CN113204709A - Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium - Google Patents
Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium Download PDFInfo
- Publication number
- CN113204709A CN113204709A CN202110595730.0A CN202110595730A CN113204709A CN 113204709 A CN113204709 A CN 113204709A CN 202110595730 A CN202110595730 A CN 202110595730A CN 113204709 A CN113204709 A CN 113204709A
- Authority
- CN
- China
- Prior art keywords
- video
- image
- short
- short video
- coefficient
- 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.)
- Withdrawn
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
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/735—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/75—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
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/783—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
- G06F16/7844—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using original textual content or text extracted from visual content or transcript of audio data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/78—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/7867—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title and artist information, manually generated time, location and usage information, user ratings
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Library & Information Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a short video search matching recommendation method, a system and a computer storage medium based on multidimensional data depth comparison analysis, which divide imported short videos into video images, count the area ratio of each category of attribute image in each video image in the short videos, analyze type attribute images corresponding to each video image in the short videos, count the number of video images of each category of attribute image in the short videos, calculate the type attribute image weight proportion coefficient of the short videos, simultaneously acquire video voice text information corresponding to each video image in the short videos, obtain each effective word in each section of video text information in the short videos, compare and analyze the word frequency of each keyword in the short videos, calculate the keyword weight proportion coefficient of the short videos, obtain the video playing duration corresponding to each keyword in the short videos, and calculate the comprehensive matching degree coincidence coefficient of the short videos, and recommending other short videos with the matching degrees meeting the highest coefficient by contrast screening.
Description
Technical Field
The invention relates to the field of short video matching recommendation, in particular to a short video searching matching recommendation method and system based on multidimensional data depth comparison analysis and a computer storage medium.
Background
With the rapid development of internet technology, we are in the era of short video information overload. The short video platform is hard to find the content really interested by the user in the face of excessive information, so how to effectively recommend the interested short video to the user is a problem to be solved urgently. At present, the existing short video search matching recommendation method basically infers the interest of a user according to all historical behaviors of the user, so as to recommend a short video list which is most likely to be interested by the user, but deduces the interest of the user from the historical behaviors, so that the problem of information hysteresis and time delay exists, and the targeted real-time recommendation of the interest points of the user cannot be realized, so that the actual requirements of the user cannot be met, the accuracy of short video recommendation is reduced, the short video viewing experience and interest of the user are influenced, and the intelligent matching recommendation level of the short video is reduced.
Disclosure of Invention
The invention aims to provide a short video search matching recommendation method, a system and a computer storage medium based on multidimensional data depth comparison analysis, the invention divides an imported short video into video images, counts the area of each category of attribute image in each video image in the short video, calculates the area ratio of each category of attribute image in each video image in the short video, analyzes the type attribute image corresponding to each video image in the short video, counts the number of video images of each category of attribute image in the short video, calculates the type attribute image weight proportion coefficient of the short video, simultaneously obtains the video voice text information corresponding to each video image in the short video, obtains each effective word in each section of video text information in the short video, contrasts and analyzes the word frequency of each keyword in the short video, calculates the keyword weight proportion coefficient of the short video, and obtains the video playing time length corresponding to each keyword in the short video, and calculating the comprehensive matching degree coincidence coefficient of the short videos, and comparing and screening other short videos with the highest matching degree coincidence coefficient for recommendation, thereby solving the problems in the background technology.
The purpose of the invention can be realized by the following technical scheme:
in a first aspect, the invention provides a short video search matching recommendation method based on multidimensional data depth comparison analysis, comprising the following steps:
s1, short video division: the method comprises the steps of importing the short video watched by a user through a short video importing module, dividing the imported short video into video images according to a set video frame dividing mode, numbering the video images in the short video in sequence according to video time sequence, and numbering the video images in the short video as a1,a2,...,ai,...,an，aiExpressed as the ith video image in the short video;
s2, video image processing: processing each video image in the short video by a video image processing module, processing each video image in the short video by adopting an image processing technology to obtain each category attribute image in each video image in the short video, and forming each category attribute image set A in each video image in the short videoiP(aip1,aip2,...,aipj,...,aipm)，aipjRepresenting the image as the jth category attribute image in the ith video image in the short video;
s3, image area ratio analysis: measuring the area of each category attribute image in each video image in the short video through an image area measuring module, counting the area of each category attribute image in each video image in the short video, extracting the standard area of the image stored in a storage database in a fixed standard form, and calculating the area ratio of each category attribute image in each video image in the short video;
s4, category attribute image statistics: comparing the area ratios of various types of attribute images in various video images in the short video through a category attribute image analysis module, screening the type attribute image with the largest area ratio in various video images in the short video, and obtaining various video images in the short videoType attribute images corresponding to the images, and a video image number set X (X) forming each type attribute image in the short video1,x2,...,xj,...,xm)，xjThe number of video images represented as jth category attribute images in the short video;
s5, analyzing image weight proportion coefficient: extracting a weight compensation coefficient corresponding to the type attribute image stored in the storage database through an analysis server, and calculating a weight proportion coefficient of the type attribute image of the short video;
s6, acquiring video character information: respectively acquiring video voice corresponding to each video image in the short video through a video voice acquisition module, identifying video voice text information corresponding to each video image in the short video, and forming each video text information set B (B) in the short video1,b2,...,bi,...,bn)，biThe text information is represented as the ith video text information in the short video;
s7, counting the number of effective words: performing word segmentation processing on each segment of video character information in the short video through a character information analysis module to obtain each effective word in each segment of video character information in the short video, and counting the occurrence frequency of each effective word in the short video to form an occurrence frequency set Y (Y) of each effective word in the short video1,y2,...,yr,...,yv)，yrExpressed as the number of occurrences of the r-th valid word in the short video;
s8, effective word frequency analysis: respectively calculating the word frequency of each effective word in the short video through an effective word frequency analysis module, comparing the word frequency of each effective word in the short video with the standard word frequency of a set keyword, if the word frequency of a certain effective word in the short video is greater than or equal to the standard word frequency of the set keyword, indicating that the effective word in the short video is the keyword, counting the word frequency of each keyword in the short video, and forming a word frequency set f ' (f ') of each keyword in the short video '1,f′2,...,f′u,...,f′l),l≤v，f′uThe word frequency is expressed as the u-th keyword in the short video;
s9, character weight proportion coefficient analysis: extracting a weight compensation coefficient corresponding to a keyword stored in a storage database through an analysis server, and calculating a keyword weight proportion coefficient of the short video;
s10, acquiring video playing time length: the video playing time length corresponding to each keyword in the short video is obtained through the video time length obtaining module to form a video playing time length set T (T) corresponding to each keyword in the short video1,t2,...,tu,...,tl)，tuRepresenting the video playing time corresponding to the u-th keyword in the short video;
s11, short video matching conformity coefficient analysis: the method comprises the steps of extracting an influence proportion coefficient of the total playing time length and the video playing time length ratio of the short videos stored in the storage data and a corresponding influence coefficient of the type attribute images and the keywords of the short videos through an analysis server, calculating a comprehensive matching degree coincidence coefficient of the short videos, comparing the comprehensive matching degree coincidence coefficient of the short videos with matching degree coincidence coefficients of other short videos, and screening other short videos with the highest matching degree coincidence coefficient for recommendation.
In a possible design of the first aspect, the image processing technique in step S2 includes performing geometric normalization processing on each video image in the short video, transforming the video image into each video image in a fixed standard format, performing multiple different filtering frames on each category attribute image contour in each transformed video image, extracting each category attribute image contour filtering frame in each video image, removing images except the each category attribute image contour filtering frame in each video image, obtaining each category attribute image in each video image, and performing filtering noise reduction processing and enhancement processing on each category attribute image in each video image.
In a possible design of the first aspect, step S3 includes forming a set S of areas of attribute images of each category in each video image in the short videoiP(sip1,sip2,...,sipj,...,sipm)，sipjExpressed as the area of the jth category attribute image within the ith video image in the short video.
In a possible design of the first aspect, the area ratio of each category attribute image in each video image in the short video is calculated as
kipjExpressed as the area ratio, s, of the jth category attribute image in the ith video image in the short videoipjExpressed as the area of the jth class-attribute image within the ith video image in the short video, SSign boardExpressed as a standard area of the image in a fixed standard format.
In one possible design of the first aspect, the type attribute image weight scaling factor of the short video is calculated as
Xi is expressed as a type attribute image weight proportion coefficient of the short video, epsilon is expressed as a weight compensation coefficient corresponding to the type attribute image, n is expressed as the total number of divided video images in the short video, and xjThe number of video images, j, represented as the jth category attribute image in the short video, is 1, 2.
In a possible design of the first aspect, the word frequency calculation formula of each valid word in the short video is
frWord frequency, y, expressed as the r-th valid word in short videorExpressed as the number of occurrences of the r-th significant word in the short video.
In one possible design of the first aspect, the keyword weight scaling factor of the short video is calculated as
ξ' is represented as the keyword weight scaling factor for short videos,
weight compensation expressed as keyword correspondencesCoefficient, yrIs expressed as the number of occurrences of the r-th valid word, f 'in the short video'uThe term frequency, u ═ 1,2,. the term frequency, l, f, expressed as the u-th keyword in short videorExpressed as the word frequency of the r-th valid word in the short video.
In a possible design of the first aspect, the comprehensive matching degree of the short video conforms to a coefficient calculation formula
Phi represents the comprehensive matching degree coincidence coefficient of the short video, alpha and beta respectively represent the type attribute image of the short video and the influence coefficient corresponding to the key words, xi represents the type attribute image weight proportion coefficient of the short video, xi' represents the key word weight proportion coefficient of the short video, mu represents the influence proportion coefficient of the video playing time length ratio, t represents the total matching degree coincidence coefficient of the short video, alpha and beta represent the key word corresponding influence coefficient of the short video respectively, xi represents the key word weight proportion coefficient of the short video, phi represents the key word weight proportion coefficient of the video playing time length ratio, anduthe video playing time length T corresponding to the u-th keyword in the short videoGeneral assemblyExpressed as the total duration of the play of the short video.
In a second aspect, the invention further provides a short video search matching recommendation system based on multi-dimensional data depth comparison analysis, which comprises a short video import module, a video image processing module, an image area measuring module, a category attribute image analysis module, a video voice acquisition module, a text information analysis module, an effective word frequency analysis module, a video duration acquisition module, an analysis server and a storage database;
the short video import module is respectively connected with the video image processing module and the video voice acquisition module, the image area measurement module is respectively connected with the video image processing module and the category attribute image analysis module, the character information analysis module is respectively connected with the video voice acquisition module and the effective word frequency analysis module, and the analysis server is respectively connected with the category attribute image analysis module, the effective word frequency analysis module, the video duration acquisition module and the storage database.
In a third aspect, the present invention further provides a computer storage medium, where a computer program is burned in the computer storage medium, and when the computer program runs in a memory of a server, the method of the present invention is implemented.
Has the advantages that:
(1) the invention provides a short video searching, matching and recommending method, a system and a computer storage medium based on multidimensional data depth comparison and analysis, which divide imported short videos into video images, count the area of each category of attribute image in each video image in the short videos, calculate the area ratio of each category of attribute image in each video image in the short videos, analyze the type attribute image corresponding to each video image in the short videos, count the number of video images of each category of attribute image in the short videos, calculate the weight proportion coefficient of the type attribute image of the short videos, provide guiding reference data for calculating the comprehensive matching degree conformity coefficient of the short videos at the later stage, simultaneously obtain the video voice text information corresponding to each video image in the short videos, obtain each effective text information in each section of video text information in the short videos, contrastively analyze the word frequency of each keyword in the short videos, and calculating the keyword weight proportion coefficient of the short video, and providing reliable reference data for calculating the comprehensive matching degree coincidence coefficient of the short video at the later stage, thereby effectively avoiding the problems of information lag and time delay and realizing targeted real-time recommendation of the user interest points.
(2) According to the short video recommendation method and device, the video playing duration corresponding to each keyword in the short video is obtained, the comprehensive matching degree coincidence coefficient of the short video is calculated, and other short videos with the highest matching degree coincidence coefficient are compared and screened for recommendation, so that the actual requirements of users are met, the accuracy of short video recommendation is improved, the short video watching experience and interest of the users are increased, and the intelligent matching recommendation level of the short videos is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of the method steps of the present invention;
fig. 2 is a schematic view of a module connection structure according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious 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.
Referring to fig. 1, a first aspect of the present invention provides a short video search matching recommendation method based on multidimensional data depth comparison analysis, including the following steps:
s1, short video division: the method comprises the steps of importing the short video watched by a user through a short video importing module, dividing the imported short video into video images according to a set video frame dividing mode, numbering the video images in the short video in sequence according to video time sequence, and numbering the video images in the short video as a1,a2,...,ai,...,an，aiDenoted as the ith video image in the short video.
S2, video image processing: processing each video image in the short video by a video image processing module, processing each video image in the short video by adopting an image processing technology to obtain each category attribute image in each video image in the short video, and forming each category attribute image set A in each video image in the short videoiP(aip1,aip2,...,aipj,...,aipm)，aipjRepresented as the jth category attribute image within the ith video image in the short video.
In this embodiment, the image processing technique in step S2 includes performing geometric normalization processing on each video image in the short video, transforming the video image into each video image in a fixed standard format, simultaneously performing multiple different filter frame screening on each category attribute image contour in each transformed video image, extracting each category attribute image contour filter frame in each video image, removing the images outside each category attribute image contour filter frame in each video image, obtaining each category attribute image in each video image, and performing filtering noise reduction processing and enhancement processing on each category attribute image in each video image.
S3, image area ratio analysis: the area of each category of attribute image in each video image in the short video is measured through an image area measuring module, the area of each category of attribute image in each video image in the short video is counted, the standard area of the image in a fixed standard form stored in a storage database is extracted, and the area ratio of each category of attribute image in each video image in the short video is calculated.
In this embodiment, the step S3 includes an area set S for forming attribute images of each category in each video image in the short videoiP(sip1,sip2,...,sipj,...,sipm)，sipjExpressed as the area of the jth category attribute image within the ith video image in the short video.
In this embodiment, the area ratio calculation formula of each category attribute image in each video image in the short video is
kipjExpressed as the area ratio, s, of the jth category attribute image in the ith video image in the short videoipjExpressed as the area of the jth class-attribute image within the ith video image in the short video, SSign boardExpressed as a standard area of the image in a fixed standard format.
S4, category attribute image statistics: comparing the area ratios of various types of attribute images in various video images in the short video with each other through a category attribute image analysis module, screening the type attribute image with the largest area ratio in various video images in the short video to obtain the category attribute image corresponding to various video images in the short video, and forming various categories of attribute images in the short videoSet of number of video images X (X) of attribute images1,x2,...,xj,...,xm)，xjThe number of video images represented as the jth category attribute image in the short video.
S5, analyzing image weight proportion coefficient: and extracting the weight compensation coefficient corresponding to the type attribute image stored in the storage database through the analysis server, and calculating the weight proportion coefficient of the type attribute image of the short video.
In this embodiment, the formula for calculating the type attribute image weight proportion coefficient of the short video is
Xi is expressed as a type attribute image weight proportion coefficient of the short video, epsilon is expressed as a weight compensation coefficient corresponding to the type attribute image, n is expressed as the total number of divided video images in the short video, and xjThe number of video images, j, represented as the jth category attribute image in the short video, is 1, 2.
Specifically, the imported short video is divided into video images, the area of each category of attribute image in each video image in the short video is counted, the area ratio of each category of attribute image in each video image in the short video is calculated, the type attribute image corresponding to each video image in the short video is analyzed, the number of video images of each category of attribute image in the short video is counted, the type attribute image weight proportion coefficient of the short video is calculated, and guiding reference data are provided for calculating the comprehensive matching degree conformity coefficient of the short video in the later period.
S6, acquiring video character information: respectively acquiring video voice corresponding to each video image in the short video through a video voice acquisition module, identifying video voice text information corresponding to each video image in the short video, and forming each video text information set B (B) in the short video1,b2,...,bi,...,bn)，biIndicated as ith video text information in the short video.
S7, counting the number of effective words: word segmentation processing is carried out on each segment of video word information in short video through word information analysis moduleObtaining each effective word in each section of video character information in the short video, and counting the occurrence frequency of each effective word in the short video to form an occurrence frequency set Y (Y) of each effective word in the short video1,y2,...,yr,...,yv)，yrExpressed as the number of occurrences of the r-th significant word in the short video.
S8, effective word frequency analysis: respectively calculating the word frequency of each effective word in the short video through an effective word frequency analysis module, comparing the word frequency of each effective word in the short video with the standard word frequency of a set keyword, if the word frequency of a certain effective word in the short video is greater than or equal to the standard word frequency of the set keyword, indicating that the effective word in the short video is the keyword, counting the word frequency of each keyword in the short video, and forming a word frequency set f ' (f ') of each keyword in the short video '1,f′2,...,f′u,...,f′l),l≤v，f′uExpressed as the word frequency of the u-th keyword in the short video.
In this embodiment, the word frequency calculation formula of each valid word in the short video is
frWord frequency, y, expressed as the r-th valid word in short videorExpressed as the number of occurrences of the r-th significant word in the short video.
S9, character weight proportion coefficient analysis: and extracting the weight compensation coefficient corresponding to the keyword stored in the storage database through the analysis server, and calculating the keyword weight proportion coefficient of the short video.
In this embodiment, the formula for calculating the keyword weight proportion coefficient of the short video is
ξ' is represented as the keyword weight scaling factor for short videos,
expressed as the weight compensation coefficient, y, corresponding to the keywordrRepresented as short videoNumber of occurrences of the r-th significant word, f'uThe term frequency, u ═ 1,2,. the term frequency, l, f, expressed as the u-th keyword in short videorExpressed as the word frequency of the r-th valid word in the short video.
Specifically, the method obtains effective words in each section of video word information in the short video by obtaining the video voice word information corresponding to each video image in the short video, contrasts and analyzes the word frequency of each keyword in the short video, calculates the keyword weight proportion coefficient of the short video, and provides reliable reference data for calculating the comprehensive matching degree coincidence coefficient of the short video in the later period, thereby effectively avoiding the problems of information lag and time delay, and realizing the targeted real-time recommendation of the user interest points.
S10, acquiring video playing time length: the video playing time length corresponding to each keyword in the short video is obtained through the video time length obtaining module to form a video playing time length set T (T) corresponding to each keyword in the short video1,t2,...,tu,...,tl)，tuAnd the video playing time length corresponding to the u-th keyword in the short video is represented.
S11, short video matching conformity coefficient analysis: the method comprises the steps of extracting an influence proportion coefficient of the total playing time length and the video playing time length ratio of the short videos stored in the storage data and a corresponding influence coefficient of the type attribute images and the keywords of the short videos through an analysis server, calculating a comprehensive matching degree coincidence coefficient of the short videos, comparing the comprehensive matching degree coincidence coefficient of the short videos with matching degree coincidence coefficients of other short videos, and screening other short videos with the highest matching degree coincidence coefficient for recommendation.
In this embodiment, the calculation formula of the comprehensive matching degree coincidence coefficient of the short video is
Psi is a comprehensive matching degree coincidence coefficient of the short video, alpha and beta are type attribute images and key corresponding influence coefficients of the short video, xi is a type attribute image weight proportion coefficient of the short video, xi' is a key weight proportion coefficient of the short video, and mu isInfluence proportion coefficient of video playing time ratio, tuThe video playing time length T corresponding to the u-th keyword in the short videoGeneral assemblyExpressed as the total duration of the play of the short video.
According to the short video recommendation method and device, the video playing duration corresponding to each keyword in the short video is obtained, the comprehensive matching degree coincidence coefficient of the short video is calculated, and other short videos with the highest matching degree coincidence coefficient are compared and screened for recommendation, so that the actual requirements of users are met, the accuracy of short video recommendation is improved, the short video watching experience and interest of the users are increased, and the intelligent matching recommendation level of the short videos is improved.
In a second aspect, the invention further provides a short video search matching recommendation system based on multi-dimensional data depth comparison analysis, which comprises a short video import module, a video image processing module, an image area measuring module, a category attribute image analysis module, a video voice acquisition module, a text information analysis module, an effective word frequency analysis module, a video duration acquisition module, an analysis server and a storage database;
the short video import module is respectively connected with the video image processing module and the video voice acquisition module, the image area measurement module is respectively connected with the video image processing module and the category attribute image analysis module, the character information analysis module is respectively connected with the video voice acquisition module and the effective word frequency analysis module, and the analysis server is respectively connected with the category attribute image analysis module, the effective word frequency analysis module, the video duration acquisition module and the storage database.
In a third aspect, the present invention further provides a computer storage medium, where a computer program is burned in the computer storage medium, and when the computer program runs in a memory of a server, the method of the present invention is implemented.
The foregoing is merely exemplary and illustrative of the principles of the present invention and various modifications, additions and substitutions of the specific embodiments described herein may be made by those skilled in the art without departing from the principles of the present invention or exceeding the scope of the claims set forth herein.
Claims (10)
1. The short video search matching recommendation method based on the multidimensional data depth comparison analysis is characterized by comprising the following steps of: the method comprises the following steps:
s1, short video division: the method comprises the steps of importing the short video watched by a user through a short video importing module, dividing the imported short video into video images according to a set video frame dividing mode, numbering the video images in the short video in sequence according to video time sequence, and numbering the video images in the short video as a1,a2,...,ai,...,an，aiExpressed as the ith video image in the short video;
s2, video image processing: processing each video image in the short video by a video image processing module, processing each video image in the short video by adopting an image processing technology to obtain each category attribute image in each video image in the short video, and forming each category attribute image set A in each video image in the short videoiP(aip1,aip2,...,aipj,...,aipm)，aipjRepresenting the image as the jth category attribute image in the ith video image in the short video;
s3, image area ratio analysis: measuring the area of each category attribute image in each video image in the short video through an image area measuring module, counting the area of each category attribute image in each video image in the short video, extracting the standard area of the image stored in a storage database in a fixed standard form, and calculating the area ratio of each category attribute image in each video image in the short video;
s4, category attribute image statistics: comparing the area ratios of various types of attribute images in various video images in the short video with each other through a category attribute image analysis module, screening the type attribute image with the largest area ratio in various video images in the short video to obtain the category attribute image corresponding to various video images in the short video, and forming the view of various types of attribute images in the short videoSet of number of frequency images X (X)1,x2,...,xj,...,xm)，xjThe number of video images represented as jth category attribute images in the short video;
s5, analyzing image weight proportion coefficient: extracting a weight compensation coefficient corresponding to the type attribute image stored in the storage database through an analysis server, and calculating a weight proportion coefficient of the type attribute image of the short video;
s6, acquiring video character information: respectively acquiring video voice corresponding to each video image in the short video through a video voice acquisition module, identifying video voice text information corresponding to each video image in the short video, and forming each video text information set B (B) in the short video1,b2,...,bi,...,bn)，biThe text information is represented as the ith video text information in the short video;
s7, counting the number of effective words: performing word segmentation processing on each segment of video character information in the short video through a character information analysis module to obtain each effective word in each segment of video character information in the short video, and counting the occurrence frequency of each effective word in the short video to form an occurrence frequency set Y (Y) of each effective word in the short video1,y2,...,yr,...,yv)，yrExpressed as the number of occurrences of the r-th valid word in the short video;
s8, effective word frequency analysis: respectively calculating the word frequency of each effective word in the short video through an effective word frequency analysis module, comparing the word frequency of each effective word in the short video with the standard word frequency of a set keyword, if the word frequency of a certain effective word in the short video is greater than or equal to the standard word frequency of the set keyword, indicating that the effective word in the short video is the keyword, counting the word frequency of each keyword in the short video, and forming a word frequency set f' (f) of each keyword in the short video1′,f′2,...,f′u,...,fl′),l≤v，fu' word frequency expressed as the u-th keyword in the short video;
s9, character weight proportion coefficient analysis: extracting a weight compensation coefficient corresponding to a keyword stored in a storage database through an analysis server, and calculating a keyword weight proportion coefficient of the short video;
s10, acquiring video playing time length: the video playing time length corresponding to each keyword in the short video is obtained through the video time length obtaining module to form a video playing time length set T (T) corresponding to each keyword in the short video1,t2,...,tu,...,tl)，tuRepresenting the video playing time corresponding to the u-th keyword in the short video;
s11, short video matching conformity coefficient analysis: the method comprises the steps of extracting an influence proportion coefficient of the total playing time length and the video playing time length ratio of the short videos stored in the storage data and a corresponding influence coefficient of the type attribute images and the keywords of the short videos through an analysis server, calculating a comprehensive matching degree coincidence coefficient of the short videos, comparing the comprehensive matching degree coincidence coefficient of the short videos with matching degree coincidence coefficients of other short videos, and screening other short videos with the highest matching degree coincidence coefficient for recommendation.
2. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the image processing technique in step S2 includes performing geometric normalization processing on each video image in the short video, converting the video image into each video image in a fixed standard format, performing multiple different screening frames on each category attribute image contour in each converted video image, extracting each category attribute image contour screening frame in each video image, removing images except the each category attribute image contour screening frame in each video image, obtaining each category attribute image in each video image, and performing filtering noise reduction processing and enhancement processing on each category attribute image in each video image.
3. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the step S3 includes an area set S for each type of attribute image in each video image in the short videoiP(sip1,sip2,...,sipj,...,sipm)，sipjExpressed as the area of the jth category attribute image within the ith video image in the short video.
4. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the area ratio calculation formula of each category attribute image in each video image in the short video is
kipjExpressed as the area ratio, s, of the jth category attribute image in the ith video image in the short videoipjExpressed as the area of the jth class-attribute image within the ith video image in the short video, SSign boardExpressed as a standard area of the image in a fixed standard format.
5. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the type attribute image weight proportion coefficient calculation formula of the short video is
Xi is expressed as a type attribute image weight proportion coefficient of the short video, epsilon is expressed as a weight compensation coefficient corresponding to the type attribute image, n is expressed as the total number of divided video images in the short video, and xjThe number of video images, j, represented as the jth category attribute image in the short video, is 1, 2.
6. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the word frequency calculation formula of each effective word in the short video is
frExpressed as the r-th significant in short videoWord frequency, y, of wordsrExpressed as the number of occurrences of the r-th significant word in the short video.
7. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the short video keyword weight proportion coefficient calculation formula is
ξ' is represented as the keyword weight scaling factor for short videos,
expressed as the weight compensation coefficient, y, corresponding to the keywordrExpressed as the number of occurrences of the r-th significant word in the short video, fu' term frequency, u ═ 1,2,. 1, l, f, expressed as the u-th keyword in short videorExpressed as the word frequency of the r-th valid word in the short video.
8. The short video search matching recommendation method based on multi-dimensional data depth comparison analysis according to claim 1, characterized in that: the calculation formula of the comprehensive matching degree coincidence coefficient of the short video is
Phi represents the comprehensive matching degree coincidence coefficient of the short video, alpha and beta respectively represent the type attribute image of the short video and the influence coefficient corresponding to the key words, xi represents the type attribute image weight proportion coefficient of the short video, xi' represents the key word weight proportion coefficient of the short video, mu represents the influence proportion coefficient of the video playing time length ratio, t represents the total matching degree coincidence coefficient of the short video, alpha and beta represent the key word corresponding influence coefficient of the short video respectively, xi represents the key word weight proportion coefficient of the short video, phi represents the key word weight proportion coefficient of the video playing time length ratio, anduthe video playing time length T corresponding to the u-th keyword in the short videoGeneral assemblyExpressed as the total duration of the play of the short video.
9. Short video search matches recommendation system based on multidimensional data depth comparison analysis, its characterized in that: the system comprises a short video import module, a video image processing module, an image area measuring module, a category attribute image analysis module, a video voice acquisition module, a character information analysis module, an effective word and frequency analysis module, a video duration acquisition module, an analysis server and a storage database;
the short video import module is respectively connected with the video image processing module and the video voice acquisition module, the image area measurement module is respectively connected with the video image processing module and the category attribute image analysis module, the character information analysis module is respectively connected with the video voice acquisition module and the effective word frequency analysis module, and the analysis server is respectively connected with the category attribute image analysis module, the effective word frequency analysis module, the video duration acquisition module and the storage database.
10. A computer storage medium, characterized in that: the computer storage medium is burned with a computer program, which when run in the memory of the server implements the method of any of the above claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110595730.0A CN113204709A (en) | 2021-05-29 | 2021-05-29 | Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110595730.0A CN113204709A (en) | 2021-05-29 | 2021-05-29 | Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113204709A true CN113204709A (en) | 2021-08-03 |
Family
ID=77023590
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110595730.0A Withdrawn CN113204709A (en) | 2021-05-29 | 2021-05-29 | Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113204709A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114398514A (en) * | 2021-12-24 | 2022-04-26 | 北京达佳互联信息技术有限公司 | Video display method and device and electronic equipment |
| CN114639051A (en) * | 2022-03-22 | 2022-06-17 | 武汉元淳传媒有限公司 | Advertisement short video quality evaluation method and system based on big data analysis and storage medium |
| CN114745593A (en) * | 2022-04-08 | 2022-07-12 | 南京福田文化传媒有限公司 | Video playing recommendation system and method based on video big data |
| CN114897566A (en) * | 2022-03-21 | 2022-08-12 | 晨雨初听(武汉)文化艺术传播有限公司 | Short video compliance online diagnosis and analysis method and diagnosis and analysis system based on big data |
| CN115174951A (en) * | 2022-07-02 | 2022-10-11 | 武汉庆实广告传媒有限公司 | Unmanned live broadcast online analysis and management system based on multi-dimensional feature capture |
| CN115209233A (en) * | 2022-06-25 | 2022-10-18 | 平安银行股份有限公司 | Video playing method and related device and equipment |
| CN116744063A (en) * | 2023-08-15 | 2023-09-12 | 四川中电启明星信息技术有限公司 | Short video push system integrating social attribute information |
| CN116887010A (en) * | 2023-05-08 | 2023-10-13 | 武汉精阅数字传媒科技有限公司 | Self-media short video material processing control system |
| CN117807270A (en) * | 2024-02-29 | 2024-04-02 | 中国人民解放军国防科技大学 | Video recommendation method, device, equipment and storage medium based on news content |
-
2021
- 2021-05-29 CN CN202110595730.0A patent/CN113204709A/en not_active Withdrawn
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114398514A (en) * | 2021-12-24 | 2022-04-26 | 北京达佳互联信息技术有限公司 | Video display method and device and electronic equipment |
| CN114398514B (en) * | 2021-12-24 | 2022-11-22 | 北京达佳互联信息技术有限公司 | Video display method and device and electronic equipment |
| CN114897566A (en) * | 2022-03-21 | 2022-08-12 | 晨雨初听(武汉)文化艺术传播有限公司 | Short video compliance online diagnosis and analysis method and diagnosis and analysis system based on big data |
| CN114639051A (en) * | 2022-03-22 | 2022-06-17 | 武汉元淳传媒有限公司 | Advertisement short video quality evaluation method and system based on big data analysis and storage medium |
| CN114745593A (en) * | 2022-04-08 | 2022-07-12 | 南京福田文化传媒有限公司 | Video playing recommendation system and method based on video big data |
| CN114745593B (en) * | 2022-04-08 | 2023-12-29 | 广州星轩传媒有限公司 | Video playing recommendation system and method based on video big data |
| CN115209233A (en) * | 2022-06-25 | 2022-10-18 | 平安银行股份有限公司 | Video playing method and related device and equipment |
| CN115209233B (en) * | 2022-06-25 | 2023-08-25 | 平安银行股份有限公司 | Video playing method, related device and equipment |
| CN115174951A (en) * | 2022-07-02 | 2022-10-11 | 武汉庆实广告传媒有限公司 | Unmanned live broadcast online analysis and management system based on multi-dimensional feature capture |
| CN116887010A (en) * | 2023-05-08 | 2023-10-13 | 武汉精阅数字传媒科技有限公司 | Self-media short video material processing control system |
| CN116887010B (en) * | 2023-05-08 | 2024-02-02 | 杭州元媒科技有限公司 | Self-media short video material processing control system |
| CN116744063A (en) * | 2023-08-15 | 2023-09-12 | 四川中电启明星信息技术有限公司 | Short video push system integrating social attribute information |
| CN116744063B (en) * | 2023-08-15 | 2023-11-03 | 四川中电启明星信息技术有限公司 | Short video push system integrating social attribute information |
| CN117807270A (en) * | 2024-02-29 | 2024-04-02 | 中国人民解放军国防科技大学 | Video recommendation method, device, equipment and storage medium based on news content |
| CN117807270B (en) * | 2024-02-29 | 2024-05-07 | 中国人民解放军国防科技大学 | Video recommendation method, device, equipment and storage medium based on news content |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113204709A (en) | Short video search matching recommendation method and system based on multidimensional data depth comparison analysis and computer storage medium | |
| US8965893B2 (en) | System and method for grouping multiple streams of data | |
| CN103559206B (en) | A kind of information recommendation method and system | |
| JP3116851B2 (en) | Information filtering method and apparatus | |
| CN107341268B (en) | Hot searching ranking method and system | |
| CN111324801B (en) | Hot event discovery method in judicial field based on hot words | |
| CN108471544B (en) | Method and device for constructing video user portrait | |
| CN109325115B (en) | Role analysis method and analysis system | |
| WO2017075912A1 (en) | News events extracting method and system | |
| CN113449146A (en) | Short video browsing recommendation method and device based on data analysis and computer storage medium | |
| CN112632396A (en) | Article recommendation method and device, electronic equipment and readable storage medium | |
| CN112102038A (en) | Optimization method for live broadcast e-commerce platform user access database based on big data | |
| CN114201598B (en) | Text recommendation method and text recommendation device | |
| KR100788642B1 (en) | Texture analysing method of digital image | |
| CN106294797A (en) | A kind of generation method and apparatus of video gene | |
| CN114588633A (en) | Content recommendation method | |
| CN109740156B (en) | Feedback information processing method and device, electronic equipment and storage medium | |
| Torres-Tramón et al. | Topic detection in Twitter using topology data analysis | |
| CN110704738A (en) | Service information pushing method and device based on judge portrait, terminal and storage medium | |
| CN115630173B (en) | User data management method based on interestingness analysis | |
| CN113282641A (en) | Webpage search data information intelligent classification management method and system based on user behavior deep analysis and computer storage medium | |
| Díaz et al. | A comparative approach between different computer vision tools, including commercial and open-source, for improving cultural image access and analysis | |
| US20240169004A1 (en) | Methods and systems for self-tuning personalization engines in near real-time | |
| CN116739654A (en) | Information acquisition device and system | |
| US20220292127A1 (en) | Information management system |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20210803 |