CN111143694A - Information pushing method and device, storage equipment and program - Google Patents
Information pushing method and device, storage equipment and program Download PDFInfo
- Publication number
- CN111143694A CN111143694A CN201911423117.XA CN201911423117A CN111143694A CN 111143694 A CN111143694 A CN 111143694A CN 201911423117 A CN201911423117 A CN 201911423117A CN 111143694 A CN111143694 A CN 111143694A
- Authority
- CN
- China
- Prior art keywords
- information
- user
- similarity
- intelligence
- semantic
- 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
Images
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/335—Filtering based on additional data, e.g. user or group profiles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
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)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to the technical field of knowledge pushing, in particular to an information pushing method, an information pushing device, a storage device and a program, wherein the method comprises the following steps: collecting user background information and the information preference degree of a user and acquiring recommended information of an information library; carrying out feature vectorization representation on the user background information and the recommendation information through feature words, and obtaining a potential candidate push set after calculating similarity according to the vector representation of the user background information and the recommendation information; semantic analysis is carried out on the recommended information, the similarity is calculated by cooperating with the information preference degree of the user, and a semantic candidate set which accords with the interest preference of the user is pushed; and distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result. The information pushing method provided by the invention effectively relieves the cold start problem and the sparse problem in knowledge pushing.
Description
Technical Field
The invention relates to the technical field of knowledge pushing, in particular to an information pushing method, an information pushing device, a storage device and a program.
Background
With the development of information technology, a large amount of information data is output, so that the problem of information overload is caused, and how to find high-quality data from mass information becomes a great challenge. The information retrieval can be realized according to the display of related information searched by keywords, the matching result is carried out according to the single characteristic of the keywords and the information content, but the results obtained by searching the same keyword by people with different professional backgrounds are the same, so that how to quickly find the information which is interesting and wanted to be known by a user is a great challenge, the information push well realizes the process, the personalized push of the user is generated by combining the user characteristic and the information content semantic information, the working efficiency of science and technology personnel is greatly improved, the time cost of the self-searching, inquiring and screening processes is saved, and the recommendation technology development of audio, video, text and electronic commerce brings great convenience to the life of people. Therefore, the need of better mining user characteristics and accurately representing information semantic information becomes a hot problem in the current knowledge push field.
Currently, the recommendation technologies commonly used in the knowledge push field mainly include content-based recommendation, collaborative filtering-based recommendation, and model-based recommendation. The above proposed techniques each have certain drawbacks:
the content-based recommendation method cannot extract the potential interests of the user and only depends on the past preference of the user for some recommended objects; the recommendation method based on the coordination filtering only depends on the interaction behavior of the user and the recommendation object, and has the problems of cold start and data sparseness; model-based recommendations are usually made by dimension reduction of the original interaction matrix into two low-dimensional matrices by matrix decomposition, with cold start and data sparseness also present.
Disclosure of Invention
In order to solve the technical problems, the invention provides an intelligence pushing method, an intelligence pushing device, a storage device and an intelligence pushing program, which not only can dig out the potential interest of a user, but also effectively relieve the problems of cold start and data sparseness.
In order to achieve the technical effects, the invention comprises the following technical scheme: in a first aspect, the present invention provides an intelligence push method, including the following steps:
collecting user background information and the information preference degree of a user and acquiring recommended information of an information library;
performing feature vectorization representation on the user background information and the recommendation information through feature words, and calculating similarity according to the vector representation of the user background information and the recommendation information to obtain a potential candidate push set;
semantic analysis is carried out on the recommended information, similarity is calculated by cooperating with information preference degree of the user, and a semantic candidate set which accords with interest preference of the user is pushed;
distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result;
and pushing the obtained pushing result to the user.
In a second aspect, the present invention provides an information pushing device,
an information acquisition unit: the system is used for collecting user background information and the intelligence preference degree of a user and acquiring recommended intelligence information of an intelligence library;
the first calculation unit: the system is used for performing feature vectorization representation on the user background information and the recommendation information through feature words, and obtaining a potential candidate push set after calculating similarity according to the vector representation of the user background information and the recommendation information;
a second calculation unit: the semantic analysis module is used for performing semantic analysis on the recommended information and calculating similarity in cooperation with information preference degree of the user, and pushing a semantic candidate set which accords with interest preference of the user;
a third calculation unit: the system is used for distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result;
a pushing unit: for pushing the obtained push result to the user.
In a third aspect, the present invention provides a computer-readable storage medium, having stored therein instructions, which, when run on a terminal device, cause the terminal device to execute the above-mentioned intelligence push method.
In a fourth aspect, the present invention provides a computer program product, which, when run on a terminal device, causes the terminal device to perform the above-mentioned knowledge push method.
By adopting the technical scheme, the method has the following beneficial effects: the information pushing method provided by the invention matches the user characteristics with the data of the information base, calculates potential candidate items in advance, and then carries out semantic analysis on the previous interesting information content according to the historical behaviors of the user so as to push the information content conforming to the interesting characteristics of the user; and finally, carrying out weight proportion distribution on the candidate sets screened by the two sets to obtain a final pushing result. The method and the system utilize the semantic information of the user interest and the information content to recommend, can quickly retrieve the candidate set of possible results, and obtain the final result according to the linear weight, thereby effectively relieving the cold start problem and the sparse problem in knowledge push.
Drawings
Fig. 1 is a general flowchart of an intelligence push method according to an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail below with reference to specific embodiments and with reference to the attached drawings.
Example (b):
the embodiment of the invention can be suitable for pushing the information in the intelligence base (knowledge base) to a specific technical staff (user condition), and the method can be executed by an intelligence pushing device which is executed by software and/or hardware and can be generally integrated in the intelligence pushing device. The information pushing method specifically comprises the following steps:
step 1): collecting user background information and the information preference degree of a user and acquiring recommended information of an information library;
user background information, such as research directions of user technologists, frequently searched labels and scientific research experiences participated in or mainly participated in, is collected, and the like, and the preference degree of the related technologists on partial information in the information base is researched.
Such as: u09 user sets up the question-answering system to the internal information of enterprise and public institution for the question-answering system research of specific scene, provides convenient and fast information inquiry for it, provides information support for it.
['167','197','220','683','757','758']
Step 2): performing feature vectorization representation on the user background information and the recommendation information through feature words, and calculating similarity according to the vector representation of the user background information and the recommendation information to obtain a potential candidate push set, which specifically comprises the following steps:
step 2.1) carrying out feature vectorization on the user background information through feature extraction, wherein the formula is as follows:
combining all user background information with word vectors through word frequency weight to obtain user characteristic vectors, wherein m represents the characteristic number in user documents, wordiVector representation, word frequency tf (word), representing a characteristic wordi) Meaning wordiThe number of occurrences in a document is divided by the total number of terms in the document.
Step 2.2) extracting characteristic words from the recommendation information and performing vectorization expression, wherein the formula is as follows:
combining each information in the information base with the word vector through the word frequency weight to obtain the information characteristic vector, wherein n represents a certain information characteristic number, wordiVector representation, word frequency tf (word), representing a characteristic word ii) Meaning wordiNumber of occurrences in a document divided byIn terms of the total number of words of the document.
And 2.3) performing cosine similarity calculation on the vectors obtained in the two steps to obtain a candidate push set.
Calculating the distance between the user characteristic vector and the information characteristic vector through cosine similarity to obtain a candidate push set uu,pRepresenting the semantic similarity of the user and the literature, namely the preference value of the user to the literature. u, p represent the vector representation of the user and the document, respectively.
Step 3): semantic analysis is carried out on the recommended information, the similarity is calculated by cooperating with the information preference degree of the user, and a semantic candidate set conforming to the interest preference of the user is pushed, and the method specifically comprises the following steps:
calculating semantic similarity between recommended information through a WMD (word movedistance) model, and cooperatively calculating the similarity according to the semantic similarity and the information preference degree of a user. WMD means word-shift distance, and is a way to measure the distance between two text documents, and is used to determine the similarity between two texts, i.e. the larger the WMD distance, the smaller the similarity, the larger the text similarity (it should be noted that the larger the distance, the farther the distance is, the smaller the similarity is).
In this embodiment, the WMD measures the distance between two text documents, i.e. the minimum distance required for a set of weighted feature words in a document to reach another document. On the basis, WMD word shift distance is utilized to calculate document similarity, semantic similarity between information contents is enhanced, different expression modes of similar semantics are identified, semantic enhancement is realized, a semantic similarity matrix between information libraries can be obtained, and the dimension of the semantic similarity matrix is the total data quantity of the information libraries. The method for calculating the similarity cooperatively according to the semantic similarity and the intelligence preference degree of the user specifically comprises the following steps:
step 3.1) acquiring a project collection according to the collected information preference degree of the user, acquiring a similar information collection according to the semantic similarity acquired by calculation, and sorting in a descending order according to the length of intersection;
d (u, K) ═ n (u) ∩ S (j, K) equation (one)
Wherein N (u) represents a collection of items with high interest level of user u; s (j, K) is K sets of similar intelligence by calculating relative distance values between the intelligence. Sorting according to the length of the intersection D (u, K) in descending order.
Through the preference degree of the user on the intelligence collected in the step 1), a primary ranking candidate set is obtained according to the intelligence content with high preference degree. For example, for a certain user, based on the partial information id 197, 220, 683, 757, 758 that the user has seen, the interest degree is 5 (assuming that the preference degree is set to 1-5, the degree is increased in sequence), k pieces of information sets most similar to the information sets are calculated to be deduplicated and the candidate set obtained by the primary screening is returned.
N(u)＝set(197，220，683，757，758)
Such as: the similar intelligence set at reference number 770 is:
S(770,30)＝(756，586，604，131，994，800，782，775，764，729，618，583，758，241，805，778，155，759，477，764，611，769，723，808，161，571，606，787，579，682)
step 3.2) carrying out secondary screening according to the information content of intersection; and performing secondary fine sequencing on the candidate sets with consistent intersection number through the corresponding preference degree characteristic value.
Fine-grained prediction is again performed for intelligence with equal intersection length by the following formula.
Wherein Du represents the document intersection set, sim (d) obtained by formula (one)p,dq) Representing the similarity of documents p and q calculated by a WMD model; scoreu,qIndicates the preference of the intelligence person u for the document q, vu,pAnd representing the predicted value of the preference of the user for the document.
Step 4): distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result, which specifically comprises the following steps: the similarity of the user and the intelligence obtained based on the user background information and the similarity obtained based on the information semantic content filtering are linearly fused, the fusion function is as follows,
Zu,p＝uu,p×α+vu,p×(1-α)
wherein u isu,pRepresenting the similarity, v, of the user and intelligence obtained based on the user context informationu,pRepresenting the similarity obtained by filtering based on the information semantic content, and the proportion of the two in the calculation mode of the fusion similarity is adjusted by a parameter α, and the value range is [0,1]，Zu,pRepresenting the recommendation result obtained by the mixed similarity calculation with the weight value of α.
Step 5): and pushing the obtained pushing result to the user.
Example two:
the embodiment provides an intelligence pushing device, which comprises an information acquisition unit, a first calculation unit, a second calculation unit, a third calculation unit and a pushing unit, wherein the first calculation unit is used for collecting user background information and intelligence preference degree of a user and acquiring recommended intelligence information of an intelligence library; the first computing unit is used for performing feature vectorization representation on the user background information and the recommendation information through feature words, and obtaining a potential candidate push set after similarity is computed according to vector representation of the user background information and the recommendation information; the second calculation unit is used for performing semantic analysis on the recommended information, calculating similarity in cooperation with information preference degree of a user and pushing a semantic candidate set which accords with interest preference of the user; the third computing unit is used for distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result; the pushing unit is used for pushing the obtained pushing result to the user.
In the above embodiment, the information obtaining unit may include a user background information collecting unit, a user interest feature analyzing unit, and an intelligence information obtaining unit, where the user background information collecting unit is configured to collect user background information, such as research directions of user science and technology personnel, frequently retrieved tags, and scientific research experiences involved or mainly involved in the tags; the user interest characteristic analysis unit is used for researching the preference degrees of related technologists to partial information in the information base, wherein the interest degrees are all 5 (assuming that the preference degrees are set to be 1-5, the preference degrees are sequentially increased); the intelligence information acquisition unit is used for acquiring intelligence documents related to the characteristic words in the intelligence library. The first calculating unit may include a feature vectorization unit configured to perform feature vectorization on the user background information through feature extraction and extract feature words from the recommendation information and perform vectorization representation, and a cosine similarity calculating unit configured to perform cosine similarity calculation on the vectors obtained in the last two steps to obtain the candidate push set.
Example three:
an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements an intelligence pushing method provided by an embodiment of the present invention, and the method includes: collecting user background information and the information preference degree of a user and acquiring recommended information of an information library; performing feature vectorization representation on the user background information and the recommendation information through feature words, and calculating similarity according to the vector representation of the user background information and the recommendation information to obtain a potential candidate push set; semantic analysis is carried out on the recommended information, similarity is calculated by cooperating with information preference degree of the user, and a semantic candidate set which accords with interest preference of the user is pushed; distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result; and pushing the obtained pushing result to the user.
Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. An intelligence push method, characterized in that the method comprises the steps of:
collecting user background information and the information preference degree of a user and acquiring recommended information of an information library;
performing feature vectorization representation on the user background information and the recommendation information through feature words, and calculating similarity according to the vector representation of the user background information and the recommendation information to obtain a potential candidate push set;
semantic analysis is carried out on the recommended information, similarity is calculated by cooperating with information preference degree of the user, and a semantic candidate set which accords with interest preference of the user is pushed;
distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result;
and pushing the obtained pushing result to the user.
2. The intelligence push method according to claim 1, wherein the user context information and the recommended intelligence information are subjected to feature vectorization representation by feature words, and a potential candidate push set is obtained after similarity is calculated according to the vector representation of the user context information and the recommended intelligence information, specifically comprising:
(1) carrying out feature vectorization on the user background information through feature extraction;
(2) extracting characteristic words from the recommended information and performing vectorization representation;
(3) and performing cosine similarity calculation on the vectors obtained in the two steps to obtain a candidate push set.
3. The intelligence push method of claim 1, wherein performing semantic analysis on the recommended intelligence information and pushing a semantic candidate set meeting user interest preference in cooperation with intelligence preference degree of a user specifically comprises:
calculating semantic similarity between recommended information through WMD model, and calculating similarity according to semantic similarity and user information preference degree.
4. The intelligence push method of claim 3, wherein the collaborative computation of the similarity according to the semantic similarity and the intelligence preference degree of the user specifically comprises:
acquiring a project collection according to the collected information preference degree of the user, acquiring a similar information collection according to the semantic similarity acquired by calculation, and sorting in a descending order according to the length of intersection;
performing secondary screening according to the information content of the intersection; and performing secondary fine sequencing on the candidate sets with consistent intersection number through the corresponding preference degree characteristic value.
5. The intelligence push method according to claim 1, wherein the weight proportion distribution of the potential candidate push set and the semantic candidate set is performed to obtain a push result, and specifically: the similarity of the user and the intelligence obtained based on the user background information and the similarity obtained based on the information semantic content filtering are linearly fused, the fusion function is as follows,
Zu,p＝uu,p×α+vu,p×(1-α)
wherein u isu,pRepresenting the similarity, v, of the user and intelligence obtained based on the user context informationu,pRepresenting the similarity obtained by filtering based on the information semantic content, and the proportion of the two in the calculation mode of the fusion similarity is adjusted by a parameter α, and the value range is [0,1]，Zu,pRepresenting the recommendation result obtained by the mixed similarity calculation with the weight value of α.
6. An information pushing device is characterized in that,
an information acquisition unit: the system is used for collecting user background information and the intelligence preference degree of a user and acquiring recommended intelligence information of an intelligence library;
the first calculation unit: the system is used for performing feature vectorization representation on the user background information and the recommendation information through feature words, and obtaining a potential candidate push set after calculating similarity according to the vector representation of the user background information and the recommendation information;
a second calculation unit: the semantic analysis module is used for performing semantic analysis on the recommended information and calculating similarity in cooperation with information preference degree of the user, and pushing a semantic candidate set which accords with interest preference of the user;
a third calculation unit: the system is used for distributing the weight proportion of the potential candidate push set and the semantic candidate set to obtain a push result;
a pushing unit: for pushing the obtained push result to the user.
7. A computer-readable storage medium having stored therein instructions which, when run on a terminal device, cause the terminal device to perform the intelligence push method of any of claims 1-5.
8. A computer program product, characterized in that the computer program product, when run on a terminal device, causes the terminal device to perform the knowledge push method of any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911423117.XA CN111143694B (en) | 2019-12-31 | 2019-12-31 | Information pushing method and device, storage device and program |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911423117.XA CN111143694B (en) | 2019-12-31 | 2019-12-31 | Information pushing method and device, storage device and program |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111143694A true CN111143694A (en) | 2020-05-12 |
| CN111143694B CN111143694B (en) | 2023-06-16 |
Family
ID=70523062
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911423117.XA Active CN111143694B (en) | 2019-12-31 | 2019-12-31 | Information pushing method and device, storage device and program |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111143694B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170161618A1 (en) * | 2015-12-08 | 2017-06-08 | Adobe Systems Incorporated | Attribute weighting for media content-based recommendation |
| CN110188208A (en) * | 2019-06-04 | 2019-08-30 | 河海大学 | A kind of the information resources inquiry recommended method and system of knowledge based map |
| CN110399563A (en) * | 2019-08-05 | 2019-11-01 | 中国人民解放军军事科学院评估论证研究中心 | A kind of preference pattern and customized information active recommended method based on user behavior analysis |
-
2019
- 2019-12-31 CN CN201911423117.XA patent/CN111143694B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170161618A1 (en) * | 2015-12-08 | 2017-06-08 | Adobe Systems Incorporated | Attribute weighting for media content-based recommendation |
| CN110188208A (en) * | 2019-06-04 | 2019-08-30 | 河海大学 | A kind of the information resources inquiry recommended method and system of knowledge based map |
| CN110399563A (en) * | 2019-08-05 | 2019-11-01 | 中国人民解放军军事科学院评估论证研究中心 | A kind of preference pattern and customized information active recommended method based on user behavior analysis |
Non-Patent Citations (1)
| Title |
|---|
| 李朔;石宇良;: "基于位置社交网络中地点聚类推荐方法", 山东大学学报(工学版) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111143694B (en) | 2023-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109885773B (en) | Personalized article recommendation method, system, medium and equipment | |
| US9465872B2 (en) | Segment sensitive query matching | |
| CN108280114B (en) | Deep learning-based user literature reading interest analysis method | |
| CN104899322A (en) | Search engine and implementation method thereof | |
| CN102243647A (en) | Extracting higher-order knowledge from structured data | |
| MX2013005056A (en) | Multi-modal approach to search query input. | |
| CN111444304A (en) | Search ranking method and device | |
| CN111475725A (en) | Method, apparatus, device, and computer-readable storage medium for searching for content | |
| Lee et al. | Efficient image retrieval using advanced SURF and DCD on mobile platform | |
| CN103744887A (en) | Method and device for people search and computer equipment | |
| CN113806588A (en) | Method and device for searching video | |
| JP7395377B2 (en) | Content search methods, devices, equipment, and storage media | |
| CN115374781A (en) | Text data information mining method, device and equipment | |
| CN112417133A (en) | Training method and device of ranking model | |
| Haak et al. | Auditing search query suggestion bias through recursive algorithm interrogation | |
| Telles et al. | Normalized compression distance for visual analysis of document collections | |
| KR101753768B1 (en) | A knowledge management system of searching documents on categories by using weights | |
| KR101651963B1 (en) | Method of generating time and space associated data, time and space associated data generation server performing the same and storage medium storing the same | |
| CN112100216A (en) | Creative keyword processing method and device | |
| CN114282119B (en) | Scientific and technological information resource retrieval method and system based on heterogeneous information network | |
| CN111143694B (en) | Information pushing method and device, storage device and program | |
| CN114218259A (en) | Multi-dimensional scientific information search method and system based on big data SaaS | |
| Zhang et al. | Interactive mobile visual search for social activities completion using query image contextual model | |
| CN113763084A (en) | Product recommendation processing method, device, equipment and storage medium | |
| CN117648444B (en) | Patent clustering method and system based on graph convolution attribute aggregation |
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 |