CN112199599A - Media portrait generation method and system - Google Patents
Media portrait generation method and system Download PDFInfo
- Publication number
- CN112199599A CN112199599A CN202011171680.5A CN202011171680A CN112199599A CN 112199599 A CN112199599 A CN 112199599A CN 202011171680 A CN202011171680 A CN 202011171680A CN 112199599 A CN112199599 A CN 112199599A
- Authority
- CN
- China
- Prior art keywords
- data
- target media
- media
- push
- target
- 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
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000004458 analytical method Methods 0.000 claims abstract description 35
- 238000012544 monitoring process Methods 0.000 claims abstract description 6
- 230000000007 visual effect Effects 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 21
- 238000007635 classification algorithm Methods 0.000 claims description 13
- 238000012795 verification Methods 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 11
- 238000013145 classification model Methods 0.000 claims description 10
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 abstract description 6
- 238000013473 artificial intelligence Methods 0.000 abstract description 4
- 238000003062 neural network model Methods 0.000 description 9
- 238000003860 storage Methods 0.000 description 7
- 238000004590 computer program Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000006399 behavior Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 206010039203 Road traffic accident Diseases 0.000 description 2
- 230000019771 cognition Effects 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000000843 powder Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000009877 rendering Methods 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 206010044565 Tremor Diseases 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
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
- G06F16/35—Clustering; Classification
-
- 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
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a media portrait generation method and a system, wherein the method comprises the following steps: monitoring a target media account to acquire push data of a target media; selectively acquiring public portrait data of the target media according to the push data; optionally establishing at least one target media preference portrait analysis model, and analyzing and acquiring target media preference portrait data according to the pushed data; creating a visual media representation based on the public representation data and the preference representation data; the invention adopts artificial intelligence technology, establishes a plurality of analysis models according to the media attributes, generates the media portraits according to the analysis models, and can provide diversified labels for each media through big data and artificial intelligence, thereby automatically embodying the portraits content of the media in the subdivision attributes, improving the analysis depth of the portraits content and improving the accuracy and diversity of the media portraits.
Description
Technical Field
The invention relates to the field of internet big data, in particular to a media portrait generation method and a media portrait generation system.
Background
Each media in the internet field has a respective main operation range, and related reports are produced according to the main operation range, so that certain differences exist among different media, service contents and service modes, and no special data and visual analysis method and tool exist for the media with different attributes at present. The analysis of the internet media attributes is beneficial to improving the knowledge of the media, and services such as advertisement putting, television rebroadcasting and the like can be more pertinently developed on the basis of big data analysis.
Disclosure of Invention
One of the objectives of the present invention is to provide a method and a system for generating a media portrait, which establish a basic attribute for a media type, and improve the media cognition by establishing portrait content through big data analysis and processing on the basic attribute.
The invention also aims to provide a media portrait generation method and a system, the invention adopts the artificial intelligence technology, establishes a plurality of analysis models according to the media attributes, generates the media portrait according to the analysis models, and can provide diversified labels for each media through big data and artificial intelligence, thereby automatically embodying the portrait content of the media in the subdivision attributes, improving the analysis depth of the portrait content and improving the accuracy and diversity of the media portrait.
It is another object of the present invention to provide a method and system for generating a media representation, which can improve the visualization effect of the media representation by displaying different attributes in a visualized manner according to a certain rule.
The invention also aims to provide a media portrait generation method and a system thereof, which can acquire the content sent by the media through multiple channels, deeply acquire the content of media service branches and fields and comprehensively improve the cognition on the media.
To achieve at least one of the above objects, the present invention further provides a media representation generation method, comprising:
monitoring a target media account to acquire push data of a target media;
selectively acquiring public portrait data of the target media according to the push data;
optionally establishing at least one target media preference portrait analysis model, and analyzing and acquiring target media preference portrait data according to the pushed data;
a visual media representation is created based on the public representation data and the preference representation data.
According to a preferred embodiment of the present invention, the common representation data includes base data including: the administrative level of the target media mechanism, the region to which the target media belongs, the type of the target media, the positioning of the target media and the language data adopted by the target media.
According to one preferred embodiment of the present invention, the public representation data includes productivity data, transmission force data, and influence data;
the productivity data acquisition method comprises the following steps: acquiring each piece of push data of a target medium, recording the release time of each piece of push data, calculating the push frequency of the target medium, and setting different productivity labels for the target media with different push frequencies;
the method for acquiring the propagation force data comprises the following steps: calculating the propagation elements of each piece of pushed data of the target media, wherein the propagation elements comprise total reading amount, total praise amount, total forwarding amount and total evaluation amount data, setting a weight value for each propagation element respectively, calculating the sum of products of each propagation element and the corresponding weight value to obtain a propagation index, and setting different propagation labels according to the magnitude of the propagation index.
The influence data acquisition method comprises the following steps: calculating influence elements of each pushed data of the target media, wherein the influence elements comprise the number of original fans, the number of newly added fans and the number of newly added fans in unit time, setting a weight value for each influence element, calculating the sum of the products of each influence element and the corresponding weight value to obtain an influence index, and setting different influence labels according to the size of the influence index.
According to a preferred embodiment of the present invention, push data of different channels of the target media account are monitored, and productivity data of the same push data of different channels are calculated respectively for calculating the target media push behavior preference data.
According to a preferred embodiment of the present invention, the method for producing the push behavior preference data comprises the following steps:
the method comprises the steps of obtaining pushing time and pushing quantity of a target media account in pushing data of different channels, calculating the pushing quantity of the target media in different channels in unit time, and calculating the ratio of the pushing quantity of each channel to the total pushing quantity, wherein the ratio is used for obtaining the pushing behavior preference data.
According to one preferred embodiment of the invention, push data content is acquired, each piece of push data is classified by adopting a text classification algorithm, and a classification label is set for each classified piece of push data to generate preference portrait data of a target medium;
the classification method comprises the following steps:
establishing a labeled push data training set, a labeled verification set and a labeled test set, and training the training set by adopting a text classification algorithm;
adjusting the hyper-parameter adjustment of the text classification algorithm by adopting a verification set;
evaluating the generalization ability of the text classification algorithm by adopting a test set, and forming a classification model;
and inputting each piece of pushed data into a preset classification model to obtain each pushed data classification label.
According to one of the preferred embodiments of the present invention, the category labels include social, life, sports, entertainment, science and technology, military, financial and time-based, and the category and number of the category labels in each target media push data are calculated.
According to a preferred embodiment of the present invention, a plurality of entity content analysis models are used to obtain entity content preferences of target media push data, wherein the entity content analysis models include a region entity analysis model, a character entity analysis model and a mechanism entity analysis model, the target media push data are respectively input into the region entity analysis model, the character entity analysis model and the mechanism entity analysis model, and a region entity tag, a character entity tag and a mechanism entity tag used to obtain the target media push data are used to form preference portrait data of the target media.
According to a preferred embodiment of the present invention, the target media push data is obtained, and the push data is input into the emergency analysis model, and the tag of the emergency recorded in each push data is obtained for forming the target media preference profile data.
According to one preferred embodiment of the present invention, the target media push data is obtained, each push data is input into a topic model, and a topic tag of each push data is obtained for forming the target media preference portrait data.
According to one preferred embodiment of the invention, the push data of another target medium is calculated, the topic tag of the target medium in the field is obtained, the similarity of topic between the two target media is calculated, a similarity threshold value is set, if the similarity of the two target media is greater than the similarity threshold value, the two target media are stored, and a media library of similar topic topics is generated.
According to a preferred embodiment of the present invention, an image-to-text model is used to convert the text of the image or video in the target media push data into text.
According to a preferred embodiment of the present invention, a voice-to-text model is used to convert the voice information in the target media push data into text words.
To achieve at least one of the above objects, the present invention further provides a media representation generation system, which employs the above media representation generation method.
Drawings
FIG. 1 is a flow chart illustrating a method for generating a media representation according to a preferred embodiment of the present invention;
FIG. 2 is a schematic diagram showing a media rendering effect of a media rendering method according to the present invention.
Detailed Description
The following description is presented to disclose the invention so as to enable any person skilled in the art to practice the invention. The preferred embodiments in the following description are given by way of example only, and other obvious variations will occur to those skilled in the art. The underlying principles of the invention, as defined in the following description, may be applied to other embodiments, variations, modifications, equivalents, and other technical solutions without departing from the spirit and scope of the invention.
It will be understood by those skilled in the art that in the present disclosure, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like are used in an orientation or positional relationship indicated in the drawings for ease of description and simplicity of description, and do not indicate or imply that the referenced devices or components must be in a particular orientation, constructed and operated in a particular orientation, and thus the above terms are not to be construed as limiting the present invention.
It is understood that the terms "a" and "an" should be interpreted as meaning that a number of one element or element is one in one embodiment, while a number of other elements is one in another embodiment, and the terms "a" and "an" should not be interpreted as limiting the number.
Please refer to fig. 1, which shows a schematic flow chart of a media portrait generating method of the present invention, the method includes:
monitoring a target media account to acquire target media push data, wherein the push data comprises but is not limited to pushed manuscript content, video content, audio content, push time and the like, and the push content comprises original content and forwarding content; the monitoring of the target media account further includes monitoring target media accounts of multiple channels, such as APP or website channels like WeChat, microblog, tremble, today's first line, and the like, and acquiring push data of the target media accounts of the multiple channels.
Further, after acquiring push data of a target account, processing and analyzing the push data to acquire public portrait data and preference portrait data of the target media account, it should be noted that the public portrait data is attribute data of the target media, where the public portrait data includes: and basic attribute data, wherein the basic attribute data comprises administrative levels of the target media mechanism, regions to which the target media belong, types of the target media, positioning of the target media and language data adopted by the target media. Wherein the target media is tagged according to the base attribute; for example: if the target media is 'sports channel with central vision', the target media administrative level label of the 'sports channel with central vision' is manually or automatically set to 'national level', the region label of the target media is set to 'Beijing', the target media type label is set to 'sports', the target media positioning is set to 'Chinese', and the language label adopted by the target media is set to 'Chinese'. It should be noted that in a preferred embodiment of the present invention, the tag can be automatically set according to the field of the media name, for example, the "central view" mentioned above can automatically set the corresponding tag by identifying the "central view" field in the media name: "national level", "Beijing", and "China".
The public representation data further comprises productivity data, transmission force data, and influence data, wherein the productivity data acquisition method comprises the steps of: acquiring each piece of push data of a plurality of target media, recording the release time of each piece of push data, calculating the push frequency of the target media, and setting different productivity labels for the target media with different push frequencies; for example: calculating a daily pushed amount S1, a near 7-day pushed amount S2, and a near 30-day pushed amount S3 in the target media, and setting a daily pushed amount weight W1, a 7-day pushed amount weight W2, and a near 30-day pushed amount weight W3, wherein W1+ W2+ W3 is 1, a productivity index P1 is S1 + W1+ S2 + W2+ S3 is W3, sorting the plurality of target media according to the value of the productivity index P1 from large to small, setting a "high-yield" tag in the target media 30% before the productivity index P1, a "low-yield" tag in the target media setting "with the productivity index P1 in the middle 60%, setting a" low-yield "tag in the target media setting at the end 10%, and it should be noted that the tags are relative to be filtered according to the frequency of all monitored target media, in another embodiment of the present invention, the target media belonging to the target media type calculation method, and calculating the productivity indexes of the target media in the types, and setting labels of high yield, medium yield and low yield according to the productivity indexes respectively for forming the public portrait data.
The method for generating the propagation force data comprises the following steps: calculating propagation elements of each piece of pushed data of the target media, wherein the propagation elements comprise but are not limited to total reading amount, total praise amount, total forwarding amount and total evaluation amount data, setting weights for each propagation element respectively, calculating the sum of products of each propagation element and the corresponding weight to obtain a propagation index, and setting different propagation labels according to the magnitude of the propagation index; in detail: and calculating the total reading quantity, the total praise quantity, the total forwarding quantity, the total appraisal quantity of the total pushed data of the target media, and the average reading quantity, the average praise quantity, the average forwarding quantity and the average appraisal quantity of the pushed data, respectively setting weights according to the 8 propagation elements, and calculating the propagation index, wherein the propagation index is the sum of the products of each propagation element and the corresponding weight. And further sequencing the images from high to low according to the propagation index, setting the target media of which the percentage is 30% in front of the propagation index as 'numerous upheaval', setting the target media of which the percentage is 60% in the middle of the propagation index as 'powder absorption', and setting the target media of which the percentage is 10% at the tail of the propagation index as 'powder shortage', wherein the target media are used for forming public portrait data.
And extracting the basic attribute label, the productivity label, the propagation force label and the influence label, and visually displaying on a display to form complete public portrait data.
It should be noted that the present invention further adopts a deep learning technique to generate a plurality of target media portrait preference analysis models, where the target media portrait preference analysis models are formed after training, verification and testing based on an existing neural network model, for example, the target media portrait preference analysis models include a classification model for classifying and screening contents of each push number and generating the portrait preference data of a target media, where the classification method includes the following steps:
establishing a labeled push data training set, a labeled verification set and a labeled test set, and training the training set by adopting a Text classification algorithm (Text CNN);
adjusting the hyper-parameter adjustment of the text classification algorithm by adopting a verification set;
evaluating the generalization ability of the text classification algorithm by adopting a test set, and forming a classification model;
and inputting each piece of pushed data into a preset classification model to obtain each pushed data classification label.
For example: the method comprises the steps of classifying the pushed contents reported by media into 8 categories of society, life, sports, entertainment, science and technology, military affairs, financial affairs and time administration, respectively setting the 8 categories of each training set, each verification set and each training set according to the pushed contents, training the classification model through the training sets, adjusting the hyper-parameters of the text classification algorithm through the verification sets, further evaluating the generalization capability of the text classification algorithm by adopting the test sets, specifically obtaining the convergence state of the model according to the loss function, and judging whether the classification model meets the requirements or not. And inputting text information pushed by the target media in the trained classification model, and acquiring a classification label of the target media for generating preference portrait data of the target media.
It is worth mentioning that the target media preference profile analysis model comprises a solid content analysis model, the solid content analysis model comprising: the method comprises the following steps of:
labeling the character entity information according to the push data of each target media;
forming a training set, a verification set and a test set by pushing data of target media marked with character entity information;
inputting the text content of the training set into a neural network model for training;
inputting the text content of the verification set into the neural network model for adjusting the hyper-parameters;
inputting text content of a test set into the neural network model for evaluating generalization capability of the neural network model;
adjusting the training set or the hyper-parameters to form a character entity model meeting the generalization requirement;
and inputting the pushed data of each target medium into the character entity model to obtain the character entity model label of each target medium, wherein the character entity label is used for constructing preference portrait data of the pushed data of the target medium.
Similarly, the establishment methods of the mechanism entity model and the region entity model both adopt the neural network model for marking and training. The mechanism entity model construction method needs to label mechanism entity information related to each pushed data content, wherein the mechanism entity information includes but is not limited to a mechanism name, a mechanism level, a mechanism geographic position and a mechanism attribute. The mechanism entity model is used for outputting mechanism entity labels of each piece of push data, calculating the types and the number of the mechanism entity labels of the target media, and acquiring mechanism preference data of the target media, wherein a region entity model construction method needs to label region entity information related to each piece of push data, the region entity information comprises but is not limited to news places and media mechanism addresses, and counts the number of the news places, the region entity model outputs the region entity labels of each piece of push data, and the region preference data of the target media is calculated according to the region entity labels. And the preference portrait data of the target media can be constructed according to the mechanism preference data and the region preference data.
Further, the emergency data related to the text content of each target media push data is calculated by using an emergency model, where the emergency data includes but is not limited to natural disasters, traffic accidents and financial events, the emergency model is based on a neural network model, and each target media push data is labeled, and the labeled content may include but is not limited to: financial crisis, volcanic eruption, tsunami, traffic accidents and the like, and is used for obtaining the type and the number of the emergency labels of each target media push data, and further using the labels to construct the preference portrait data of the target media.
In one preferred embodiment of the present invention, the text information in the target media push data is obtained, and the keyword in the text information is extracted by using the keyword model, wherein the keyword model can be obtained according to the pre-training, which is not described in detail herein. And acquiring keyword distribution of the target media push data by adopting a keyword model, and constructing preference portrait data of the target media.
Analyzing and acquiring a topic label in each target media push data by adopting a topic model, wherein the topic model is constructed based on the existing neural network model, and the construction method comprises the following steps: and inputting the training set, the verification set and the test set of the marked topic into the neural network model, and constructing the topic model meeting the generalization requirement by adjusting input data and neural network parameters. Inputting each media push data into the constructed topic model, obtaining the corresponding topic tag, further counting the type and the number of the topic tags of the target media, and using the statistics to construct the preference portrait data of the target media.
Further, counting the subject topic labels of all target media, further calculating the number of the same topic labels in any two target media, calculating the ratio of the same topic labels to all the labels of a single target media, and setting the phaseA similarity threshold, if the ratio is greater than the similarity threshold, storing the two target media as a similar topic media library, for example, if the similarity threshold is 75%, the total push amount of the first target media is P2The total push amount of the second target media is P3Calculate P2And P3Number Z of topics of the same subject1If Z is1/P2Or Z1/P3And if the target media is larger than or equal to 75 percent, storing the first target media and the second target media into a similar media library, and displaying the similar media library in the target media portrait in a visual mode.
It should be noted that the target media push data includes, but is not limited to, text data, image data, video data, and voice data, and in a preferred embodiment of the present invention, an image-to-text technology may be used to convert text images in the image data or the video data into text information, further analyze the text information, and set different tags. Or, converting voice data in the target media push data into recognizable text information by adopting a voice-to-text technology, further analyzing according to the converted text information, and setting different labels.
In particular, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication section, and/or installed from a removable medium. The computer program, when executed by a Central Processing Unit (CPU), performs the above-described functions defined in the method of the present application. It should be noted that the computer readable medium mentioned above in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. 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 of the computer readable storage medium may include, but are not limited to: 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 present application, 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. In this application, however, 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, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
It will be understood by those skilled in the art that the embodiments of the present invention described above and illustrated in the drawings are given by way of example only and not by way of limitation, the objects of the invention having been fully and effectively achieved, the functional and structural principles of the present invention having been shown and described in the embodiments, and that various changes or modifications may be made in the embodiments of the present invention without departing from such principles.
Claims (14)
1. A method for generating a media representation, the method comprising:
monitoring a target media account to acquire push data of a target media;
selectively acquiring public portrait data of the target media according to the push data;
optionally establishing at least one target media preference portrait analysis model, and analyzing and acquiring target media preference portrait data according to the pushed data;
a visual media representation is created based on the public representation data and the preference representation data.
2. A media representation generation method in accordance with claim 1, wherein said public representation data includes base data, said base data including: the administrative level of the target media mechanism, the region to which the target media belongs, the type of the target media, the positioning of the target media and the language data adopted by the target media.
3. A media representation generation method in accordance with claim 1, wherein said public representation data includes productivity data, transmission force data, and influence data;
the productivity data acquisition method comprises the following steps: acquiring each piece of push data of a target medium, recording the release time of each piece of push data, calculating the push frequency of the target medium, and setting different productivity labels for the target media with different push frequencies;
the method for acquiring the propagation force data comprises the following steps: calculating the propagation elements of each piece of pushed data of the target media, wherein the propagation elements comprise total reading amount, total praise amount, total forwarding amount and total evaluation amount data, setting weights for each propagation element respectively, calculating the sum of products of each propagation element and the corresponding weight to obtain a propagation index, and setting different propagation labels according to the magnitude of the propagation index;
the influence data acquisition method comprises the following steps: calculating influence elements of each pushed data of the target media, wherein the influence elements comprise the number of original fans, the number of newly added fans and the number of newly added fans in unit time, setting a weight value for each influence element, calculating the sum of the products of each influence element and the corresponding weight value to obtain an influence index, and setting different influence labels according to the size of the influence index.
4. The method of claim 1, wherein the push data of the target media account in different channels are monitored, and the productivity data of the same push data in different channels is calculated for calculating the push behavior preference data of the target media.
5. A method for generating a media representation as claimed in claim 4, wherein said push behavior preference data is generated by a method comprising the steps of:
the method comprises the steps of obtaining pushing time and pushing quantity of a target media account in pushing data of different channels, calculating the pushing quantity of the target media in different channels in unit time, and calculating the ratio of the pushing quantity of each channel to the total pushing quantity, wherein the ratio is used for obtaining the pushing behavior preference data.
6. The method of claim 4, wherein the content of the push data is obtained, each push data is classified by a text classification algorithm, and each classified push data is provided with a classification tag for generating the preferred image data of the target media;
the classification method comprises the following steps:
establishing a labeled push data training set, a labeled verification set and a labeled test set, and training the training set by adopting a text classification algorithm;
adjusting the hyper-parameter adjustment of the text classification algorithm by adopting a verification set;
evaluating the generalization ability of the text classification algorithm by adopting a test set, and forming a classification model;
and inputting each piece of pushed data into a preset classification model to obtain each pushed data classification label.
7. The method of claim 6, wherein said category labels comprise social, life, sports, entertainment, science, military, financial, and time-based, and said calculating the category and number of said category labels in each target media push data.
8. The method of claim 4, wherein a plurality of entity content analysis models are used to obtain entity content preferences of the target media push data, wherein the entity content analysis models include a region entity analysis model, a character entity analysis model and a mechanism entity analysis model, the target media push data are respectively input into the region entity analysis model, the character entity analysis model and the mechanism entity analysis model, and a region entity tag, a character entity tag and a mechanism entity tag used to obtain the target media push data are used to form preference profile data of the target media.
9. The method of claim 4, wherein the target media push data is obtained and input into the emergency analysis model, and the tag of the emergency recorded in each push data is obtained for forming the target media preference image data.
10. The method of claim 1, wherein the target media push data is obtained, each push data is input into a topic model, and a topic tag is obtained for each push data to form the target media preference image data.
11. The method as claimed in claim 10, wherein the pushed data of another target media is calculated, the topic tag of the target media in the domain is obtained, the similarity of the topic between the two target media is calculated, a similarity threshold is set, and if the similarity of the two target media is greater than the similarity threshold, the two target media are saved to generate a media library of similar topic topics.
12. The method of claim 1, wherein an image-to-text model is used to convert text of the image or video in the target media push data into text.
13. The method of claim 1, wherein a speech-to-text model is used to convert the speech information in the target media push data into text.
14. A media representation generation system, characterized in that said system employs a media representation generation method as claimed in any of the claims 1-13.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011171680.5A CN112199599A (en) | 2020-10-28 | 2020-10-28 | Media portrait generation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011171680.5A CN112199599A (en) | 2020-10-28 | 2020-10-28 | Media portrait generation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112199599A true CN112199599A (en) | 2021-01-08 |
Family
ID=74011824
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011171680.5A Pending CN112199599A (en) | 2020-10-28 | 2020-10-28 | Media portrait generation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112199599A (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140189000A1 (en) * | 2013-01-02 | 2014-07-03 | Microsoft Corporation | Social media impact assessment |
| CN105787025A (en) * | 2016-02-24 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Network platform public account classifying method and device |
| CN106779073A (en) * | 2016-12-27 | 2017-05-31 | 西安石油大学 | Media information sorting technique and device based on deep neural network |
| CN109544396A (en) * | 2019-01-10 | 2019-03-29 | 腾讯科技（深圳）有限公司 | Account recommended method, device, server, terminal and storage medium |
| US20190147058A1 (en) * | 2017-11-16 | 2019-05-16 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for pushing multimedia content |
| CN110019774A (en) * | 2017-08-23 | 2019-07-16 | 腾讯科技（深圳）有限公司 | Label distribution method, device, storage medium and electronic device |
| US20200089812A1 (en) * | 2018-09-19 | 2020-03-19 | International Business Machines Corporation | Updating social media post based on subsequent related social media content |
| CN111177533A (en) * | 2019-12-05 | 2020-05-19 | 哈勃智远(北京)科技有限公司 | Method and equipment for generating converged media influence analysis report |
-
2020
- 2020-10-28 CN CN202011171680.5A patent/CN112199599A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140189000A1 (en) * | 2013-01-02 | 2014-07-03 | Microsoft Corporation | Social media impact assessment |
| CN105787025A (en) * | 2016-02-24 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Network platform public account classifying method and device |
| CN106779073A (en) * | 2016-12-27 | 2017-05-31 | 西安石油大学 | Media information sorting technique and device based on deep neural network |
| CN110019774A (en) * | 2017-08-23 | 2019-07-16 | 腾讯科技（深圳）有限公司 | Label distribution method, device, storage medium and electronic device |
| US20190147058A1 (en) * | 2017-11-16 | 2019-05-16 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for pushing multimedia content |
| US20200089812A1 (en) * | 2018-09-19 | 2020-03-19 | International Business Machines Corporation | Updating social media post based on subsequent related social media content |
| CN109544396A (en) * | 2019-01-10 | 2019-03-29 | 腾讯科技（深圳）有限公司 | Account recommended method, device, server, terminal and storage medium |
| CN111177533A (en) * | 2019-12-05 | 2020-05-19 | 哈勃智远(北京)科技有限公司 | Method and equipment for generating converged media influence analysis report |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107391760B (en) | User interest recognition methods, device and computer readable storage medium | |
| CN107193974B (en) | Regional information determination method and device based on artificial intelligence | |
| CN109919244B (en) | Method and apparatus for generating a scene recognition model | |
| CN108021660B (en) | Topic self-adaptive microblog emotion analysis method based on transfer learning | |
| CN107273295B (en) | Software problem report classification method based on text chaos | |
| CN112001184B (en) | User emotion difference region detection method and system for video bullet screen | |
| CN108363748B (en) | Topic portrait system and topic portrait method based on knowledge | |
| CN105809186A (en) | Emotion classification method and system | |
| CN112995690B (en) | Live content category identification method, device, electronic equipment and readable storage medium | |
| CN111460267B (en) | Object identification method, device and system | |
| CN113064995A (en) | Text multi-label classification method and system based on deep learning of images | |
| CN110475158B (en) | Video learning material providing method and device, electronic equipment and readable medium | |
| Cheng | Data-mining research in education | |
| TW201640383A (en) | Internet events automatic collection and analysis method and system thereof | |
| CN114022923A (en) | Intelligent collecting and editing system | |
| CN117113973A (en) | Information processing method and related device | |
| CN112199599A (en) | Media portrait generation method and system | |
| US11574125B2 (en) | Method for automatically determining target entities from unstructured conversation using natural language understanding | |
| CN112800230B (en) | Text processing method and device, computer readable storage medium and electronic equipment | |
| Kurniawan et al. | Exploring Tourist Feedback on Riau Attractions Through Indonesian Language YouTube Opinion Using Naïve Bayes Algorithm | |
| CN112506405B (en) | Artificial intelligent voice large screen command method based on Internet supervision field | |
| CN111143688B (en) | Evaluation method and system based on mobile news client | |
| CN114648688A (en) | Method, system and equipment for evaluating landscape level along high-speed rail and readable storage medium | |
| CN114240250A (en) | Intelligent management method and system for vocational evaluation | |
| CN113569153A (en) | Image-text content classification method and device, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210108 |
|
| RJ01 | Rejection of invention patent application after publication |