CN110096482B - Data analysis method and device - Google Patents
Data analysis method and device Download PDFInfo
- Publication number
- CN110096482B CN110096482B CN201910383032.7A CN201910383032A CN110096482B CN 110096482 B CN110096482 B CN 110096482B CN 201910383032 A CN201910383032 A CN 201910383032A CN 110096482 B CN110096482 B CN 110096482B
- Authority
- CN
- China
- Prior art keywords
- label
- information
- weight
- multimedia file
- acquiring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
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
- G06F16/10—File systems; File servers
- G06F16/16—File or folder operations, e.g. details of user interfaces specifically adapted to file systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a data analysis method and a data analysis device, relates to the field of internet, and aims to solve the problem that in the prior art, the accuracy rate of determining the field of interest of a publisher only by selecting a label is low. The technical scheme provided by the embodiment of the invention comprises the following steps: classifying each multimedia file of the publisher account respectively to obtain at least one information label and corresponding information weight of each multimedia file; acquiring a middle layer label and a middle layer weight according to at least one information label and a corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library; acquiring a display layer label and a display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library; and obtaining the interest classification of the publisher account according to the display layer label and the display layer weight. The method can be applied to personalized information services such as medium popularization, network marketing, e-commerce recommendation and personalized information retrieval.
Description
Technical Field
The invention relates to the field of internet, in particular to a data analysis method and device.
Background
In recent years, many publishers have begun to share their thoughts with others or show themselves through social media. Thus, to some extent, social behavior of the publisher may show the publisher's areas of interest. The interest of the publisher is the key for performing personalized information services such as media promotion, network marketing, e-commerce recommendation and personalized information retrieval through the social media KOL, reflects the personal characteristics and interest preference of the publisher, and is an important basis for performing social media propagation services.
In the prior art, the interest field is generally determined by a label set by a publisher. However, because there is a difference between the publisher self-located interest field and the publisher located interest field of the business side of the social media dissemination service, and the publisher interests are greatly influenced by living habits, time, place, weather, and other factors, in addition, different publishers have many different personal understandings and tendencies for the same set of self-chosen tags, and the accuracy of determining the interest field of the publisher based on the self-chosen tags is low.
Disclosure of Invention
In view of the above, the main objective of the present invention is to solve the problem of low accuracy in determining the interest field of a publisher based on the self-chosen tag of the publisher.
In one aspect, a data analysis method provided in an embodiment of the present invention includes: classifying each multimedia file of the publisher account respectively to obtain at least one information label and corresponding information weight of each multimedia file; acquiring a middle layer label and a middle layer weight according to at least one information label and a corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library; acquiring a display layer label and a display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library; and obtaining the interest classification of the publisher account according to the display layer label and the display layer weight.
In another aspect, an embodiment of the present invention provides a data analysis apparatus, including:
the analysis module is used for classifying each multimedia file of the publisher account respectively to obtain at least one information label and corresponding information weight of each multimedia file;
the middle layer conversion module is connected with the classification module and is used for acquiring a middle layer label and a middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library;
the display module is connected with the middle layer conversion module and is used for acquiring a display layer label and a display layer weight according to the middle layer label and the middle layer weight and a preset middle layer label-display layer label mapping library;
and the classification module is connected with the display module and is used for acquiring the interest classification of the publisher account according to the display layer label and the display layer weight.
In summary, the data analysis method and apparatus provided by the present invention classify multimedia files and convert layer by layer, so as to finally achieve the acquisition of interest classification of publisher accounts. The service classification of the display layer is realized through the middle layer, so that the technical scheme provided by the embodiment of the invention can be suitable for various systems; in addition, according to the technical scheme provided by the embodiment of the invention, all multimedia files of the account of the publisher are analyzed to obtain the interest classification, the interest classification can be updated in real time according to the favor of the publisher, and the problem that the accuracy rate of determining the interest field of the publisher only according to the label in the prior art is low is solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flow chart of a data analysis method provided by an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a data analysis apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it is to be understood that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the present invention provides a data analysis method, comprising:
step 101, classifying each multimedia file of the publisher account respectively to obtain at least one information tag and corresponding information weight of each multimedia file.
In this embodiment, the multimedia file of the publisher account in step 101 may include a multimedia file originally created by the publisher, or may include a multimedia file forwarded by the publisher, which is not limited herein. For more accurate data analysis, different comprehensive coefficients may be set for multimedia files from different sources, which is not described herein any more. Wherein the multimedia file includes: one or more of text, pictures, video and audio. For convenience of processing, the audio, the text type video and the characters can be processed in a unified mode, and are combined and analyzed after being converted into the characters.
When the multimedia file comprises a video, step 101 comprises: acquiring at least one information tag from a video through a preset video analysis model, and acquiring the occurrence time ratio t, the average confidence b and the occurrence times c of information corresponding to the at least one information tag; the information includes: people, things, and actions; acquiring information weight corresponding to at least one information label according to the occurrence time ratio t, the average confidence b and the occurrence times c of the information corresponding to the at least one information label; c is a positive integer of 0<b<1，0<t<1. If the video further includes text, step 101 further includes: intercepting a video into a first picture containing characters; extracting a first text from the first picture; and performing text analysis on the first text to obtain at least one information label and corresponding information weight. The video analysis model can be obtained by training in advance by using a machine learning method. Obtaining information weight corresponding to at least one information label through video
When the multimedia file includes audio, step 101 includes: converting the audio into a second text; and performing text analysis on the second text to obtain at least one information label and corresponding information weight.
When the multimedia file includes the third text, step 101 includes: and performing text analysis on the third text to obtain at least one information label and corresponding information weight.
When the multimedia file includes a second picture, step 101 includes: carrying out target detection on the second picture to obtain a corresponding target object; and acquiring at least one information label corresponding to the target object and the corresponding information weight. The second picture may be subjected to target detection through a deep neural network model, such as fast-RCNN, YOLO, SSD, and the like, which is not limited herein.
The first/second/third text may be analyzed by using a natural language processing model such as BERT, fastnext, and the like, which is not limited herein.
And 102, acquiring a middle layer label and a middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library.
In this embodiment, the manner of obtaining the middle layer label and the middle layer weight by 102 may be that, after each information label is mapped to the middle layer label according to a preset information label-middle layer label mapping library, the same weight of the middle layer label is combined to obtain a final middle layer label and a final middle layer weight; or, after combining the information weights of the same information tags, mapping each information tag to an intermediate layer tag according to a preset information tag-intermediate layer tag mapping library to obtain a final intermediate layer tag and an intermediate layer weight.
In this embodiment, the value of the intermediate layer weight may be the same as the value of the corresponding information weight, i.e., the value of the intermediate layer conversion coefficient is 1; in order to respectively and uniformly point a plurality of information labels to one middle layer label, corresponding middle layer conversion coefficients may be set for different contents, and the specific process of obtaining the middle layer label and the middle layer weight through step 102 may include: respectively acquiring the content type of each multimedia file; respectively acquiring a middle layer conversion coefficient corresponding to each content type; acquiring a middle layer conversion value corresponding to at least one information label of each multimedia file according to the middle layer conversion coefficient corresponding to each content type and the information weight corresponding to at least one information label of each multimedia file; and acquiring the middle layer label and the middle layer weight according to a preset information label-middle layer label mapping library and a middle layer conversion value corresponding to at least one information label of each multimedia file.
The content type of the multimedia file can comprise video, characters, audio, pictures and the like; for different content types, the same/different intermediate layer conversion coefficients can be preset; for any information tag, the intermediate conversion value corresponding to the information tag may be equal to the product of the intermediate conversion coefficient corresponding to the information tag and the information weight; specifically, there may be a plurality of the middle layer transform coefficients, for example, two, the middle layer transform value corresponding to the information tag is the first middle layer transform coefficient and the information weight is the second middle layer transform coefficient. Wherein the first intermediate layer conversion coefficient is more than or equal to 0 and less than or equal to 1; the second intermediate layer conversion factor is a rational number. According to the middle layer conversion value corresponding to at least one information label of each multimedia file, the middle layer label and the middle layer weight are obtained in a mode that the middle layer label corresponding to the information label of each multimedia file is respectively obtained according to a preset information label-middle layer label mapping library; and taking the sum of the middle layer conversion values of the same middle layer labels in all the multimedia files as the weight of the information label to obtain the middle layer label and the middle layer weight.
Specifically, step 102 may also obtain the middle layer tag corresponding to the information tag first, and then perform weight conversion, which is similar to the above process and is not described herein again.
And 103, acquiring a display layer label and a display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library.
In this embodiment 103, the manner of obtaining the display layer labels and the display layer weights may be that, after mapping each intermediate layer label to a display layer label according to a preset intermediate layer label-display layer label mapping library, the weights of the same display layer labels are combined to obtain final display layer labels and display layer weights; or, after combining the same information weights of the middle layer labels, mapping each middle layer label to a display layer label according to a preset middle layer label-display layer label mapping library to obtain a final display layer label and a final display layer weight.
In this embodiment, the value of the middle layer weight may be the same as the value of the corresponding display layer weight, i.e. the value of the display layer conversion coefficient is 1; in order to respectively and uniformly point a plurality of middle layer labels to one display layer label, the specific process of obtaining the display layer label and the display layer weight through step 103 may include: acquiring a display layer label corresponding to the middle layer label according to a preset middle layer label-display layer label mapping library; acquiring a display layer conversion coefficient corresponding to the middle layer label; and acquiring the display layer weight corresponding to the display layer label according to the display layer conversion coefficient corresponding to the middle layer label and the middle layer weight.
Wherein, for different middle layer labels, the same/different display layer conversion coefficients can be preset; for any intermediate layer label, the display layer weight corresponding to the display layer label may be equal to the product of the display layer conversion coefficient corresponding to the intermediate layer label and the intermediate layer weight; in particular, the display layer transition coefficient may be multiple, for example, two, display layer weights corresponding to the display layer label are the first display layer transition coefficient and the middle layer weightSecond display layer conversion factor。
Specifically, step 103 may also perform weight conversion, and then obtain a display layer label corresponding to the middle layer label, which is similar to the corresponding process in step 102 and is not described in detail here.
And 104, acquiring interest classification of the publisher account according to the display layer label and the display layer weight.
In this embodiment, the display layer weights of the same display layer label may be summed as the final weight of the display layer label in step 104. In particular, the display layer tags with the final weight values larger than the preset threshold value may be used as interest categories of the publisher account, which is not described in detail herein.
When the summary is performed in step 104, different time coefficients can be set according to the publication time of the multimedia file, so that the finally obtained interest classification can be updated in real time according to the preference of the publisher; meanwhile, other coefficients can be set so that interest classification is more practical, and are not described in detail herein.
Specifically, for example, a publisher publishes a piece of singing video on social media, namely whether you would like tomorrow, yesterday writing diary, tomorrow you still have activity, what you are most crying ever (the publisher plays guitar in the video and has banana placed beside it), and the video is matched with the text "guitar music heard on the internet".
Classifying videos through step 101 to obtain an information label 'guitar, banana and singing' of a video file, wherein the occurrence time ratio of the guitar is 1, the average confidence coefficient is 0.9, and the occurrence frequency is 1; the banana appearance time ratio is 1, the average confidence coefficient is 0.6, and the appearance frequency is 1; the occurrence time of singing is 2/3, the average confidence coefficient is 0.8, and the occurrence frequency is 1. The weight of the guitar was found to be 1 × 0.9/ln2 ═ 1.30; the weight of banana is 1 × 0.6/ln2 ═ 0.87; the singing weight is 2/3 × 0.8/ln2 ═ 0.77.
Converting the audio into characters through step 101 to obtain 'whether you would like tomorrow, yesterday's diary, tomorrow's still active, ever most crying you'; combining the characters converted by the audio frequency with the characters, wherein the complete characters are 'guitar music heard on the internet, whether you would like tomorrow, a diary written by yesterday, whether you still have activity tomorrow, and a person who love crying once'; performing text analysis on the complete characters to obtain an information label 'musical instrument' with the weight of 0.4; the information label "singing", weight 0.5.
The labels of guitar, banana and singing, musical instrument and singing can be gathered into the labels of middle layer label of musical instrument, fruit and singing by step 102, taking the middle layer conversion coefficients of video content as 0.8 and 0.9, the middle layer conversion coefficients of text content as 0.7 and 0.5 as examples, and the middle layer weight of the instrument of the middle layer label is the sum of the middle conversion value of 0.8, 1.3, 0.9 ^ 1.01 of the guitar and the middle conversion value of 0.7, 0.4, 0.5^ 0.44 of the instrument of the information label as 1.45; the weight of the middle layer of the fruit in the middle layer label is 0.8 x 0.87 x 0.9-0.71 of the middle layer conversion value of the banana in the information label; the middle layer weight of the middle layer label singing is 1.12 of the sum of the middle layer conversion value 0.8 x 0.77 x 0.9 x 0.63 of the video information label singing and the middle layer conversion value 0.7 x 0.5 x 0.49 of the text information label singing.
Taking the example that the middle layer label 'musical instrument and singing' is mapped to the display layer label music, the display layer conversion coefficients are 1 and 2, the middle layer label fruit is mapped to the display layer label diet, the display layer conversion coefficients are 1.5 and 2.5, and the display layer weight of the display layer label music is 1^ 1.45^2+ 1^ 1.12^ 2^ 3.36; display slice weight of display slice label diet was 1.5 x 0.71 x 2.5 x 0.64.
Thus, this social information pertains to music and diet of the display layer, with music weight of 3.36 and diet weight of 0.64.
Assuming that the publisher publishes n multimedia files, the above process can be repeated to obtain the label and weight of the real layer of each multimedia file, and the labels and the weights are summarized on the display layer; when the display layer is summarized, the weight can be adjusted according to the publishing time of the multimedia file and the coefficient corresponding to the situations such as the original or forwarding and the like. And taking the label with the aggregated weight larger than a preset threshold value as the interest classification of the publisher.
In summary, the data analysis method and apparatus provided by the present invention classify multimedia files and convert layer by layer, so as to finally achieve the acquisition of interest classification of publisher accounts. The service classification of the display layer is realized through the middle layer, so that the technical scheme provided by the embodiment of the invention can be suitable for various systems; in addition, according to the technical scheme provided by the embodiment of the invention, all multimedia files of the account of the publisher are analyzed to obtain the interest classification, the interest classification can be updated in real time according to the favor of the publisher, and the problem that the accuracy rate of determining the interest field of the publisher only according to the label in the prior art is low is solved.
Example 2
As shown in fig. 2, the present invention provides a data analysis apparatus including:
the analysis module 201 is configured to classify each multimedia file of the publisher account, respectively, to obtain at least one information tag and a corresponding information weight of each multimedia file;
the middle layer conversion module 202 is connected with the classification module and is used for acquiring a middle layer label and a middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library;
the display module 203 is connected with the middle layer conversion module and is used for acquiring the display layer label and the display layer weight according to the middle layer label and the middle layer weight and a preset middle layer label-display layer label mapping library;
and the classification module 204 is connected with the display module and is used for obtaining interest classifications of the publisher accounts according to the display layer labels and the display layer weights.
In this embodiment, the process of obtaining the interest classification of the publisher account is implemented by the parsing module 201, the middle layer conversion module 202, the display module 203, and the classification module 204, which is similar to the process provided in embodiment 1 of the present invention and is not described in detail herein.
In summary, the data analysis method and apparatus provided by the present invention classify multimedia files and convert layer by layer, so as to finally achieve the acquisition of interest classification of publisher accounts. The service classification of the display layer is realized through the middle layer, so that the technical scheme provided by the embodiment of the invention can be suitable for various systems; in addition, according to the technical scheme provided by the embodiment of the invention, all multimedia files of the account of the publisher are analyzed to obtain the interest classification, the interest classification can be updated in real time according to the favor of the publisher, and the problem that the accuracy rate of determining the interest field of the publisher only according to the label in the prior art is low is solved.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (2)
1. A method of data analysis, comprising:
classifying each multimedia file of the publisher account respectively to obtain at least one information label and corresponding information weight of each multimedia file;
acquiring a middle layer label and a middle layer weight according to at least one information label and a corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library;
acquiring a display layer label and a display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library;
obtaining interest classification of the publisher account according to the display layer label and the display layer weight;
when the multimedia file includes a video, the classifying each multimedia file of the publisher account respectively includes: acquiring at least one information tag from the video through a preset video analysis model, and acquiring the occurrence time ratio t, the average confidence b and the occurrence times c of information corresponding to the at least one information tag; the information includes: people, things, and actions; acquiring information weight corresponding to at least one information label according to the occurrence time ratio t, the average confidence b and the occurrence times c of the information corresponding to the at least one information label; c is a positive integer of 0<b<1，0<t<1; the information weight =
；
When the video contains the characters, the method also comprises the following steps of classifying each multimedia file of the account of the publisher respectively: intercepting the video into a first picture containing characters; extracting a first text from the first picture; performing text analysis on the first text to obtain at least one information label and corresponding information weight;
when the multimedia file includes audio, the classifying each multimedia file of the publisher account respectively includes: converting the audio into a second text; performing text analysis on the second text to obtain at least one information label and corresponding information weight;
when the multimedia file includes the third text, the classifying each multimedia file of the publisher account respectively includes: performing text analysis on the third text to obtain at least one information label and corresponding information weight;
when the multimedia file includes the second picture, the classifying each multimedia file of the publisher account respectively includes: carrying out target detection on the second picture to obtain a corresponding target object; acquiring at least one information label corresponding to the target object and corresponding information weight;
the acquiring the middle layer label and the middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library comprises the following steps: respectively acquiring the content type of each multimedia file; respectively acquiring a middle layer conversion coefficient corresponding to each content type; acquiring a middle layer conversion value corresponding to at least one information label of each multimedia file according to the middle layer conversion coefficient corresponding to each content type and the information weight corresponding to at least one information label of each multimedia file; acquiring an intermediate layer label and an intermediate layer weight according to a preset information label-intermediate layer label mapping library and an intermediate layer conversion value corresponding to at least one information label of each multimedia file;
the acquiring of the display layer label and the display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library comprises: acquiring a display layer label corresponding to a middle layer label according to a preset middle layer label-display layer label mapping library; acquiring a display layer conversion coefficient corresponding to the middle layer label; and acquiring the display layer weight corresponding to the display layer label according to the display layer conversion coefficient and the intermediate layer weight corresponding to the intermediate layer label.
2. A data analysis apparatus, comprising:
the analysis module is used for classifying each multimedia file of the publisher account respectively to obtain at least one information label and corresponding information weight of each multimedia file;
the middle layer conversion module is connected with the classification module and is used for acquiring a middle layer label and a middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library;
the display module is connected with the middle layer conversion module and is used for acquiring a display layer label and a display layer weight according to the middle layer label and the middle layer weight and a preset middle layer label-display layer label mapping library;
the classification module is connected with the display module and used for acquiring interest classification of the publisher account according to the display layer label and the display layer weight;
when the multimedia file includes a video, the classifying each multimedia file of the publisher account respectively includes: acquiring at least one information tag from the video through a preset video analysis model, and acquiring the occurrence time ratio t, the average confidence b and the occurrence times c of information corresponding to the at least one information tag; the information includes: people, things, and actions; acquiring information weight corresponding to at least one information label according to the occurrence time ratio t, the average confidence b and the occurrence times c of the information corresponding to the at least one information label; c is a positive integer of 0<b<1，0<t<1; the information weight =
；
When the video contains the characters, the method also comprises the following steps of classifying each multimedia file of the account of the publisher respectively: intercepting the video into a first picture containing characters; extracting a first text from the first picture; performing text analysis on the first text to obtain at least one information label and corresponding information weight;
when the multimedia file includes audio, the classifying each multimedia file of the publisher account respectively includes: converting the audio into a second text; performing text analysis on the second text to obtain at least one information label and corresponding information weight;
when the multimedia file includes the third text, the classifying each multimedia file of the publisher account respectively includes: performing text analysis on the third text to obtain at least one information label and corresponding information weight;
when the multimedia file includes the second picture, the classifying each multimedia file of the publisher account respectively includes: carrying out target detection on the second picture to obtain a corresponding target object; acquiring at least one information label corresponding to the target object and corresponding information weight;
the acquiring the middle layer label and the middle layer weight according to at least one information label and corresponding information weight of each multimedia file and a preset information label-middle layer label mapping library comprises the following steps: respectively acquiring the content type of each multimedia file; respectively acquiring a middle layer conversion coefficient corresponding to each content type; acquiring a middle layer conversion value corresponding to at least one information label of each multimedia file according to the middle layer conversion coefficient corresponding to each content type and the information weight corresponding to at least one information label of each multimedia file; acquiring an intermediate layer label and an intermediate layer weight according to a preset information label-intermediate layer label mapping library and an intermediate layer conversion value corresponding to at least one information label of each multimedia file;
the acquiring of the display layer label and the display layer weight according to the intermediate layer label and the intermediate layer weight and a preset intermediate layer label-display layer label mapping library comprises: acquiring a display layer label corresponding to a middle layer label according to a preset middle layer label-display layer label mapping library; acquiring a display layer conversion coefficient corresponding to the middle layer label; and acquiring the display layer weight corresponding to the display layer label according to the display layer conversion coefficient and the intermediate layer weight corresponding to the intermediate layer label.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910383032.7A CN110096482B (en) | 2019-05-09 | 2019-05-09 | Data analysis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910383032.7A CN110096482B (en) | 2019-05-09 | 2019-05-09 | Data analysis method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110096482A CN110096482A (en) | 2019-08-06 |
| CN110096482B true CN110096482B (en) | 2020-10-30 |
Family
ID=67447523
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910383032.7A Active CN110096482B (en) | 2019-05-09 | 2019-05-09 | Data analysis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110096482B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140289202A1 (en) * | 2013-03-21 | 2014-09-25 | Nextbit Systems Inc. | Utilizing user devices for backing up and retrieving data in a distributed backup system |
| CN105095431A (en) * | 2015-07-22 | 2015-11-25 | 百度在线网络技术（北京）有限公司 | Method and device for pushing videos based on behavior information of user |
| CN106055617A (en) * | 2016-05-26 | 2016-10-26 | 乐视控股（北京）有限公司 | Data pushing method and device |
-
2019
- 2019-05-09 CN CN201910383032.7A patent/CN110096482B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140289202A1 (en) * | 2013-03-21 | 2014-09-25 | Nextbit Systems Inc. | Utilizing user devices for backing up and retrieving data in a distributed backup system |
| CN105095431A (en) * | 2015-07-22 | 2015-11-25 | 百度在线网络技术（北京）有限公司 | Method and device for pushing videos based on behavior information of user |
| CN106055617A (en) * | 2016-05-26 | 2016-10-26 | 乐视控股（北京）有限公司 | Data pushing method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110096482A (en) | 2019-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20240281462A1 (en) | Content summarization leveraging systems and processes for key moment identification and extraction | |
| CN105677844B (en) | A kind of orientation of moving advertising big data pushes and user is across screen recognition methodss | |
| WO2021190174A1 (en) | Information determining method and apparatus, computer device, and storage medium | |
| Cole-Lewis et al. | Assessing electronic cigarette-related tweets for sentiment and content using supervised machine learning | |
| US20190102374A1 (en) | Predicting future trending topics | |
| JP6261547B2 (en) | Determination device, determination method, and determination program | |
| US8311997B1 (en) | Generating targeted paid search campaigns | |
| CN107145485B (en) | Method and apparatus for compressing topic models | |
| CN108021660B (en) | Topic self-adaptive microblog emotion analysis method based on transfer learning | |
| US11615485B2 (en) | System and method for predicting engagement on social media | |
| Franke | Applications of meta-analysis for marketing and public policy: A review | |
| Xu et al. | How do you say it matters? A multimodal analytics framework for product return prediction in live streaming e-commerce | |
| JP2017201535A (en) | Determination device, learning device, determination method, and determination program | |
| Al Sari et al. | Sentiment analysis for cruises in Saudi Arabia on social media platforms using machine learning algorithms | |
| Dorochowicz et al. | Employing subjective tests and deep learning for discovering the relationship between personality types and preferred music genres | |
| US20190080354A1 (en) | Location prediction based on tag data | |
| CN110096482B (en) | Data analysis method and device | |
| Chen | Multimodal cooperative learning for micro-video advertising click prediction | |
| Hamacher et al. | The Application of Artificial Intelligence to Automate Sensory Assessments Combining Pretrained Transformers with Word Embedding Based on the Online Sensory Marketing Index | |
| Park et al. | Automatic Movie Tag Generation System for Improving the Recommendation System | |
| Kaur et al. | A Review on Sentimental Analysis on Facebook Comments by using Data Mining Technique | |
| CN114996571A (en) | Information pushing method and device, storage medium and electronic equipment | |
| Radhika et al. | Personalized language-independent music recommendation system | |
| Pandit et al. | How good is your model ‘really’? on ‘wildness’ of the in-the-wild speech-based affect recognisers | |
| GB2608112A (en) | System and method for providing media content |
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 |