CN106095949A - A kind of digital library's resource individuation recommendation method recommended based on mixing and system - Google Patents
A kind of digital library's resource individuation recommendation method recommended based on mixing and system Download PDFInfo
- Publication number
- CN106095949A CN106095949A CN201610422756.4A CN201610422756A CN106095949A CN 106095949 A CN106095949 A CN 106095949A CN 201610422756 A CN201610422756 A CN 201610422756A CN 106095949 A CN106095949 A CN 106095949A
- Authority
- CN
- China
- Prior art keywords
- user
- data
- resource
- recommendation
- library
- 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
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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a kind of digital library's resource individuation recommendation method recommended based on mixing and system, including following five steps: step 1, user information management and maintenance;Step 2, library resource data are extracted and management;Step 3, library resource content-data is analyzed;Step 4, user behavior data analysis;Step 5, library resource personalized recommendation.The present invention utilizes existing digital information in digital library, user oriented carries out Individual book resource recommendation, for effectively utilize library resource provide a kind of new thinking, and in this way based on, it is proposed that a kind of based on mixing recommend personalized resource recommendation system.Carry out data analysis and resource recommendation by the method, can effectively help the user discover that the books and reference materials that oneself is interested, be effectively improved circulation utilization rate and the user satisfaction of books and reference materials simultaneously, provide effective technical support for State Intellectual innovation.
Description
Technical field
The present invention relates to information technology and Internet technical field, in particular to a kind of number recommended based on mixing
Word Library Resources personalized recommendation method and system.
Background technology
Library resource is the important carrier of knowledge dissemination, many traditional librarys and the Reference Room of enterprises and institutions
Having substantial amounts of physical book data, these data are the sources of a State Intellectual innovation.People are using these resources
During, leave and substantial amounts of borrow the historical datas such as record, along with the modernization of management system, these historical records major parts
Use electronic computer to store and managed.Although physical book resource digitized to be realized, on the one hand because relating to
The various problems probabilities such as copyright are little, on the other hand huge due to radix, the most unrealistic for workload, but such as
What better profit from them, the problem needing in strictly one reality to consider in a hurry.If it is considered that those are substantial amounts of
Through completing digitized historical record, then it can be the thinking utilizing the offer of these conventional entity books resources new.
Along with the progress of computer technology, the most increasing books have used digital form to publish, number
Word publication has had become as the collection emphasis of many librarys, also become many librarys purchase fund takies most
One of part.
Along with China's expanding economy, people are more and more higher to the requirement of cultural life, and country has also put into substantial amounts of money
Gold, for Library, collects the increasing resource such as physical book and digital publication in library, this is also for figure
Management and the circulation of book shop resource bring higher requirement.The utilization rate improving them is in the urgent need to address asking
Topic, is also mission and the task of library, especially the purpose embodying and existing of Library Value.
The Reference Room of library and enterprises and institutions creates substantial amounts of collection situation in running, borrows note
The historical datas such as record, these contents have used computer management, it is achieved that digitized.For digitalization resource, it is
System also saves the information such as substantial amounts of user profile, user behavior, reader-based assessment, but lacks these data in present stage
Effectively analyze and utilize, causing the waste of this part data value.
Summary of the invention
For the problems referred to above, the present invention proposes a kind of Reference Room utilizing digital library and enterprises and institutions
Existing digital information, user oriented carries out the method for personalized resource recommendation, provides one for effectively utilizing library resource
Kind new thinking, and in this way based on, it is proposed that a kind of personalized resource recommendation system recommended based on mixing.By this
Method carries out data analysis and resource recommendation, can effectively help to the user discover that the books and reference materials that oneself is interested, have simultaneously
Effect improves circulation utilization rate and the user satisfaction of books and reference materials, provides effective technical support for State Intellectual innovation.
The present invention is based on existing book management system, as an independent assembly and original system Seamless integration-, maximum
Limit utilizes the present resource of system, it is achieved a general system.
The invention provides a kind of digital library's resource individuation recommendation method recommended based on mixing and system, bag
Include:
Step 1, user information management and maintenance, user is core and the service object of system, in order to be preferably user
Provide personalized service, need subscriber data data to be managed and safeguards;
Step 2, the extraction of library resource data and management, this method carries out integrated with book management system, extracts books
The metadata such as the bibliography of data, allow in system or propose books and reference materials content information in copyright allowed band, for realizing base
Recommendation in content is prepared;
Step 3, the analysis of library resource content-data, library resource content-data is mainly text data, and this method uses
Text data processing method processes, in view of common content-data is mainly Chinese data, this method will mainly in
Literary composition data are carried out;
Step 4, user behavior data analysis, this method is analyzed based on user behavior data and models, by user's
Behavior speculates the interest of user as implicit feedback, before making up based on library resource own content loss of learning not
Foot;
Step 5, library resource personalized recommendation, main consideration recommendation based on books and reference materials content, based on user behavior
Recommendation, and the mixing way of recommendation of the two, draw final recommendation list.
Further improve as the present invention, in step 1, subscriber data data include user Demographic data and
The Borrowing History data of user;
Wherein,
Substantially the information that describes of the Demographic data of user, i.e. individual subscriber, as sex, the age, occupation, address,
Contact method etc., but need not gather the sensitive identity information of any user, can effectively protect privacy of user;
The Borrowing History data of user, by the process of Borrowing History data, it appeared that interest.Our borrowing user
The content of text information of the books and reference materials read is as user interest preference vector.
Further improving as the present invention, in step 2, library resource data include metadata and the books of library resource
The content-data of resource;
Wherein,
Substantially the information that describes of the metadata of library resource, i.e. library resource, such as title, author, publishing house, book number, letter
Jie, keyword etc., can carry out basic content-based recommendation based on the contents of the section;
The content-data of library resource, will allow in system or realize this function in copyright allowed band, if there being figure
Supplementing of book resource content, it is possible to achieve preferably content-based recommendation.
Further improving as the present invention, step 3 specifically includes:
Step 301, participle, the Main Function of participle is to be independent feature by entire chapter document process, and this method will use
Common Chinese word cutting method is carried out;
Step 302, removes stop words, and removing stop words is to delete information useless for classification in document, various special symbol
With character, and some word frequently occurred in a document and words, as " ", " ", the auxiliary word of " obtaining " etc, they are for literary composition
The result of shelves classification cuts little ice, and this kind of word is collectively referred to as " stop words ", and their set is referred to as " disabling vocabulary ";
Step 303, feature selection and feature extraction, in the vector space model method for expressing of text, go out in training set
Existing any vocabulary all will be likely to become the characteristic item representing text, the vector space that the whole characteristic items in training set are constituted
Dimension is at a relatively high, generally, all can reach several ten thousand dimensions.Therefore, dimension reduction method is often adopted before being text classifier study
A kind of method, main purpose is to ensure or improving on the premise of classifier performance, effectively reducing the dimension of vector space.
Dimension reduction method can be divided into again feature selection and feature extraction, and the characteristic set after feature selection is the one of original feature space
Individual subset, and the feature that feature extraction obtains is not the most the feature in original feature space, but the combination of primitive character or
The brand-new feature obtained after person's conversion.This method mainly uses Feature Selection to carry out dimensionality reduction, and accelerates the execution of system.
Step 304, characteristic weighing, characteristic weighing refers to each feature in text data, according to it to classification contribution
The height of degree gives the process of certain weight.This method is mainly used in information retrieval field widely used tf-idf power
Weight;
Step 305, content similarity calculates, and this method mainly uses calculating based on cosine similarity.
Further improving as the present invention, step 5 specifically includes:
Step 501, recommendation based on books and reference materials content first according to the text feature of the Borrowing History data of user to
Amount, the interest vector of structuring user's, secondly the content of text vector according to books and reference materials content is similar to user interest vector
Degree, obtains the content-based recommendation list of user;
Step 502, recommendation based on user behavior, first find its neighbor user for every user, then based on its neighbours
User, estimates user's implicit scores to non-scoring item, and introduces time function, neighbor user in the scoring of different time
Giving different weights, so that targeted customer is marked by the scoring that neighbor user is given in the recent period, the impact estimated is higher than it
The scoring be given before long period.Produce behavior such as user the most in systems, then can utilize the demography of user
Data, searching has similar demographics and learns the neighbor user of feature, estimates user's implicit scores to non-scoring item with it,
Can preferably solve " cold start-up " problem of system;
Step 503, the two mixes the way of recommendation, provides recommendation results list, in conjunction with before based on library resource content and
The critical datas such as the books similarity drawn based on user behavior and implicit scores, matching also produces their mark, by mark
The front L part sorted from big to small, as final recommendation results, forms the recommendation results list of a length of L and returns to user,
I.e. complete to recommend.
Accompanying drawing explanation
Fig. 1 is a kind of individualized resource recommendation side of digital library recommended based on mixing described in the embodiment of the present invention
Method and the flow chart of system;
Fig. 2 is a kind of individualized resource recommendation side of digital library recommended based on mixing described in the embodiment of the present invention
The flow chart of step 3 in method and system;
Fig. 3 is a kind of individualized resource recommendation side of digital library recommended based on mixing described in the embodiment of the present invention
The relation table of the feasible user behavior of step 4 and corresponding implicit scores value in method and system;
Fig. 4 is a kind of individualized resource recommendation side of digital library recommended based on mixing described in the embodiment of the present invention
The flow chart of step 5 in method and system.
Detailed description of the invention
Below by specific embodiment and combine accompanying drawing the present invention is described in further detail.
Embodiment 1, as it is shown in figure 1, a kind of digital library's resource recommended based on mixing of the embodiment of the present invention
Propertyization recommends method and system, including:
Step 1, user information management and maintenance, user is core and the service object of system, in order to be preferably user
Provide personalized service, need subscriber data data to be managed and safeguards;
Subscriber data data include the Demographic data of user and the Borrowing History data of user;
Wherein,
Substantially the information that describes of the Demographic data of user, i.e. individual subscriber, as sex, the age, occupation, address,
Contact method etc., but need not gather the sensitive identity information of any user, can effectively protect privacy of user;
The Borrowing History data of user, by the process of Borrowing History data, it appeared that interest.Our borrowing user
The content of text information of the books and reference materials read is as user interest preference vector.
Step 2, the extraction of library resource data and management, this method carries out integrated with book management system, extracts books
The metadata such as the bibliography of data, allow in system or propose books and reference materials content information in copyright allowed band, for realizing base
Recommendation in content is prepared;
Library resource data include the metadata of library resource and the content-data of library resource;
Wherein,
Substantially the information that describes of the metadata of library resource, i.e. library resource, such as title, author, publishing house, book number, letter
Jie, keyword etc., can carry out basic content-based recommendation based on the contents of the section;
The content-data of library resource, will allow in system or realize this function in copyright allowed band, if there being figure
Supplementing of book resource content, it is possible to achieve preferably content-based recommendation.
Step 3, the analysis of library resource content-data, library resource content-data is mainly text data, and this method uses
Text data processing method processes, in view of common content-data is mainly Chinese data, this method will mainly in
Literary composition data are carried out;
As in figure 2 it is shown, specifically include following steps:
Step 301, participle, the Main Function of participle is to be independent feature by entire chapter document process, and this method will use
Common Chinese word cutting method is carried out;
Step 302, removes stop words, and removing stop words is to delete information useless for classification in document, various special symbol
With character, and some word frequently occurred in a document and words, as " ", " ", the auxiliary word of " obtaining " etc, they are for literary composition
The result of shelves classification cuts little ice, and this kind of word is collectively referred to as " stop words ", and their set is referred to as " disabling vocabulary ";
Step 303, feature selection and feature extraction, in the vector space model method for expressing of text, go out in training set
Existing any vocabulary all will be likely to become the characteristic item representing text, the vector space that the whole characteristic items in training set are constituted
Dimension is at a relatively high, generally, all can reach several ten thousand dimensions.Therefore, dimension reduction method is often adopted before being text classifier study
A kind of method, main purpose is to ensure or improving on the premise of classifier performance, effectively reducing the dimension of vector space.
Dimension reduction method can be divided into again feature selection and feature extraction, and the characteristic set after feature selection is the one of original feature space
Individual subset, and the feature that feature extraction obtains is not the most the feature in original feature space, but the combination of primitive character or
The brand-new feature obtained after person's conversion.This method mainly uses Feature Selection to carry out dimensionality reduction, and accelerates the execution of system.
Step 304, characteristic weighing, characteristic weighing refers to each feature in text data, according to it to classification contribution
The height of degree gives the process of certain weight.This method is mainly used in information retrieval field widely used tf-idf power
Weight;
Step 305, content similarity calculates, and this method mainly uses calculating based on cosine similarity.
Step 4, user behavior data analysis, this method is analyzed based on user behavior data and models, by user's
Behavior speculates the interest of user as implicit feedback, before making up based on library resource own content loss of learning not
Foot;
Produced Various types of data when the behavioral data of user, i.e. user and management system interact, including borrowing note
Record, browse record, retrieval record and collection record etc..Due to management system or the difference of user habit, at present should in reality
The situation explicitly marked library resource in is more rare, and this is unfavorable for the analysis quantifying user behavior.Right
We take implicit scores mechanism, by formulating relatively reasonable strategy, by the non-scoring to library resource common for user for this
Behavior is converted into scoring behavior, thus carries out place mat for analysis afterwards and calculating.Fig. 3 illustrates a kind of feasible user behavior
Relation table with corresponding implicit scores value.User's implicit scores to library resource just can be constructed according to such relation
Matrix R=(ri,j), each r in matrixi,jRepresent the i-th bit user implicit scores to jth part library resource.Can send out
Existing, if user is the highest to the implicit scores of certain part library resource, then it represents that he is the biggest to the interest level of this library resource.
Step 5, library resource personalized recommendation, main consideration recommendation based on books and reference materials content, based on user behavior
Recommendation, and the mixing way of recommendation of the two, draw final recommendation list.
As shown in Figure 4, following steps are specifically included:
Step 501, recommendation based on books and reference materials content, first according to the text feature of the Borrowing History data of user to
Amount, the interest vector of structuring user's, further according to the similarity that the content of text vector of books and reference materials content is vectorial with user interest,
Obtain the content-based recommendation list of user;
Step 502, recommendation based on user behavior, first find its neighbor user for every user.It is to say, we
Can be according to implicit scores matrix R=(ri,j) analyze and take the post as the similarity between two users.Higher with certain user's similarity
Other users, be referred to as this user neighbor user.This method uses Pearson correlation coefficient to calculate the similarity between user
Sim (u, v), its formula is:
Wherein ru,i、rv,iRepresent user u and the v scoring to library resource i respectively,Represent user u and v couple respectively
The average score of all library resources, Iu,vRepresent the library resource set that user u and v marked jointly.
Based on its neighbor user, estimate user to the implicit scores Q of non-scoring item (u, i), its formula is:
Wherein (u v) represents targeted customer u and the similarity of neighbor user v, r to SimviRepresent that books are provided by neighbor user v
The score value of source i,Represent targeted customer u and the neighbor user v average score to all library resources, N respectivelyuRepresent
The neighbor user set of targeted customer u.
Formula also introduces time function, and its formula is:
Wherein t is the time being evaluated, and k is factor of influence.It gives neighbor user not in the scoring of different time
With weight so that the scoring that neighbor user is given in the recent period targeted customer is marked estimation impact be higher than its longer time
The scoring be given before between.
Produce behavior such as user the most in systems, then can utilize the Demographic data of user, find and have with it
There is similar demographics to learn the neighbor user of feature, estimate user's implicit scores to non-scoring item, can preferably solve
" cold start-up " problem of system.
Step 503, the two mixes the way of recommendation, provides recommendation results list, in conjunction with before based on library resource content and
The critical datas such as the books similarity drawn based on user behavior and implicit scores, matching and produce mark Score (u, i), its
Formula is:
Score (u, i)=α P (u, i)+β Q (u, i)
Wherein P (u, i) be the project of borrowing of user u to not borrowing the similarity based on content of project i, (u i) is Q
The user u implicit scores to project i, α, β are factor of influence.
It can easily be seen that if user u is for the mark Score of certain part library resource i, (u, i) the highest, then representing user more has
May like it.Therefore, calculate user u for every library resource i mark Score (u, i) after, by mark from greatly to
The front L part of little sequence, as final recommendation results, forms the recommendation results list of a length of L and returns to user u, i.e. completing
Recommend.
The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, for the skill of this area
For art personnel, the present invention can have various modifications and variations.All within the spirit and principles in the present invention, that is made any repaiies
Change, equivalent, improvement etc., should be included within the scope of the present invention.
Claims (5)
1. the digital library's resource individuation recommendation method recommended based on mixing and system, it is characterised in that including:
Step 1, user information management and maintenance, user is core and the service object of system, in order to preferably provide the user
Personalized service, needs to be managed subscriber data data and safeguard;
Step 2, the extraction of library resource data and management, this method carries out integrated with book management system, extracts books and reference materials
The metadata such as bibliography, allow in system or in copyright allowed band, propose books and reference materials content information, for realizing based on interior
The recommendation held is prepared;
Step 3, the analysis of library resource content-data, library resource content-data is mainly text data, and this method uses text
Data processing method processes, and in view of common content-data is mainly Chinese data, this method will be mainly for Chinese number
According to carry out;
Step 4, user behavior data analysis, this method is analyzed based on user behavior data and models, by the behavior of user
The interest of user is speculated, deficiency based on library resource own content loss of learning before making up as implicit feedback;
Step 5, library resource personalized recommendation, main consider recommendation based on books and reference materials content, based on user behavior push away
Recommend, and the mixing way of recommendation of the two, draw final recommendation list.
The digital library's resource individuation recommendation method recommended based on mixing the most according to claim 1 and system,
It is characterized in that, in step 1, subscriber data data include the Demographic data of user and the Borrowing History data of user;
Wherein,
Substantially the information that describes of the Demographic data of user, i.e. individual subscriber, such as sex, age, occupation, address, contact
Mode etc., but need not gather the sensitive identity information of any user, can effectively protect privacy of user;
The Borrowing History data of user, by the process of Borrowing History data, it appeared that interest.We borrow user's
The content of text information of books and reference materials is as user interest preference vector.
The digital library's resource individuation recommendation method recommended based on mixing the most according to claim 1 and system,
It is characterized in that, in step 2, library resource data include the metadata of library resource and the content-data of library resource;
Wherein,
Substantially the information that describes of the metadata of library resource, i.e. library resource, such as title, author, publishing house, book number, brief introduction, pass
Key words etc., can carry out basic content-based recommendation based on the contents of the section;
The content-data of library resource, will allow in system or realize this function in copyright allowed band, if there being books to provide
Supplementing of source contents, it is possible to achieve preferably content-based recommendation.
The digital library's resource individuation recommendation method recommended based on mixing the most according to claim 1 and system,
It is characterized in that, step 3 specifically includes:
Step 301, participle, the Main Function of participle is to be independent feature by entire chapter document process, and this method will use generally
Chinese word cutting method carry out;
Step 302, removes stop words, and removing stop words is to delete in document for useless information, various special symbol and word of classifying
Symbol, and some word frequently occurred in a document and words, as " ", " ", the auxiliary word of " obtaining " etc, they divide for document
The result of class cuts little ice, and this kind of word is collectively referred to as " stop words ", and their set is referred to as " disabling vocabulary ";
Step 303, feature selection and feature extraction, in the vector space model method for expressing of text, occur in training set
Any vocabulary all will be likely to become the characteristic item representing text, the vector space dimension that the whole characteristic items in training set are constituted
At a relatively high, generally, several ten thousand dimensions all can be reached.Therefore, dimension reduction method be text classifier study before through frequently with
A kind of method, main purpose is on the premise of ensureing or improving classifier performance, effectively reduces the dimension of vector space.Dimensionality reduction
Method can be divided into again feature selection and feature extraction, and the characteristic set after feature selection is a son of original feature space
Collection, and the feature that feature extraction obtains is not the most the feature in original feature space, but the combination of primitive character or change
The brand-new feature obtained after changing.This method mainly uses Feature Selection to carry out dimensionality reduction, and accelerates the execution of system.
Step 304, characteristic weighing, characteristic weighing refers to each feature in text data, according to it to classification percentage contribution
Height give certain weight process.This method is mainly used in information retrieval field widely used tf-idf weight;
Step 305, content similarity calculates, and this method mainly uses calculating based on cosine similarity.
The digital library's resource individuation recommendation method recommended based on mixing the most according to claim 1 and system,
It is characterized in that, step 5 specifically includes:
Step 501, recommendation based on books and reference materials content is first according to the Text eigenvector of the Borrowing History data of user, structure
Make the interest vector of user, secondly according to content of text vector and the similarity of user interest vector of books and reference materials content, obtain
Content-based recommendation list to user;
Step 502, recommendation based on user behavior, first find its neighbor user for every user, then based on its neighbor user,
Estimate user's implicit scores to non-scoring item, and introduce time function, neighbor user is given in the scoring of different time
Different weights, so that targeted customer is marked by the scoring that neighbor user is given in the recent period, to be higher than it longer for the impact estimated
The scoring be given before time.Produce behavior such as user the most in systems, then can utilize the Demographic data of user,
Searching has similar demographics and learns the neighbor user of feature with it, estimates user's implicit scores to non-scoring item, permissible
Preferably solve " cold start-up " problem of system;
Step 503, the two mixes the way of recommendation, provides recommendation results list, in conjunction with before based on library resource content and based on
The critical data such as books similarity that user behavior draws and implicit scores, matching also produces their mark, by mark from greatly
To the front L part of little sequence as final recommendation results, form the recommendation results list of a length of L and return to user, the completeest
Become to recommend.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610422756.4A CN106095949A (en) | 2016-06-14 | 2016-06-14 | A kind of digital library's resource individuation recommendation method recommended based on mixing and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610422756.4A CN106095949A (en) | 2016-06-14 | 2016-06-14 | A kind of digital library's resource individuation recommendation method recommended based on mixing and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106095949A true CN106095949A (en) | 2016-11-09 |
Family
ID=57846486
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610422756.4A Pending CN106095949A (en) | 2016-06-14 | 2016-06-14 | A kind of digital library's resource individuation recommendation method recommended based on mixing and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106095949A (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106599265A (en) * | 2016-12-22 | 2017-04-26 | 国家图书馆 | Multimedia booklist data organization system and method oriented to sharding service |
CN106802915A (en) * | 2016-12-09 | 2017-06-06 | 宁波大学 | A kind of academic resources based on user behavior recommend method |
CN107392812A (en) * | 2017-09-01 | 2017-11-24 | 安徽教育网络出版有限公司 | A kind of education of middle and primary schools resource application service system based on semanteme |
CN107943910A (en) * | 2017-11-18 | 2018-04-20 | 电子科技大学 | A kind of Individual book based on combinational algorithm recommends method |
CN108055279A (en) * | 2017-12-26 | 2018-05-18 | 云栖科技(广州)有限公司 | A kind of book data integrated management service system |
CN108090528A (en) * | 2016-11-22 | 2018-05-29 | 上海阿法迪智能标签系统技术有限公司 | The RFID device and supplying system of Intelligent bookshelf |
CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
CN108763241A (en) * | 2018-03-23 | 2018-11-06 | 上海优景智能科技股份有限公司 | The digital management system of intelligent library |
CN109255520A (en) * | 2018-08-10 | 2019-01-22 | 成都明途科技有限公司 | A kind of data push method based on user job behavior |
WO2019090741A1 (en) * | 2017-11-10 | 2019-05-16 | 深圳市华阅文化传媒有限公司 | Method and apparatus for recommending books for user to read |
CN109766465A (en) * | 2018-12-26 | 2019-05-17 | 中国矿业大学 | A kind of picture and text fusion book recommendation method based on machine learning |
CN109961347A (en) * | 2018-10-19 | 2019-07-02 | 大连九州创智科技有限公司 | A kind of method for managing book information of library based on radio frequency tag technology |
CN110111224A (en) * | 2019-05-21 | 2019-08-09 | 南阳理工学院 | A kind of library management system based on big data |
CN110825974A (en) * | 2019-11-22 | 2020-02-21 | 厦门美柚股份有限公司 | Recommendation system content ordering method and device |
CN110955775A (en) * | 2019-11-11 | 2020-04-03 | 南通大学 | Drawing book recommendation method based on implicit inquiry |
CN111008278A (en) * | 2019-11-22 | 2020-04-14 | 厦门美柚股份有限公司 | Content recommendation method and device |
CN112214686A (en) * | 2020-09-29 | 2021-01-12 | 东北石油大学 | Reader interest preference-based research method for information retrieval personalized recommendation service |
CN112765476A (en) * | 2020-07-27 | 2021-05-07 | 上海斐杰教育科技有限公司 | Parent-child reading resource recommendation service system and method |
CN112818037A (en) * | 2021-02-01 | 2021-05-18 | 上海阿法迪智能数字科技股份有限公司 | Book recommendation system and method |
CN112860991A (en) * | 2021-01-25 | 2021-05-28 | 杭州博享科技有限公司 | Book optimization method and device based on user habits |
CN113065711A (en) * | 2021-04-13 | 2021-07-02 | 河南工程学院 | Book purchasing optimization decision system and book purchasing decision method |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101290626A (en) * | 2008-06-12 | 2008-10-22 | 昆明理工大学 | Text categorization feature selection and weight computation method based on field knowledge |
US20100076983A1 (en) * | 2008-09-08 | 2010-03-25 | Apple Inc. | System and method for playlist generation based on similarity data |
CN104992181A (en) * | 2015-06-29 | 2015-10-21 | 昆明理工大学 | Method for recommending books in real time according to habits of library user |
-
2016
- 2016-06-14 CN CN201610422756.4A patent/CN106095949A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101290626A (en) * | 2008-06-12 | 2008-10-22 | 昆明理工大学 | Text categorization feature selection and weight computation method based on field knowledge |
US20100076983A1 (en) * | 2008-09-08 | 2010-03-25 | Apple Inc. | System and method for playlist generation based on similarity data |
CN104992181A (en) * | 2015-06-29 | 2015-10-21 | 昆明理工大学 | Method for recommending books in real time according to habits of library user |
Non-Patent Citations (4)
Title |
---|
刘晓晓: "基于协同过滤算法在图书馆学的应用", 《商》 * |
孙彦超 等: "基于协同过滤算法的个性化图书推荐系统的研究", 《图书馆理论与实践》 * |
成军: "面向电子商务的协同过滤推荐算法与推荐系统研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
熊拥军: "数字图书馆个性化服务资源推荐模式分析", 《图书馆(LIBRARY)》 * |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108090528A (en) * | 2016-11-22 | 2018-05-29 | 上海阿法迪智能标签系统技术有限公司 | The RFID device and supplying system of Intelligent bookshelf |
CN108090528B (en) * | 2016-11-22 | 2020-11-10 | 上海阿法迪智能标签系统技术有限公司 | RFID device and push system of intelligent bookshelf |
CN106802915A (en) * | 2016-12-09 | 2017-06-06 | 宁波大学 | A kind of academic resources based on user behavior recommend method |
CN106802915B (en) * | 2016-12-09 | 2020-07-28 | 宁波大学 | Academic resource recommendation method based on user behaviors |
CN106599265B (en) * | 2016-12-22 | 2020-02-21 | 国家图书馆 | Multimedia book data organization system and method for fragment service |
CN106599265A (en) * | 2016-12-22 | 2017-04-26 | 国家图书馆 | Multimedia booklist data organization system and method oriented to sharding service |
CN107392812A (en) * | 2017-09-01 | 2017-11-24 | 安徽教育网络出版有限公司 | A kind of education of middle and primary schools resource application service system based on semanteme |
WO2019090741A1 (en) * | 2017-11-10 | 2019-05-16 | 深圳市华阅文化传媒有限公司 | Method and apparatus for recommending books for user to read |
CN107943910A (en) * | 2017-11-18 | 2018-04-20 | 电子科技大学 | A kind of Individual book based on combinational algorithm recommends method |
CN107943910B (en) * | 2017-11-18 | 2021-09-07 | 电子科技大学 | Personalized book recommendation method based on combined algorithm |
CN108055279A (en) * | 2017-12-26 | 2018-05-18 | 云栖科技(广州)有限公司 | A kind of book data integrated management service system |
CN108256119A (en) * | 2018-02-14 | 2018-07-06 | 北京方正阿帕比技术有限公司 | A kind of construction method of resource recommendation model and the resource recommendation method based on the model |
CN108763241A (en) * | 2018-03-23 | 2018-11-06 | 上海优景智能科技股份有限公司 | The digital management system of intelligent library |
CN109255520A (en) * | 2018-08-10 | 2019-01-22 | 成都明途科技有限公司 | A kind of data push method based on user job behavior |
CN109961347A (en) * | 2018-10-19 | 2019-07-02 | 大连九州创智科技有限公司 | A kind of method for managing book information of library based on radio frequency tag technology |
CN109766465A (en) * | 2018-12-26 | 2019-05-17 | 中国矿业大学 | A kind of picture and text fusion book recommendation method based on machine learning |
CN110111224A (en) * | 2019-05-21 | 2019-08-09 | 南阳理工学院 | A kind of library management system based on big data |
CN110955775A (en) * | 2019-11-11 | 2020-04-03 | 南通大学 | Drawing book recommendation method based on implicit inquiry |
CN111008278A (en) * | 2019-11-22 | 2020-04-14 | 厦门美柚股份有限公司 | Content recommendation method and device |
CN110825974A (en) * | 2019-11-22 | 2020-02-21 | 厦门美柚股份有限公司 | Recommendation system content ordering method and device |
CN110825974B (en) * | 2019-11-22 | 2022-06-21 | 厦门美柚股份有限公司 | Recommendation system content ordering method and device |
CN111008278B (en) * | 2019-11-22 | 2022-06-21 | 厦门美柚股份有限公司 | Content recommendation method and device |
CN112765476A (en) * | 2020-07-27 | 2021-05-07 | 上海斐杰教育科技有限公司 | Parent-child reading resource recommendation service system and method |
CN112214686A (en) * | 2020-09-29 | 2021-01-12 | 东北石油大学 | Reader interest preference-based research method for information retrieval personalized recommendation service |
CN112860991A (en) * | 2021-01-25 | 2021-05-28 | 杭州博享科技有限公司 | Book optimization method and device based on user habits |
CN112818037A (en) * | 2021-02-01 | 2021-05-18 | 上海阿法迪智能数字科技股份有限公司 | Book recommendation system and method |
CN113065711A (en) * | 2021-04-13 | 2021-07-02 | 河南工程学院 | Book purchasing optimization decision system and book purchasing decision method |
CN113065711B (en) * | 2021-04-13 | 2023-05-12 | 河南工程学院 | Book purchasing optimization decision system and book purchasing decision method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106095949A (en) | A kind of digital library's resource individuation recommendation method recommended based on mixing and system | |
Bansal et al. | On predicting elections with hybrid topic based sentiment analysis of tweets | |
Wan et al. | Aminer: Search and mining of academic social networks | |
CN103744981B (en) | System for automatic classification analysis for website based on website content | |
CN102982153B (en) | A kind of information retrieval method and device thereof | |
Kano et al. | Coliee-2018: Evaluation of the competition on legal information extraction and entailment | |
Pu et al. | Subject categorization of query terms for exploring Web users' search interests | |
CN105930469A (en) | Hadoop-based individualized tourism recommendation system and method | |
CN104866554B (en) | A kind of individuation search method and system based on socialization mark | |
Mirani et al. | Sentiment analysis of isis related tweets using absolute location | |
CN105468649B (en) | Method and device for judging matching of objects to be displayed | |
Shani et al. | Mining recommendations from the web | |
KR100954842B1 (en) | Method and System of classifying web page using category tag information and Recording medium using by the same | |
Zhu et al. | Real-time personalized twitter search based on semantic expansion and quality model | |
JP2006318398A (en) | Vector generation method and device, information classifying method and device, and program, and computer readable storage medium with program stored therein | |
JP2014085862A (en) | Prediction server, program, and method for predicting number of future comments on prediction target content | |
KR100876214B1 (en) | Apparatus and method for context aware advertising and computer readable medium processing the method | |
KR101652433B1 (en) | Behavioral advertising method according to the emotion that are acquired based on the extracted topics from SNS document | |
Li et al. | A probabilistic topic-based ranking framework for location-sensitive domain information retrieval | |
Zhang et al. | A generic pseudo relevance feedback framework with heterogeneous social information | |
Hou et al. | HITSZ-ICRC at NTCIR-11 Temporalia Task. | |
KR101318843B1 (en) | Blog category classification method and apparatus using time information | |
Lau et al. | Mining the web for business intelligence: Homepage analysis in the internet era | |
Cheng et al. | A model for age and gender profiling of social media accounts based on post contents | |
Zhang | A personalized recommendation algorithm based on text mining |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161109 |