CN109726294A - A kind of App entity alignment schemes, device and electronic equipment - Google Patents
A kind of App entity alignment schemes, device and electronic equipment Download PDFInfo
- Publication number
- CN109726294A CN109726294A CN201811472348.5A CN201811472348A CN109726294A CN 109726294 A CN109726294 A CN 109726294A CN 201811472348 A CN201811472348 A CN 201811472348A CN 109726294 A CN109726294 A CN 109726294A
- Authority
- CN
- China
- Prior art keywords
- app
- aligned
- entity
- picture
- name
- 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
Landscapes
- User Interface Of Digital Computer (AREA)
Abstract
The embodiment of the invention provides a kind of App entity alignment schemes, device and electronic equipments.Wherein, this method comprises: according to the name of App entity in knowledge mapping, App entity is aligned, obtains two or more consistent App entities to be aligned of name；Obtain the icon picture of each App entity to be aligned；The icon picture of each App entity to be aligned is compared, judges whether the icon picture between each App entity to be aligned matches；If the icon picture between each App entity to be aligned matches, the App entity to be aligned is determined as the same App entity.It can fast and accurately realize that App entity is aligned using the embodiment of the present invention.
Description
Technical field
The present invention relates to the technical fields being managed to the App entity in knowledge mapping, real more particularly to a kind of App
Body alignment schemes, device and electronic equipment.
Background technique
Knowledge mapping is a kind of semantic network of relationship between announcement entity, can things to real world and its mutually
Relationship is formally described.Present knowledge mapping has been used to refer to various large-scale knowledge bases.
Currently, network A pp (Application, application program) quantity is various, know to preferably manage App and can establish
Know map.Knowledge mapping would generally be established first according to the information of each App, in knowledge mapping, the corresponding App of each App
Entity.Then, the alignment of App entity then to knowledge mapping is carried out, obtains final knowledge mapping.
In the prior art, the mode for carrying out the alignment of App entity to knowledge mapping is to be carried out according to the name of App entity pair
Together.Specific method is: for each App entity in knowledge mapping, calculating each App entity using text similarity algorithm
Multiple App entities that similarity is greater than preset threshold are determined as same by the similarity of name according to the similarity of name
App entity only retains an App entity in knowledge mapping, deletes other App entities identical with the App entity.
The name of App entity is generated according to the description information of App, some App when being adjusted and updating, due to
For the different based on different operating system describing mode of same App, or for same App, different operating system is to it
The description for carrying out different angle, will lead to the name that multiple App entities are generated to same App, and the name of some App entities is poor
It is very not big.Such as: the name of iqiyi.com App, Android client are " the new song of iqiyi.com-China ", and the name of Ios client is
" iqiyi.com-prolongs auspiciousness strategy ".Since the name difference of the two App entities is very big, it is aligned according to the name of App entity
When, the two App entities are not determined to the same App entity, that is, App entity occur and be aligned inaccurate problem.As it can be seen that
The method accuracy that the prior art carries out entity alignment according to the name of App entity is not high.
Summary of the invention
The embodiment of the present invention is designed to provide a kind of App entity alignment schemes, device and electronic equipment, to improve
The accuracy of App entity alignment.Specific technical solution is as follows:
In order to achieve the above objectives, in a first aspect, the present invention provides a kind of App entity alignment schemes, the method includes；
According to the name of the App entity in knowledge mapping, App entity is aligned, obtain name it is consistent two or
More than two App entities to be aligned；
Obtain the icon picture of each App entity to be aligned；
The icon picture of each App entity to be aligned is compared, judge each App entity to be aligned it
Between icon picture whether match；
If the icon picture between each App entity to be aligned matches, the App entity to be aligned is determined as
The same App entity.
Optionally, the name according to the App entity in knowledge mapping, is aligned App entity, obtains name one
The step of two or more App entities to be aligned caused, comprising:
Obtain the name of each App entity in the knowledge mapping；
Calculate the name similarity in the knowledge mapping between the name of each App entity；
Obtain two or more App entities that the name similarity is less than default name similarity threshold；By institute
It is consistent to be aligned as name less than two or more App entities of default name similarity threshold to state name similarity
App entity.
Optionally, the icon picture to each App entity to be aligned is compared, and judges that each App to be aligned is real
The step of whether icon picture between body matches, comprising:
Calculate the picture diversity factor between the icon picture of each App entity to be aligned；
Judge whether calculated each picture diversity factor is less than preset threshold；
Determine the icon picture matching that picture diversity factor is less than between each App entity to be aligned of preset threshold.
Optionally, the step of calculating the picture diversity factor between the icon picture of each App entity to be aligned, packet
It includes:
Calculate the Hamming distance between the icon picture of each App entity to be aligned, using calculated Hamming distance as
Picture diversity factor between icon picture.
Optionally, the step of Hamming distance between the icon picture for calculating each App entity to be aligned, comprising:
The icon picture size of each App entity to be aligned is converted into preset icon picture normal size；
The grayscale image of the icon picture of each App entity to be aligned is obtained respectively；
Based on the grayscale image, the binary map of the icon picture of each App entity to be aligned is obtained respectively；
The binary map is indicated with vector form respectively；
It is indicated based on the vector form as a result, the Hamming distance between the App entity icon picture is calculated, as figure
Piece diversity factor.
Optionally, the step of calculating the picture diversity factor between the icon picture of each App entity to be aligned, packet
It includes:
The Jie Kade distance between the icon picture of each App entity to be aligned is calculated, by calculated Jie Kade distance
As the picture diversity factor between icon picture.
Optionally, the Jie Kade between the icon picture for calculating each App entity to be aligned apart from the step of, packet
It includes:
The icon picture size of each App entity to be aligned is converted into the preset icon picture normal size；
The grayscale image of the icon picture of each App entity to be aligned is obtained respectively；
Based on the grayscale image, the key point and feature of the icon picture of each App entity to be aligned are obtained respectively
Vector；
Based on the key point and feature vector, the set of characteristic points being mutually matched are obtained respectively；
Based on the set of characteristic points being mutually matched, the Jie Kade of the icon picture of each App entity to be aligned is calculated
Distance, as picture diversity factor.
Second aspect, the present invention also provides a kind of App entity alignment means, described device includes:
Name alignment module is aligned App entity for the name according to the App entity in knowledge mapping, obtains
Two or more consistent App entities to be aligned of name；
Icon picture obtains module, for obtaining the icon picture of each App entity to be aligned；
Icon picture comparison module is compared for the icon picture to each App entity to be aligned, and judgement is each
Whether the icon picture between a App entity to be aligned matches；
Determining module, if for the icon picture matching between each App entity to be aligned, it will be described to be aligned
App entity is determined as the same App entity.
Optionally, the name alignment module includes:
Name acquiring unit, for obtaining the name of each App entity in the knowledge mapping；
Name similarity calculated, for calculating the name in the knowledge mapping between the name of each App entity
Similarity；
Name alignment unit, the similarity for obtaining the name are less than two or two of default name similarity threshold
A above App entity；The similarity of the name is less than to two or more App entities of default name similarity threshold
As the consistent App entity to be aligned of name.
Optionally, the icon picture comparison module includes:
Icon picture diversity factor computing unit, between the icon picture for calculating each App entity to be aligned
Picture diversity factor；
Picture diversity factor judging unit, for judging whether calculated each picture diversity factor is less than preset threshold；
Determination unit, the icon being less than between each App entity to be aligned of preset threshold for determining picture diversity factor
Picture match.
Optionally, the icon picture diversity factor computing unit is specifically used for calculating the icon of each App entity to be aligned
Hamming distance between picture；Using calculated Hamming distance as the picture diversity factor between icon picture, comprising:
First size conversion subunit, it is real for the App entity icon picture size to be converted to the preset App
Body icon picture normal size；
First grayscale image obtains subelement, for obtaining the grayscale image of the App entity icon picture respectively；
Binary map obtains subelement, for being based on the grayscale image, obtains the two-value of the App entity icon picture respectively
Figure；
Vector indicates subelement, for indicating the binary map with vector form；
Hamming distance computation subunit, for indicating to mark on a map as a result, calculating the App sterogram based on the vector form
Hamming distance between piece, using calculated Hamming distance as picture diversity factor.
Optionally, the icon picture diversity factor computing unit is specifically used for calculating the icon of each App entity to be aligned
Jie Kade distance between picture regard calculated Jie Kade distance as picture diversity factor, comprising:
Second size conversion subunit, it is real for the App entity icon picture size to be converted to the preset App
Body icon picture normal size；
Second grayscale image obtains subelement, for obtaining the grayscale image of the App entity icon picture respectively；
Key point obtains subelement, for being based on the grayscale image, obtains the key of the App entity icon picture respectively
Point and feature vector；
Characteristic point obtains subelement, for being based on the key point and feature vector, obtains the feature being mutually matched respectively
Point set；
Jie Kade is apart from computation subunit, for calculating the App entity icon based on the set of characteristic points being mutually matched
Jie Kade distance between picture regard calculated Jie Kade distance as picture diversity factor.
Present invention implementation additionally provides a kind of computer readable storage medium, storage in the computer readable storage medium
There is computer program, the computer program realizes a kind of any of the above-described App entity alignment side when being executed by processor
The step of method.
The embodiment of the invention also provides a kind of computer programs comprising instruction to make when run on a computer
It obtains computer and executes a kind of any of the above-described App entity alignment schemes.
A kind of App entity alignment schemes, device and electronic equipment provided in an embodiment of the present invention, according in knowledge mapping
The name of App entity carries out App entity to obtain App entity to be aligned after name alignment, obtains each App to be aligned respectively
The icon picture of entity is compared the icon picture of each App entity to be aligned, judge each App entity to be aligned it
Between icon picture whether match, determine the matched each App entity to be aligned of icon picture be the same App entity.Due to
The icon picture of same App entity differs often very little, therefore compared with the existing technology may be used according to icon picture to be aligned
To improve the accuracy of App entity alignment.
Certainly, implement any of the products of the present invention or method it is not absolutely required at the same reach all the above excellent
Point.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described.
Fig. 1 is a kind of flow chart of App entity alignment schemes provided in an embodiment of the present invention；
Fig. 2 is a kind of specific flow chart of step 101 in embodiment illustrated in fig. 1；
Fig. 3 is that flow chart of the Hamming distance as the picture diversity factor between icon picture is calculated in embodiment illustrated in fig. 1；
Fig. 4 is the process that Jie Kade distance is calculated in embodiment illustrated in fig. 1 as the picture diversity factor between icon picture
Figure；
Fig. 5 is a kind of structure chart of App entity alignment means provided in an embodiment of the present invention；
Fig. 6 is a kind of electronic equipment structure chart provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description.
In order to improve the accuracy of App entity alignment, the embodiment of the invention discloses a kind of App entity alignment schemes, dress
It sets and electronic equipment, is described in detail below.
Referring to Fig. 1, Fig. 1 is a kind of flow chart of App entity alignment schemes provided in an embodiment of the present invention, including following step
It is rapid:
Step 101；According to the name of the App entity in knowledge mapping, App entity is aligned, it is consistent to obtain name
Two or more App entities to be aligned.
Step 102；Obtain the icon picture of each App entity to be aligned.
It is to be aligned using the icon picture of App entity in the embodiment of the present invention.Pass through App entity to be aligned first
Name the initial data for obtaining the icon picture of each App entity to be aligned is searched into database, further according to icon picture
Initial data obtain the icon picture of each App entity to be aligned.
Step 103: the icon picture of each App entity to be aligned is compared, judge each App entity to be aligned it
Between icon picture whether match.
To the concrete mode that the icon picture of each App entity to be aligned is compared, it is each to be aligned to can be calculating
The picture diversity factor of icon picture between App entity, judges whether calculated each picture diversity factor is less than preset threshold,
Determine that picture diversity factor is less than the icon picture matching of each App entity to be aligned of preset threshold.In the embodiment of the present invention, tool
Body can be using Hamming distance between calculation icon picture or Jie Kade distance as picture diversity factor, naturally it is also possible to use other
Calculate picture diversity factor algorithm calculate picture diversity factor, the embodiment of the present invention to specific algorithm without limitation.
Step 104: if the icon picture between each App entity to be aligned matches, the App to be aligned is real
Body is determined as the same App entity.
It, can be matched each to be aligned by icon picture according to the similarity judging result of icon picture in this step
App entity is determined as the same App entity.Retain an App entity, delete other same App entities, completes App entity pair
Together.
A kind of App entity alignment schemes provided in an embodiment of the present invention, App entity is carrying out name alignment in knowledge mapping
App entity to be aligned is obtained later, obtains the icon picture of each App entity to be aligned respectively, by comparing the figure of each App entity
Piece of marking on a map judges whether the icon picture between each App entity matches, and determines the matched each App entity to be aligned of icon picture
The alignment of App entity is completed for the same App entity.Since the icon picture of same App entity differs often very little, basis
The accuracy of App entity alignment can be improved to be aligned compared with the existing technology for icon picture.
In embodiment illustrated in fig. 1, step 101 carries out name alignment according to the App entity in knowledge mapping, obtains name
The step of two or more consistent App entities to be aligned, it specifically may refer to Fig. 2, Fig. 2 is in embodiment illustrated in fig. 1
A kind of specific flow chart of step 101, includes the following steps；
Step 201: obtaining the name of each App entity in knowledge mapping.
The name data for being stored in App entity in knowledge mapping database by reading is each in knowledge mapping to obtain
The name of App entity.
Step 202: the name similarity in calculation knowledge map between the name of each App entity.
Step 203: obtaining two or more App entities that name similarity is less than default name similarity threshold.
Step 204: using name similarity be less than two or more App entities of default name similarity threshold as
The consistent App entity to be aligned of name.
Calculating name similarity can be completed using remaining rotation similarity algorithm, and detailed process can be described as:
(1) keyword of each App physical name is found out；
(2) each keyword of each App entity is merged into a set, calculates each App physical name for this collection
The word frequency of word in conjunction；
(3) each respective word frequency vector of App physical name is generated；
(4) cosine similarity of vector is calculated separately, value means that more greatly more similar.
Such as: the name A of the App entity of acquisition is that " the new song of iqiyi.com-China ", B are " iqiyi.com-prolongs auspiciousness strategy ",
Its calculating process, which may be summarized to be, to be segmented to obtain A=iqiyi.com/China to A and B/new/song B=iqiyi.com/prolonging auspiciousness/is attacked
Slightly, all words are listed to obtain (iqiyi.com China prolongs auspiciousness new strategy song) calculating word frequency, write out word frequency vector A={ 1101
The vector expression of A and B is substituted into formula by 01 } { 101010 } B=:
θ=0.289 cos is obtained, which illustrates that A is closer with B closer to 1, and presetting name similarity threshold is 0.9, then
Calculated result obtained in the present embodiment is less than default name similarity threshold, and it is similar less than default name to obtain name similarity
Spend two App to be aligned of threshold value.
Step 103 in embodiment illustrated in fig. 1 can calculate the Chinese between the icon picture of each App entity to be aligned
Prescribed distance or Jie Kade distance；By calculated Hamming distance or Jie Kade distance as the picture difference between icon picture
Degree.
Wherein, the process for calculating Hamming distance, may refer to Fig. 3, and Fig. 3 is to calculate Hamming distance in embodiment illustrated in fig. 1
Flow chart as the picture diversity factor between icon picture, comprising:
Step 301: the icon picture size of each App entity to be aligned is converted into preset icon picture normal size；
The effect of this step is to remove the details of picture, only retains structure, the essential informations such as light and shade, abandon different sizes,
Ratio bring picture difference, so that calculated Hamming distance is more accurate.
Step 302: obtaining the grayscale image of the icon picture of each App entity to be aligned respectively；
In image procossing, with tri- components (R, G, B) of RGB, i.e. (R:Red, G:Green, B:Blue), i.e. red, green, blue
Three primary colors indicate true color, and R component, G component, the value range of B component is 0~255, allow each in pixel matrix
A pixel all meets lower relation of plane: R=G=B just obtains grayscale image.
Step 303: being based on grayscale image, obtain the binary map of the icon picture of each App entity to be aligned respectively；
Specifically, can first calculate the average value of the pixel value of grayscale image all pixels point in this step, then traverse ash
Each pixel value of degree figure is recorded as 255 (whites) compared with the average value, greater than average value, is otherwise 0 (black), is formed
Binary map, that is, allow whole image that only black and white effect is presented.
Step 304: respectively indicating each binary map with vector form；
Specifically, white portion in binary map can be denoted as to 1 in this step, black portions are denoted as 0 and obtain icon picture
The vector of binary map indicates.
Step 305: indicated based on vector form as a result, calculating the Hamming between each App entity icon picture to be aligned
Distance, as picture diversity factor.
In this step, if App entity to be aligned is two, the Hamming distance value between two App is directly calculated, if to
Being aligned App entity is three or three or more, then App entity to be aligned is carried out comparison two-by-two and obtain Hamming distance between any two
From value.
Such as: App entity to be aligned is iqiyi.com and iqiyi.com pps, and alignment procedure can be described as:
The icon picture for obtaining iqiyi.com App and iqiyi.com pps App first, narrows down to 3*3's for icon picture respectively
Size, 9 pixels in total, to size, treated that icon picture carries out gray processing and binary conversion treatment obtains black and white respectively
Icon picture, is indicated to obtain that iqiyi.com App is { 0,1,0,1,1,1,0,0,0 }, iqiyi.com pps App is with vector form
{ 0,1,1,1,1,1,0,0,1 }, then Hamming distance calculated result is 2, sets predetermined threshold as 5, it is determined that group App to be aligned
Entity is the matched App entity of icon picture.
Such as: App entity to be aligned is iqiyi.com pps, the new song of iqiyi.com-China and iqiyi.com-prolong auspiciousness strategy, right
Neat process can be described as:
The icon that iqiyi.com pps App, the new song App of iqiyi.com-China and iqiyi.com-prolong auspiciousness strategy App is obtained first
Icon picture is narrowed down to the size of 3*3 by picture respectively, in total 9 pixels, respectively to size treated icon picture into
Row gray processing and binary conversion treatment obtain the icon picture of black and white, be indicated to obtain iqiyi.com pps App with vector form be
The new song App of { 0,1,1,1,1,1,0,0,1 }, iqiyi.com-China is { 0,1,0,1,1,1,0,0,0 }, iqiyi.com-prolongs auspiciousness strategy
App is { 0,1,0,1,1,1,0,0,0 }, and App entity to be aligned is compared two-by-two, then iqiyi.com pps App and iqiyi.com-
Chinese new song App Hamming distance calculated result is 2, and iqiyi.com pps App and iqiyi.com-prolong auspiciousness strategy App Hamming distance meter
Calculating result is 2, and it is 0 that the new song App of iqiyi.com-China, which prolongs auspiciousness strategy App Hamming distance calculated result with iqiyi.com-, and setting is pre-
Determining threshold value is 5, it is determined that group App entity to be aligned is the matched App entity of icon picture.
In the present embodiment, App to be aligned is determined by calculating the Hamming distance between App entity icon picture to be aligned
Whether the icon picture between entity matches, and not only accurately determines whether App entity icon picture to be aligned matches, Er Qieji
Calculation amount is small, and the icon that can be realized rapidly and efficiently compares.
Wherein, the process for calculating Jie Kade distance, may refer to Fig. 4, and Fig. 4 is to calculate Jie Kade in embodiment illustrated in fig. 1
Flow chart of the distance as the picture diversity factor between icon picture, comprising:
Step 401: the icon picture size of each App entity to be aligned is converted into preset icon picture normal size.
The effect of this step is to remove the details of picture, only retains structure, the essential informations such as light and shade, abandon different sizes,
Ratio bring picture difference, so that calculated Jie Kade is apart from more accurate.
Step 402: obtaining the grayscale image of the icon picture of each App entity to be aligned respectively.
In image procossing, with tri- components (R:Red, G:Green, B:Blue) of RGB, i.e., Red Green Blue carrys out table
Show true color, R component, G component, the value range of B component is 0~255, allow each of pixel matrix pixel all
Meet lower relation of plane: R=G=B just obtains grayscale image.
Step 403: being based on grayscale image, obtain the key point and feature of the icon picture of each App entity to be aligned respectively
Vector.
Specifically, this step can be based on Sift algorithm (Scale-invariant feature transform, scale
Invariant features conversion) Lai Shixian, Sift algorithm is lookup key point (characteristic point) on different scale spaces, and calculates pass
The direction of key point, realization process includes scale space extremum extracting, crucial point location, direction is determining and key point describes.This
The key point information keypoints=P of the icon picture of each App entity to be aligned is obtained in step based on Sift algorithmn(n
For keypoint quantity) it include size, direction, dimensional information.Again to each key point using the description of 4*4*8 totally 128 dimensional vectors
Son carries out key point and characterizes to obtain feature vector descriptors=Dn×128。
Step 404: being based on key point and feature vector, obtain the set of characteristic points being mutually matched respectively；
The matching of characteristic point is realized by the Euclidean distance for the key point feature vector for calculating 128 dimensions of two groups of characteristic points
's.Matched m set of characteristic points can be expressed as goodMatches={ (PA, PB)1... (PA, PB)m}.Wherein A, B are represented
The icon picture of two App entities to be aligned.
Step 405: based on the set of characteristic points being mutually matched, calculating the outstanding person of the icon picture of each App entity to be aligned
Card moral distance, as picture diversity factor；
Jie Kade distance (Jaccard Distance) is a kind of index for measuring two set difference opposite sex, it is outstanding
The supplementary set of card moral similarity factor, Jie Kade distance account for the ratio of all elements with elements different in two set to measure two collection
The discrimination of conjunction.I.e.
If App entity to be aligned is two, the Jie Kade distance value between two App is directly calculated, if App to be aligned
Entity is three or three or more, then by App entity to be aligned carry out comparison two-by-two respectively obtain Jie Kade between any two away from
From value.
Such as: App entity to be aligned is iqiyi.com and iqiyi.com pps, and alignment procedure can be described as:
The icon picture for obtaining iqiyi.com App and iqiyi.com pps App first, narrows down to 4*4's for icon picture respectively
Size carries out gray processing to size treated icon picture respectively and handles to obtain grayscale image, obtains with SIFT algorithm each
The key point information P of App entity to be aligned16, obtain feature vector D16×128.By calculating iqiyi.com App and iqiyi.com pps
10 set of characteristic points { (P that the Euclidean distance of the key point feature vector of 128 dimensions of two groups of characteristic points of App is mutually matchedIqiyi.com, PIqiyi.com pps)1... (PIqiyi.com, PIqiyi.com pps)10}.Iqiyi.com App and iqiyi.com pps App are regarded as two set, 1 indicates
Set includes this feature point, and 0 indicates that set does not include this feature point, then iqiyi.com App is (1,1,1,1,1,1,1,0,0,1),
Iqiyi.com pps App is that (1,1,1,1,1,1,0,1,1,0) then Jie Kade similarity factor isWherein p: iqiyi.com
App and iqiyi.com pps App is 1 number, and q: iqiyi.com App is 1 and iqiyi.com pps App is 0 number, r: iqiyi.com
App is 0 and iqiyi.com pps App is 1 number.Obtain J=0.6, then Jie Kade distance be d=0.4, set predetermined threshold as
0.5, it is determined that group App entity to be aligned is the matched App entity of icon picture.
In the present embodiment, determined by calculating the Jie Kade distance between App entity to be aligned App entity to be aligned it
Between icon picture whether match, whether the icon picture for not only accurately determining App entity to be aligned matches, but also calculation amount
Small, the icon picture that can be realized rapidly and efficiently compares.
Referring to Fig. 5, Fig. 5 is a kind of structure chart of App entity alignment means of the embodiment of the present invention, comprising:
Name alignment module 501 is aligned App entity for the name according to the App entity in knowledge mapping,
Obtain two or more consistent App entities to be aligned of name；
Icon picture obtains module 502, for obtaining the icon picture of each App entity to be aligned；
Icon picture comparison module 503 is compared for the icon picture to each App entity to be aligned, and judgement is each
Whether the icon picture between a App entity to be aligned matches；
Determining module 504, if for the icon picture matching between each App entity to be aligned, will it is described to
Alignment App entity is determined as the same App entity.
Specifically, name alignment module may include:
Name acquiring unit, for obtaining the name of each App entity in knowledge mapping；
Name similarity calculated, it is similar for the name between the name of App entity each in calculation knowledge map
Degree；
Name alignment unit, the similarity for obtaining name be less than two or two of default name similarity threshold with
Upper App entity；The name similarity is less than two or more App entities of default name similarity threshold as name
The consistent App entity to be aligned of word.
Specifically, icon picture comparison module may include:
Icon picture diversity factor computing unit, the picture between icon picture for calculating each App entity to be aligned
Diversity factor；
Icon picture result judging unit, for judging whether calculated each picture diversity factor is less than preset threshold；
Determination unit, the icon being less than between each App entity to be aligned of preset threshold for determining picture diversity factor
Picture match.
Specifically, in a specific embodiment, icon picture diversity factor computing unit may include:
First size handles subelement, marks on a map for App entity icon picture size to be converted to preset App sterogram
Piece normal size；
First grayscale image obtains subelement, for obtaining the grayscale image of App entity icon picture respectively；
Binary map obtains subelement, for being based on grayscale image, obtains the binary map of the App entity icon picture respectively；
Vector indicates subelement, for indicating binary map with vector form；
Hamming distance computation subunit, for being indicated based on vector form as a result, calculating between App entity icon picture
Hamming distance, using calculated Hamming distance as picture diversity factor.
Specifically, in another specific embodiment, icon picture diversity factor computing unit may include:
Second size handles subelement, marks on a map for converting preset App sterogram for App entity icon picture size
Piece normal size；
Second grayscale image obtains subelement, for obtaining the grayscale image of App entity icon picture respectively；
Key point obtains subelement, for being based on grayscale image, obtains the key point and feature of App entity icon picture respectively
Vector；
Characteristic point obtains subelement, for being based on key point and feature vector, obtains the feature point set being mutually matched respectively
It closes；
Jie Kade is apart from computation subunit, for calculating App entity icon picture based on the set of characteristic points being mutually matched
Between Jie Kade distance, regard calculated Jie Kade distance as picture diversity factor.
A kind of App entity alignment means provided in an embodiment of the present invention, wherein name alignment module is completed in knowledge mapping
App entity to be aligned is obtained after the alignment of App physical name, icon picture obtains module and obtains each App entity to be aligned respectively
Icon picture, whether icon picture comparison module judge the icon picture between each App by comparing the icon picture of each App
Matching determines that the matched each App entity to be aligned of icon picture is that the same App entity completes the alignment of App entity.Due to same
The icon picture of one App differs often very little, therefore can be improved compared with the existing technology according to icon picture to be aligned
The accuracy of App entity alignment.
For device embodiment, since it is substantially similar to the method embodiment, related so being described relatively simple
Place illustrates referring to the part of embodiment of the method.
The embodiment of the invention also provides a kind of electronic equipment, can be server, computer, mobile phone, tablet computer, bluetooth
The electronic products such as bracelet.
Referring to Fig. 6, Fig. 6 is a kind of electronic equipment structure chart that inventive embodiments provide, including processor 601, communication connect
Mouth 602, memory 603 and communication bus 604, wherein processor 601, communication interface 602, memory 603 pass through communication bus
604 complete mutual communication,
Memory 603, for storing computer program；
Processor 601 when for executing the program stored on memory 603, realizes following steps:
According to the name of the App entity in knowledge mapping, App entity is aligned, obtain name it is consistent two or
More than two App entities to be aligned；
Obtain the icon picture of each App entity to be aligned；
The icon picture of each App entity to be aligned is compared, is judged between each App entity to be aligned
Whether icon picture matches；
If the icon picture between each App entity to be aligned matches, the App entity to be aligned is determined as
The same App entity.
Optionally, according to the name of the App entity in knowledge mapping, App entity is aligned, it is consistent to obtain name
The step of two or more App entities to be aligned, may include:
Obtain the name of each App entity in knowledge mapping；
Name similarity in calculation knowledge map between the name of each App entity；
Obtain two or more App entities that name similarity is less than default name similarity threshold；By the name
Word similarity is less than two or more App entities of default name similarity threshold as the consistent App to be aligned of name
Entity.
Optionally, the icon picture of each App entity to be aligned is compared, judge each App entity to be aligned it
Between icon picture the step of whether matching, may include:
Calculate the picture diversity factor between the icon picture of each App entity to be aligned；
Judge whether calculated each picture diversity factor is less than preset threshold；
Determine the icon picture matching that picture diversity factor is less than between each App entity to be aligned of preset threshold.
Optionally, the step of calculating the picture diversity factor between the icon picture of each App entity to be aligned, can wrap
It includes:
Calculate the Hamming distance between the icon picture of each App entity to be aligned；Using calculated Hamming distance as
Picture diversity factor between icon picture.
The step of calculating the Hamming distance between the icon picture of each App entity to be aligned, comprising:
The icon picture size of each App entity to be aligned is converted into preset icon picture normal size；
The grayscale image of the icon picture of each App entity to be aligned is obtained respectively；
Based on grayscale image, the binary map of the icon picture of each App entity to be aligned is obtained respectively；
Binary map is indicated with vector form respectively；
It is indicated based on vector form as a result, the Hamming distance between App entity icon picture is calculated, as picture difference
Degree.
Optionally, the step of calculating the picture diversity factor between the icon picture of each App entity to be aligned, can wrap
It includes:
Calculate the Jie Kade distance between the icon picture of each App entity to be aligned；By calculated Jie Kade distance
As the picture diversity factor between icon picture.
Calculate Jie Kade between the icon picture of each App entity to be aligned apart from the step of, comprising:
The icon picture size of each App entity to be aligned is converted into preset icon picture normal size；
The grayscale image of the icon picture of each App entity to be aligned is obtained respectively；
Based on grayscale image, the key point and feature vector of the icon picture of each App entity to be aligned are obtained respectively；
Based on key point and feature vector, the set of characteristic points being mutually matched are obtained respectively；
Based on the set of characteristic points being mutually matched, the Jie Kade distance of the icon picture of each App entity to be aligned is calculated,
As picture diversity factor.
The communication bus that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect picture standard (Peripheral Component
Interconnect, PCI) bus or extension industry as quasi- structure (Extended Industry Standard
Architecture, EISA) bus etc..The communication bus can be divided into address bus, data/address bus, control bus etc..For just
It is only indicated with a thick line in expression, figure, it is not intended that an only bus or a type of bus.
Communication interface is for the communication between above-mentioned electronic equipment and other equipment.
Memory may include random access memory (Random Access Memory, RAM), also may include non-easy
The property lost memory (Non-Volatile Memory, NVM), for example, at least a magnetic disk storage.Optionally, memory may be used also
To be storage device that at least one is located remotely from aforementioned processor.
Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit,
CPU), network processing unit (Network Processor, NP) etc.；It can also be digital signal processor (Digital Signal
Processing, DSP), it is specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing
It is field programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete
Door or transistor logic, discrete hardware components.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
Each embodiment in this specification is all made of relevant mode and describes, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device reality
For applying example, since it is substantially similar to the method embodiment, so being described relatively simple, related place is referring to embodiment of the method
Part explanation.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention
It is interior.
Claims (9)
1. a kind of App entity alignment schemes, which is characterized in that the described method includes:
According to the name of the App entity in knowledge mapping, App entity is aligned, obtains name consistent two or two
The above App entity to be aligned；
Obtain the icon picture of each App entity to be aligned；
The icon picture of each App entity to be aligned is compared, is judged between each App entity to be aligned
Whether icon picture matches；
If the icon picture between each App entity to be aligned matches, the App entity to be aligned is determined as same
A App entity.
2. the method according to claim 1, wherein the name according to the App entity in knowledge mapping, right
The step of App entity is aligned, and name consistent two or more App entities to be aligned are obtained, comprising:
Obtain the name of each App entity in the knowledge mapping；
Calculate the name similarity in the knowledge mapping between the name of each App entity；
Two or more App entities that the name similarity is less than default name similarity threshold are obtained, and will be described
Two or more App entities that name similarity is less than default name similarity threshold are consistent to be aligned as name
App entity.
3. the method according to claim 1, wherein the icon picture to each App entity to be aligned into
Row compares, and judges the step of whether icon picture between each App entity to be aligned matches, comprising:
Calculate the picture diversity factor between the icon picture of each App entity to be aligned；
Judge whether calculated each picture diversity factor is less than preset threshold；
Determine the icon picture matching that picture diversity factor is less than between each App entity to be aligned of preset threshold.
4. according to the method described in claim 3, it is characterized in that, calculating the icon picture of each App entity to be aligned
Between picture diversity factor the step of, comprising:
Calculate the Hamming distance or Jie Kade distance between the icon picture of each App entity to be aligned；By calculated Hamming
Distance or Jie Kade distance are as the picture diversity factor between icon picture.
5. a kind of App entity alignment means characterized by comprising
Name alignment module is aligned App entity for the name according to the App entity in knowledge mapping, obtains name
Two or more consistent App entities to be aligned；
Icon picture obtains module, for obtaining the icon picture of each App entity to be aligned；
Icon picture comparison module is compared for the icon picture to each App entity to be aligned, judges each institute
Whether the icon picture stated between App entity to be aligned matches；
Determining module, if for the icon picture matching between each App entity to be aligned, by the App to be aligned
Entity is determined as the same App entity.
6. device as claimed in claim 5, which is characterized in that the name alignment module includes:
Name acquiring unit, for obtaining the name of each App entity in the knowledge mapping；
Name similarity calculated, it is similar for calculating name in the knowledge mapping between the name of each App entity
Degree；
Name alignment unit, the similarity for obtaining the name be less than two or two of default name similarity threshold with
Upper App entity；Using the similarity of the name be less than two or more App entities of default name similarity threshold as
The consistent App entity to be aligned of name.
7. device as claimed in claim 5, which is characterized in that the icon picture comparison module includes:
Icon picture diversity factor computing unit, the picture between icon picture for calculating each App entity to be aligned
Diversity factor；
Picture diversity factor judging unit, for judging whether calculated each picture diversity factor is less than preset threshold；
Determination unit, the icon picture being less than between each App entity to be aligned of preset threshold for determining picture diversity factor
Matching.
8. device as claimed in claim 7, which is characterized in that the icon picture diversity factor computing unit is specifically used for: meter
Calculate the Hamming distance or Jie Kade distance between the icon picture of each App entity to be aligned；By calculated Hamming distance or
Jie Kade distance is as the picture diversity factor between icon picture.
9. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor realizes side described in any one of the claims 1-4 when executing described program
Method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811472348.5A CN109726294A (en) | 2018-12-04 | 2018-12-04 | A kind of App entity alignment schemes, device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811472348.5A CN109726294A (en) | 2018-12-04 | 2018-12-04 | A kind of App entity alignment schemes, device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109726294A true CN109726294A (en) | 2019-05-07 |
Family
ID=66294804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811472348.5A Pending CN109726294A (en) | 2018-12-04 | 2018-12-04 | A kind of App entity alignment schemes, device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109726294A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111550015A (en) * | 2020-04-30 | 2020-08-18 | 广东三维家信息科技有限公司 | Automatic adsorption method and device for entities in home design system and electronic equipment |
| CN111563192A (en) * | 2020-04-28 | 2020-08-21 | 腾讯科技（深圳）有限公司 | Entity alignment method and device, electronic equipment and storage medium |
| CN111651972A (en) * | 2020-05-06 | 2020-09-11 | 腾讯科技（深圳）有限公司 | Entity alignment method, device, computer readable medium and electronic equipment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324697A (en) * | 2013-06-07 | 2013-09-25 | 北京掌汇天下科技有限公司 | Method for removing copycatting applications in android application search and based on icon contrast |
| CN104991893A (en) * | 2014-11-06 | 2015-10-21 | 哈尔滨安天科技股份有限公司 | Heuristic detection method and system for self-extracting package and installation package |
| CN106598710A (en) * | 2016-10-28 | 2017-04-26 | 努比亚技术有限公司 | Application management device and method, and mobile terminal |
| CN107368735A (en) * | 2017-07-24 | 2017-11-21 | 广东欧珀移动通信有限公司 | One kind applies installation method, mobile terminal and computer-readable recording medium |
| US20170364238A1 (en) * | 2016-06-17 | 2017-12-21 | Samsung Electronics Co., Ltd. | User input processing method and electronic device performing the same |
| CN108416212A (en) * | 2018-03-01 | 2018-08-17 | 腾讯科技（深圳）有限公司 | Method for identifying application program and device |
-
2018
- 2018-12-04 CN CN201811472348.5A patent/CN109726294A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324697A (en) * | 2013-06-07 | 2013-09-25 | 北京掌汇天下科技有限公司 | Method for removing copycatting applications in android application search and based on icon contrast |
| CN104991893A (en) * | 2014-11-06 | 2015-10-21 | 哈尔滨安天科技股份有限公司 | Heuristic detection method and system for self-extracting package and installation package |
| US20170364238A1 (en) * | 2016-06-17 | 2017-12-21 | Samsung Electronics Co., Ltd. | User input processing method and electronic device performing the same |
| CN106598710A (en) * | 2016-10-28 | 2017-04-26 | 努比亚技术有限公司 | Application management device and method, and mobile terminal |
| CN107368735A (en) * | 2017-07-24 | 2017-11-21 | 广东欧珀移动通信有限公司 | One kind applies installation method, mobile terminal and computer-readable recording medium |
| CN108416212A (en) * | 2018-03-01 | 2018-08-17 | 腾讯科技（深圳）有限公司 | Method for identifying application program and device |
Non-Patent Citations (1)
| Title |
|---|
| 田萍芳: "《面向云出版的语义关键技术》", 30 April 2015 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111563192A (en) * | 2020-04-28 | 2020-08-21 | 腾讯科技（深圳）有限公司 | Entity alignment method and device, electronic equipment and storage medium |
| CN111550015A (en) * | 2020-04-30 | 2020-08-18 | 广东三维家信息科技有限公司 | Automatic adsorption method and device for entities in home design system and electronic equipment |
| CN111651972A (en) * | 2020-05-06 | 2020-09-11 | 腾讯科技（深圳）有限公司 | Entity alignment method, device, computer readable medium and electronic equipment |
| CN111651972B (en) * | 2020-05-06 | 2022-06-17 | 腾讯科技（深圳）有限公司 | Entity alignment method, device, computer readable medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10528618B2 (en) | Image processing and matching | |
| Wang et al. | MARCH: Multiscale-arch-height description for mobile retrieval of leaf images | |
| WO2020125216A1 (en) | Pedestrian re-identification method, device, electronic device and computer-readable storage medium | |
| Mouine et al. | A shape-based approach for leaf classification using multiscaletriangular representation | |
| CN104318259B (en) | A kind of equipment, method and computing device for recognizing Target Photo | |
| KR102678975B1 (en) | Content-based search and retrieval of trademark images | |
| CN104991954B (en) | Shopping search system based on picture search and its implementation | |
| CN111783770B (en) | Image correction method, device and computer readable storage medium | |
| CN109726294A (en) | A kind of App entity alignment schemes, device and electronic equipment | |
| Cao et al. | Similarity based leaf image retrieval using multiscale R-angle description | |
| CN106203242A (en) | A kind of similar image recognition methods and equipment | |
| CN113762280A (en) | Image category identification method, device and medium | |
| US10528844B2 (en) | Method and apparatus for distance measurement | |
| CN112597978B (en) | Fingerprint matching method and device, electronic equipment and storage medium | |
| CN103294676A (en) | Content duplicate detection method of network image based on GIST (generalized search tree) global feature and SIFT (scale-invariant feature transform) local feature | |
| CN110059722A (en) | Checking method, device, equipment and the readable storage medium storing program for executing of seal image | |
| CN109165657A (en) | A kind of image feature detection method and device based on improvement SIFT | |
| Chen et al. | Invariant leaf image recognition with histogram of Gaussian convolution vectors | |
| CN108647264A (en) | A kind of image automatic annotation method and device based on support vector machines | |
| US20200005078A1 (en) | Content aware forensic detection of image manipulations | |
| Bhardwaj et al. | Palette power: Enabling visual search through colors | |
| JP6541226B2 (en) | Information terminal device and program | |
| Chan et al. | F-sort: An alternative for faster geometric verification | |
| CN110059708A (en) | Generate method and device, the storage medium of description | |
| TW202244844A (en) | Image matching 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190507 |
|
| RJ01 | Rejection of invention patent application after publication |