodiment to recommend, and is recommended in the third embodiment applied to educational information, is imitated in the first embodiment
Mark material is the high-quality research methods document delivered;Criterion material is the high-quality medical science delivered in a second embodiment
Document or healthy popular science document;Criterion material is the quality education pertinent literature or education delivered in the third embodiment
Related popular science document;It is described in detail below for a kind of information recommendation method in first embodiment:
It is specific as follows when a kind of information recommendation method is applied to scientific research information recommendation in the first embodiment:
Using the high-quality research methods document delivered as criterion material, criterion material number n0 is obtained, from a parts of materials
Middle extraction keyword obtains keyword quantity n1a, a=1,2 ..., n0 meet n1a>=2, while obtaining a parts of criterion materials
B keyword k0ab, a=1,2 ..., n0, b=1,2 ..., n1a。
For example, one on t examine statistical method document, be extracted keyword " t inspections ", " test of normality ",
" F inspections " and " Wilcoxon rank tests ".
To all keyword k0 extracted in a parts of criterion materialsabMatch two-by-two, obtain keyword to quantityC-th of keyword of a parts of criterion materials is obtained to p0 simultaneouslyac:
p0ac={ k0ad, k0ae,
A=1,2 ..., n0, c=1,2 ..., n2a, d=1,2 ..., n1a-1, e=2,3 ..., n1a, d < e.
For example, keyword " t inspections ", " test of normality ", " F inspections " and " Wilcoxon rank tests " is matched two-by-two,
Obtain keyword is to quantityRespectively { t examine, test of normality }, { t is examined, and F is examined }, t is examined,
Wilcoxon rank tests }, { test of normality, F examine }, { test of normality, Wilcoxon rank tests } and F inspections,
Wilcoxon rank tests }.
Between associated ratings between the keyword of setting keyword centering in advance, the keyword for obtaining keyword centering
Associated ratings, obtain associated ratings quantity for n3, while obtaining f-th of associated ratings lf, f=1,2 ..., n3 remember lfIt must gather
For L.
For example, making associated ratings quantity n3=2 using the degree of correlation of precondition according to statistical method, it is respectively
" having correlation " and " without related ", corresponding value is " 1 " and " 0 ".
C-th of keyword of a parts of criterion materials is obtained to associated ratings r0acWith weight w0ac, obtain new keyword pair
p1ac:
p1ac={ k0ad, k0ae, r0aciw0ac}i
r0ac∈ L, a=1,2 ..., n0, c=1,2 ..., n2a, d=1,2 ..., n1a-1, e=2,3 ..., n1a, d
< e.
For example, obtaining associated ratings and weight to value to above-mentioned 6 keywords of example, 6 new keywords pair are obtained, point
Wei not { t examine, test of normality, 1,1 }, { t is examined, and F is examined, 1,1 }, { t be examined, Wilcoxon rank tests, 0,1 }, { just
State property examine, F examine, 0,1 }, { test of normality, Wilcoxon rank tests, 0,1 } and F examine, Wilcoxon sums of ranks examine
Test, 0,1 }.
According to associated ratings and weight by keyword to Classifying Sum, all keywords are to p1acClassify by associated ratings, obtain
To associated ratings classification quantity n4.All keywords are to p1acAccording to associated ratings and keyword to the (keyword of keyword centering
Location swap still for same keyword to) classification, obtain the keyword of g-th of associated ratings to classification quantity n5G, g=1,
2 ..., n4.Further, g-th of associated ratings, h-th of keyword is obtained to p2gh={ k1gi, k1gj, r1g, w1gh, r1g∈
L,g=1,2 ..., n4, h=1,2 ..., n5g.Wherein, in g-th of associated ratings, h-th of keyword in classification, being closed
Keyword is to quantity n6gh, g=1,2 ..., n4, h=1,2 ..., n5g.Obtain g-th of associated ratings, h-th of information keywords simultaneously
Word is to m-th of keyword in classification to weight w2ghm, g=1,2 ..., n4, h=1,2 ..., n5g, m=1,2 ..., n6gh,
And then obtain
For example, the above-mentioned associated ratings classification quantity n4=2 for having obtained keyword pair, correspondence associated ratings are respectively " 1 "
" 0 ".In " 1 " grade, each according to keyword to Classifying Sum, obtain keyword to classification t examine, test of normality,
1,1 }, { t is examined, and F is examined, 1,1 }, in " 0 " grade, obtain keyword to { t examine, Wilcoxon rank tests, 0,1 },
{ test of normality, F is examined, 0,1 }, { test of normality, Wilcoxon rank tests, 0,1 } and { F inspections, Wilcoxon sums of ranks
Examine, 0,1 }.If in " 1 " grade in classification, there are 3 keywords to { t is examined in " t inspections-test of normality " keyword
Test, test of normality, 1,1,1 keyword is obtained after collecting to weight, and to classification, { t is examined, test of normality, 1, (1+1+
1)/3}。
User's known demand is obtained, obtained keyword quantity is n7.Obtain keyword k2 simultaneouslyq, q=1,2 ..., n7,
Remember k2qCollection be combined into K2.
For example, the known keyword " t inspections " of user's input, obtains user's keyword set { t inspections }.
Using user keyword set K2, r1 is specified in matchinggAll keywords in associated ratings are to p2gh, comprising
During user's keyword set K2 keyword pair, another keyword of keyword centering is extracted, keyword set first is obtained
K40.Then from keyword set K4 first0Middle removal user keyword set K2, obtains the preliminary keyword set of non-duplicate removal.
Duplicate key word is removed in the non-preliminary keyword set of duplicate removal, retains the maximum keyword of respective weights in identical keyword,
Obtain preliminary keyword set K50, by user's keyword set K2 and preliminary keyword set K50It is incorporated as known keyword
Set K30.Non- Keywords matching first is carried out, there is known keyword set K31=K30, the specified r1 of matchinggIn associated ratings
All keywords are to p2gh, obtain including known keyword set K31Keyword pair, extract keyword centering another key
Word, obtains non-keyword set K4 first1.Then from non-keyword set K4 first1Middle removal known keyword set K31, obtain
To non-duplicate removal pre-output keyword set.Duplicate key word is removed in non-duplicate removal pre-output keyword set, retains identical close
The maximum keyword of respective weights, obtains pre-output keyword set K5 in keyword1, whenWhen, obtain new keywords
Set K6=K31.Otherwise, proceed non-Keywords matching first, there is known keyword K3s=K3s-1∪K5s-1, s >=2,
With specified r1gAll keywords in associated ratings are to p2gh, obtain including known keyword set K3sKeyword pair, extract
Another keyword of keyword centering, obtains non-keyword set K4 firsts.Then from non-keyword set K4 firstsIt is middle to remove
Known keyword set K3s, obtain non-duplicate removal pre-output keyword set.Weight is removed in non-duplicate removal pre-output keyword set
Returning to customs keyword, retains the maximum keyword of respective weights in same keyword, obtains pre-output keyword set K5s, untilWhen, stop non-Keywords matching first, obtain new keywords set K6=K3s。
Example 1, when user's keyword set is combined into { t inspections }, carries out Keywords matching first, obtains 2 keywords to { t is examined
Test, test of normality, 1,1 and { t is examined, and F is examined, 1,1 }.Another keyword of keyword centering is extracted, obtains crucial first
Set of words { { test of normality, 1,1 }, { F is examined, 1,1 } }.Then user's keyword set is removed from keyword set first
Close, obtain the preliminary keyword set of non-duplicate removal.Duplicate key word is removed in the preliminary keyword set of non-duplicate removal, retains identical
The maximum keyword of respective weights in keyword, obtain preliminary keyword set { test of normality, 1,1 }, and F is examined, 1,
1}}.User's keyword set and preliminary keyword set are incorporated as known keyword set, { { t inspections }, { normal state is obtained
Property examine, 1,1 }, { F examine, 1,1 }.Non- Keywords matching first is carried out, 4 keywords are obtained to { t is examined, normality inspection
Test, 1,1 }, { t is examined, and F is examined, 1,1 }, { t examine, test of normality, 1,1 } and { t is examined, F inspections, 1,1 }.Remove known
Keyword set { { t inspections }, { test of normality, 1,1 }, { F is examined, 1,1 } }, obtains non-duplicate removal pre-output keyword set.
Duplicate key word is removed in non-duplicate removal pre-output keyword set, retains the maximum key of respective weights in same keyword
Word, obtains pre-output keyword set and is combined into empty set.Stop non-Keywords matching first, obtain new keywords collection and be combined into known key
Set of words { { t inspections }, { test of normality, 1,1 }, { F is examined, 1,1 } }.
Example 2, when user's keyword set is combined into { F inspections }, carries out Keywords matching first, obtains 1 keyword to { t is examined
Test, F is examined, 1,1 }.Another keyword of keyword centering is extracted, keyword set { { t is examined, 1,1 } } first is obtained.Then
User's keyword set is removed from keyword set first, the preliminary keyword set of non-duplicate removal is obtained.Tentatively closed in non-duplicate removal
Duplicate key word is removed in keyword set, retains the maximum keyword of respective weights in identical keyword, obtains preliminary key
Set of words { { t is examined, 1,1 } }.User's keyword set and preliminary keyword set are incorporated as known keyword set, obtained
To { { F inspections }, { t is examined, 1,1 } }.Carry out non-Keywords matching first, obtain 3 keywords to t is examined, and F is examined, 1,
1 }, { t is examined, test of normality, 1,1 } and { t is examined, and F is examined, 1,1 }.Another keyword of keyword centering is extracted, is obtained
Non- keyword set first { t is examined, 1,1 }, { test of normality, 1,1 } and { F is examined, 1,1 }.Remove known keyword set
{ { F inspections }, { t is examined, 1,1 } }, obtains non-duplicate removal pre-output keyword set.In non-duplicate removal pre-output keyword set
Except duplicate key word, retain the maximum keyword of respective weights in same keyword, obtain pre-output keyword set { { normal state
Property examine, 1,1 } }.By known keyword set { { F inspections }, { t is examined, 1,1 } } and pre-output keyword set { { normality
Examine, 1,1 } new known keyword set is incorporated as, proceed non-Keywords matching first, obtain 4 keywords to { t
Examine, F is examined, 1,1 }, { t examine, test of normality, 1,1 }, { t is examined, and F is examined, 1,1 } and t inspections, test of normality,
1,1}.Extract another keyword of keyword centering, obtain non-keyword set first { t examine, 1,1 }, test of normality,
1,1 }, { F is examined, 1,1 } and { t is examined, 1,1 }.Remove known keyword set { { F inspections }, { t is examined, 1,1 }, { normality
Examine, 1,1 }, non-duplicate removal pre-output keyword set is obtained, duplicate key is removed in non-duplicate removal pre-output keyword set
Word, retains the maximum keyword of respective weights in same keyword, obtains pre-output keyword set and be combined into empty set.Stopping is non-first
Keywords matching, obtain new keywords collection be combined into known keyword set { F inspections }, { t examine, 1,1 }, test of normality,
1,1}}。
Example 3, when user's keyword set is combined into { F is examined, test of normality }, carries out Keywords matching first, obtains 2 passes
Keyword is to { t is examined, and F is examined, 1,1.5 } and { t is examined, test of normality, 1,1 }.Another keyword of keyword centering is extracted,
Obtain keyword set { { t is examined, 1,1.5 }, { t is examined, 1,1 } } first.Then user is removed from keyword set first
Keyword set, obtains the preliminary keyword set of non-duplicate removal { { t is examined, 1,1.5 }, { t is examined, 1,1 } }.Tentatively closed in non-duplicate removal
Duplicate key word is removed in keyword set, retains the maximum keyword of respective weights in identical keyword, obtains preliminary key
Set of words { { t is examined, 1,1.5 } }.User's keyword set and preliminary keyword set are incorporated as known keyword set,
Obtain { { F inspections }, { test of normality }, { t is examined, 1,1.5 } }.Non- Keywords matching first is carried out, 4 keywords pair are obtained
{ t inspections, F is examined, 1,1.5 }, { t is examined, test of normality, 1,1 }, { t is examined, and F is examined, 1,1.5 } and { t inspections, normality
Examine, 1,1 }.Another keyword of keyword centering is extracted, non-keyword set first { t is examined, 1,1.5 } is obtained, { t is examined
Test, 1,1 }, { F examine, 1,1.5 } and { test of normality, 1,1 }.Known keyword set is removed, non-duplicate removal pre-output is obtained and closes
Keyword set is empty set.Duplicate key word is removed in non-duplicate removal pre-output keyword set, retains correspondence in same keyword
The maximum keyword of weight, obtains pre-output keyword set and is combined into empty set.Stop non-Keywords matching first, obtain new keywords
Collection is combined into known keyword set { { F inspections }, { test of normality }, { t is examined, 1,1.5 } }.
It is new crucial according to the new keywords set of example 1 { { t inspections }, { test of normality, 1,1 }, { F is examined, 1,1 } } or example 2
Set of words { { F inspections }, { t is examined, 1,1 }, { test of normality, 1,1 } } or the new keywords set of example 3 { { F inspections }, { normality
Examine, { t is examined, 1,1.5 } }, in the default information bank containing a large amount of scientific documents and other e-texts on index
The information that the keyword in set is associated is stated, and is presented to user.
Information recommendation method tool in information recommendation method and first embodiment in second embodiment and 3rd embodiment
Body step is identical, and here is omitted.
As shown in Figure 3-4, a kind of information recommendation system of the present patent application, the system includes:Criterion material acquisition module,
Keyword is to extraction module, associated ratings and Weight Acquisition module, keyword to Classifying Sum module, customer requirement retrieval mould
Block, Keywords matching module, result recommending module;
Criterion material acquisition module obtains criterion material, and criterion material acquisition module sends criterion material to keyword pair
Extraction module, keyword extracts some keywords to extraction module in criterion material, and generates key according to some keywords
Word pair, associated ratings and Weight Acquisition module obtain the associated ratings and weight of keyword pair, and keyword is to Classifying Sum module
According to associated ratings and weight by keyword to Classifying Sum, keyword is obtained to storehouse;
Customer requirement retrieval module obtains the keyword of user's request, i.e. user's keyword;Keywords matching module according to
User's keyword matches new keywords in keyword is to storehouse, obtains new keywords set, and as a result recommending module is according to new pass
Keyword set indexes related information in default information bank, obtains recommendation information, and recommendation information is pushed into user terminal.
Associated ratings and Weight Acquisition module include associated ratings acquiring unit and Weight Acquisition unit, and associated ratings are obtained
Unit is used for the associated ratings for obtaining keyword pair, and Weight Acquisition unit is used for the weight for obtaining keyword pair.
As a result recommending module includes information index unit and push unit, and information index unit is used for according to new keywords collection
Close the index related information in default information bank and obtain recommendation information, push unit is used to recommendation information pushing to user
End.
The a kind of electronic equipment of the present invention, the equipment includes:Processor;
Memory;And program, its Program is stored in memory, and is configured to by computing device, journey
Sequence includes being used to perform the information recommendation method in invention.
A kind of computer-readable recording medium of the present invention, is stored thereon with computer program, and computer program is processed
Device performs the information recommendation method in the present invention.
A kind of information recommendation method and system of the present patent application, if by obtaining in criterion material extraction criterion material
Dry keyword generates keyword pair, the associated ratings and weight of keyword pair is obtained, according to associated ratings and weight to keyword
Keyword is obtained to Classifying Sum to storehouse, some new keys are matched in keyword is to storehouse according to the keyword of user's request
Word, obtains new keywords set, and being integrated into index related information in default information bank according to new keywords obtains recommendation information,
Recommendation information is pushed into user terminal, user is only inputted the keyword of known demand, the key with demand just can be obtained
The recommendation information of the associated keyword of word, making the relevant information of search keyword does not have limitation, make it is scattered throughout with
The related key word information of the keyword of user's request is easier to be found by user.
It will be apparent to those skilled in the art that technical scheme that can be as described above and design, make other various
It is corresponding to change and deformation, and all these change and deformation should all belong to the protection domain of the claims in the present invention
Within.
Claims (8)
1. a kind of information recommendation method, it is characterised in that including:
Obtain criterion material;
Some keywords are extracted in the criterion material, keyword pair is generated according to some keywords, described close is obtained
The associated ratings and weight of keyword pair, obtain the keyword to Classifying Sum according to the associated ratings and the weight
Keyword is to storehouse;
Obtain the keyword of user's request, i.e. user's keyword;
Some new keywords are matched in the keyword is to storehouse according to user's keyword, new keywords set is obtained,
It is integrated into according to the new keywords in default information bank and indexes related information, obtains recommendation information, and by the recommendation
Breath pushes to user terminal.
2. a kind of information recommendation method as claimed in claim 1, it is characterised in that:It is described according to user's keyword in keyword
Include carrying out Keywords matching first according to user's keyword to matching new keywords set in storehouse, obtain preliminary key
Set of words, known keyword set is incorporated as by user's keyword and the preliminary keyword set, and progress is non-first
Keywords matching, obtains pre-output keyword set, and whether judge the pre-output keyword set is empty set, if so, then will
The known keyword set is as new keywords set, if it is not, the known keyword set and the pre-output is crucial
Set of words merges, and obtains new known keyword set, using the new known keyword as known keyword set, repeats
The non-Keywords matching first;
The Keywords matching first is that keyword set first is matched in keyword is to storehouse according to user's keyword,
User's keyword is removed in the keyword set first, preliminary keyword set is obtained;
The non-Keywords matching first is to be integrated into keyword according to the known keyword non-to close first to matching in storehouse
Keyword set, removes the known keyword set in the non-keyword set first, obtains pre-output keyword set.
3. a kind of information recommendation method as claimed in claim 1, it is characterised in that:It is described and recommendation information is pushed into user
End includes sorting to the recommendation information according to the sequence of the associated ratings and weight of the new keywords, by the recommendation information
User terminal is pushed to according to the sequence.
4. a kind of electronic equipment, it is characterised in that including:Processor;
Memory;And program, wherein described program is stored in the memory, and is configured to be held by processor
OK, described program includes being used for the method described in perform claim requirement 1-3 any one.
5. a kind of computer-readable recording medium, is stored thereon with computer program, it is characterised in that:The computer program quilt
Method of the computing device as described in claim 1-3 any one.
6. a kind of information recommendation system, it is characterised in that including:Criterion material acquisition module, keyword are to extraction module, correlation
Grade and Weight Acquisition module, keyword are to Classifying Sum module, customer requirement retrieval module, Keywords matching module, result
Recommending module;
The criterion material acquisition module obtains criterion material, the criterion material acquisition module by the criterion material send to
The keyword is to extraction module, and the keyword extracts some keywords, and root to extraction module in the criterion material
According to some keyword generation keywords pair, the associated ratings obtain the related of the keyword pair to Weight Acquisition module
Grade and weight, the keyword converge the keyword to classification according to the associated ratings and weight to Classifying Sum module
Always, keyword is obtained to storehouse;
The customer requirement retrieval module obtains the keyword of user's request, i.e. user's keyword;The Keywords matching module
New keywords are matched in the keyword is to storehouse according to user's keyword, new keywords set, the result is obtained
Recommending module is integrated into default information bank according to the new keywords and indexes related information, obtains recommendation information, and by institute
State recommendation information and push to user terminal.
7. a kind of information recommendation system as claimed in claim 6, it is characterised in that:The associated ratings and Weight Acquisition module
Including associated ratings acquiring unit and Weight Acquisition unit, the associated ratings acquiring unit is used to obtain the keyword pair
Associated ratings, the Weight Acquisition unit is used for the weight for obtaining the keyword pair.
8. a kind of information recommendation system as claimed in claim 6, it is characterised in that:The result recommending module includes information rope
Draw unit and push unit, described information indexing units, which are used to be integrated into default information bank according to the new keywords, to be indexed
Related information obtains recommendation information, and the push unit is used to the recommendation information pushing to user terminal.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710444314.4A CN107180111B (en) | 2017-06-13 | 2017-06-13 | A kind of information recommendation method, electronic equipment, storage medium and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710444314.4A CN107180111B (en) | 2017-06-13 | 2017-06-13 | A kind of information recommendation method, electronic equipment, storage medium and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107180111A true CN107180111A (en) | 2017-09-19 |
CN107180111B CN107180111B (en) | 2019-10-25 |
Family
ID=59836902
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710444314.4A Active CN107180111B (en) | 2017-06-13 | 2017-06-13 | A kind of information recommendation method, electronic equipment, storage medium and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107180111B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108052666A (en) * | 2017-12-29 | 2018-05-18 | 北京奇虎科技有限公司 | A kind of content recommendation method and device |
CN109242427A (en) * | 2018-08-30 | 2019-01-18 | 出门问问信息科技有限公司 | Contract information archiving method, device, electronic equipment and computer storage medium |
CN111737582A (en) * | 2020-07-29 | 2020-10-02 | 腾讯科技(深圳)有限公司 | Content recommendation method and device |
CN111767378A (en) * | 2020-06-24 | 2020-10-13 | 北京墨丘科技有限公司 | Method and device for intelligently recommending scientific and technical literature |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8452798B2 (en) * | 2009-03-26 | 2013-05-28 | Korea Advanced Institute Of Science And Technology | Query and document topic category transition analysis system and method and query expansion-based information retrieval system and method |
CN103823893A (en) * | 2014-03-11 | 2014-05-28 | 北京大学 | User comment-based product search method and system |
CN105005564A (en) * | 2014-04-17 | 2015-10-28 | 北京搜狗科技发展有限公司 | Data processing method and apparatus based on question-and-answer platform |
-
2017
- 2017-06-13 CN CN201710444314.4A patent/CN107180111B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8452798B2 (en) * | 2009-03-26 | 2013-05-28 | Korea Advanced Institute Of Science And Technology | Query and document topic category transition analysis system and method and query expansion-based information retrieval system and method |
CN103823893A (en) * | 2014-03-11 | 2014-05-28 | 北京大学 | User comment-based product search method and system |
CN105005564A (en) * | 2014-04-17 | 2015-10-28 | 北京搜狗科技发展有限公司 | Data processing method and apparatus based on question-and-answer platform |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108052666A (en) * | 2017-12-29 | 2018-05-18 | 北京奇虎科技有限公司 | A kind of content recommendation method and device |
CN109242427A (en) * | 2018-08-30 | 2019-01-18 | 出门问问信息科技有限公司 | Contract information archiving method, device, electronic equipment and computer storage medium |
CN111767378A (en) * | 2020-06-24 | 2020-10-13 | 北京墨丘科技有限公司 | Method and device for intelligently recommending scientific and technical literature |
CN111737582A (en) * | 2020-07-29 | 2020-10-02 | 腾讯科技(深圳)有限公司 | Content recommendation method and device |
CN111737582B (en) * | 2020-07-29 | 2020-12-08 | 腾讯科技(深圳)有限公司 | Content recommendation method and device |
Also Published As
Publication number | Publication date |
---|---|
CN107180111B (en) | 2019-10-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Sigurbjörnsson et al. | Flickr tag recommendation based on collective knowledge | |
CN107180111B (en) | A kind of information recommendation method, electronic equipment, storage medium and system | |
Moghaddam et al. | AQA: aspect-based opinion question answering | |
Bouadjenek et al. | Sopra: A new social personalized ranking function for improving web search | |
JP6420268B2 (en) | Image evaluation learning device, image evaluation device, image search device, image evaluation learning method, image evaluation method, image search method, and program | |
Dumani et al. | Quality-aware ranking of arguments | |
Yang | Metadata effectiveness in internet discovery: An analysis of digital collection metadata elements and internet search engine keywords | |
An et al. | Feature selection with rough sets for web page classification | |
Heymann et al. | Contrasting Controlled Vocabulary and Tagging | |
Imhof et al. | Multimodal social book search | |
Williams et al. | Classifying and ranking search engine results as potential sources of plagiarism | |
Guha | Related Fact Checks: a tool for combating fake news | |
Oliveira et al. | Automatic tag suggestion based on resource contents | |
Zhou et al. | Evaluating large-scale distributed vertical search | |
Azpiazu et al. | Is readability a valuable signal for hashtag recommendations? | |
Li et al. | CoPatE: a novel contrastive learning framework for patent embeddings | |
CN112597106A (en) | Document page skipping method and system | |
Nishida et al. | Hierarchical auto-tagging: organizing Q&A knowledge for everyone | |
Abbasi et al. | RichVSM: enRiched vector space models for folksonomies | |
Chauhan et al. | Review of Indexing Techniques in Information Retrieval | |
Pisal et al. | AskUs: An opinion search engine | |
Djuana et al. | Ontology learning from user tagging for tag recommendation making | |
CN109684474B (en) | Method, device, equipment and storage medium for providing written topics | |
Chakraverty et al. | Dual lexical chaining for context based text classification | |
Ming et al. | Vocabulary filtering for term weighting in archived question search |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |