CN109639452A - Social modeling training method, device, server and storage medium - Google Patents
Social modeling training method, device, server and storage medium Download PDFInfo
- Publication number
- CN109639452A CN109639452A CN201811283737.3A CN201811283737A CN109639452A CN 109639452 A CN109639452 A CN 109639452A CN 201811283737 A CN201811283737 A CN 201811283737A CN 109639452 A CN109639452 A CN 109639452A
- Authority
- CN
- China
- Prior art keywords
- frame
- sniffer
- wireless exploration
- social
- mobile terminal
- 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
- 238000012549 training Methods 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000003062 neural network model Methods 0.000 claims abstract description 34
- 238000004590 computer program Methods 0.000 claims description 18
- 238000012795 verification Methods 0.000 claims description 10
- 238000011156 evaluation Methods 0.000 claims description 8
- 230000015654 memory Effects 0.000 claims description 7
- 238000012216 screening Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 abstract description 13
- 238000005516 engineering process Methods 0.000 abstract description 2
- 230000005540 biological transmission Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 210000005036 nerve Anatomy 0.000 description 4
- 238000001914 filtration Methods 0.000 description 3
- 210000004209 hair Anatomy 0.000 description 3
- 241000039077 Copula Species 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 239000000523 sample Substances 0.000 description 2
- 241001062472 Stokellia anisodon Species 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W8/00—Network data management
- H04W8/22—Processing or transfer of terminal data, e.g. status or physical capabilities
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L2101/00—Indexing scheme associated with group H04L61/00
- H04L2101/60—Types of network addresses
- H04L2101/618—Details of network addresses
- H04L2101/622—Layer-2 addresses, e.g. medium access control [MAC] addresses
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Databases & Information Systems (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The present invention is applicable in server technology field, provide a kind of social modeling training method based on mobile terminal, device, server and storage medium, this method comprises: obtaining the wireless exploration claim frame that all mobile terminals are sent around Wi-Fi sniffer by Wi-Fi sniffer, according to the corresponding contextual information of MAC Address all in wireless exploration claim frame acquisition wireless exploration claim frame, all MAC Address and corresponding contextual information are input to Skip-Gram neural network model and carry out social networks assertiveness training, according to the social networks term vector of the Skip-Gram neural network model after cosine similarity adjusting training, to obtain the social modeling around Wi-Fi sniffer, and export the social modeling around Wi-Fi sniffer, to logical It crosses the distance between mobile terminal and simplifies the building process of social modeling, and then improve the accuracy rate of social networks building.
Description
Technical field
The invention belongs to server technology field more particularly to a kind of social modeling training sides based on mobile terminal
Method, device, server and storage medium.
Background technique
In recent years, more and more mechanisms focus on the research to human behavior, the behavioral data of the mankind includes abundant
Semantic information, it embodies the variation of society to a certain extent, closes for example, the track data of people can embody the social of people
System, the track data of people are also applicable in the fields such as personalized recommendation, safety.And the track data of people, it can be by people Chang Suishen
The mobile terminal of carrying is monitored.Currently, obtain mobile terminal track data mode have GPS positioning, base station location,
Wi-Fi positioning etc., still, the data that GPS positioning obtains are only limitted to outdoor environment, and base station location is high to the degree of dependence of signal,
The local position error of signal difference is larger, and Wi-Fi positioning applies in general to indoor environment, and can get use without intrusive
The location information at family, but influence of the setting accuracy vulnerable to indoor environment.Therefore, it is obtained by the location information on mobile terminal
The track data for taking mobile terminal, re-maps out the social networks of each mobile terminal user, and the social of mobile terminal user closes
The building process of system is cumbersome and accuracy rate is lower.
Summary of the invention
The social modeling training method that the purpose of the present invention is to provide a kind of based on mobile terminal, device, service
Device and storage medium, it is intended to solve that a kind of effective social modeling training method can not be provided due to the prior art, lead
The problem that process is cumbersome, accuracy rate is lower when the social networks of mobile terminal user being caused to construct.
On the one hand, the present invention provides a kind of social modeling training method based on mobile terminal, the method packet
Include following step:
The wireless exploration request that all mobile terminals are sent around the Wi-Fi sniffer is obtained by Wi-Fi sniffer
Frame, the wireless exploration claim frame include MAC Address, sending time and signal strength;
According to the corresponding context of MAC Address all in the wireless exploration claim frame acquisition wireless exploration claim frame
Information, the contextual information include the distance between MAC Address mobile terminal corresponding with other MAC Address;
By all MAC Address and the corresponding contextual information be input to Skip-Gram neural network model into
Row social networks assertiveness training;
The social of the Skip-Gram neural network model closes after adjusting social networks assertiveness training according to cosine similarity
Copula vector to obtain the social modeling around the Wi-Fi sniffer, and exports around the Wi-Fi sniffer
Social modeling.
On the other hand, the present invention provides a kind of the social modeling training device based on mobile terminal, described device
Include:
Claim frame acquiring unit, for obtaining all mobile terminals around the Wi-Fi sniffer by Wi-Fi sniffer
The wireless exploration claim frame of transmission;
Information acquisition unit owns for being obtained in the wireless exploration claim frame according to the wireless exploration claim frame
The corresponding contextual information of MAC Address;
Model training unit, for all MAC Address and the corresponding contextual information to be input to Skip-
Gram neural network model carries out social networks assertiveness training;And
Output unit is adjusted, it is refreshing for the Skip-Gram after adjusting social networks assertiveness training according to cosine similarity
Social networks term vector through network model, to obtain the social modeling around the Wi-Fi sniffer, and described in output
Social modeling around Wi-Fi sniffer.
On the other hand, the present invention also provides a kind of server, including memory, processor and it is stored in the storage
In device and the computer program that can run on the processor, the processor are realized as above when executing the computer program
The step of stating the social modeling training method based on mobile terminal.
On the other hand, the present invention also provides a kind of computer readable storage medium, the computer readable storage mediums
It is stored with computer program, such as the above-mentioned social networks based on mobile terminal are realized when the computer program is executed by processor
The step of model training method.
The present invention obtains the wireless exploration that all mobile terminals are sent around Wi-Fi sniffer by Wi-Fi sniffer and asks
Frame is sought, according to the corresponding contextual information of MAC Address all in wireless exploration claim frame acquisition wireless exploration claim frame, will be owned
MAC Address and corresponding contextual information are input to Skip-Gram neural network model and carry out social networks assertiveness training, according to
The social networks term vector of Skip-Gram neural network model after cosine similarity adjusting training, to obtain Wi-Fi sniffer
The social modeling of surrounding, and the social modeling around Wi-Fi sniffer is exported, thus by between mobile terminal
Distance simplifies the building process of social modeling, and then improves the accuracy rate of social networks building.
Detailed description of the invention
Fig. 1 is the implementation process for the social modeling training method based on mobile terminal that the embodiment of the present invention one provides
Figure;
Fig. 2 is the structural representation of the social modeling training device provided by Embodiment 2 of the present invention based on mobile terminal
Figure;
Fig. 3 is the structural representation for the social modeling training device based on mobile terminal that the embodiment of the present invention three provides
Figure;And
Fig. 4 is a kind of structural schematic diagram for server that the embodiment of the present invention four provides.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Specific implementation of the invention is described in detail below in conjunction with specific embodiment:
Embodiment one:
Fig. 1 shows the realization of the social modeling training method based on mobile terminal of the offer of the embodiment of the present invention one
Process, for ease of description, only parts related to embodiments of the present invention are shown, and details are as follows:
In step s101, it is sent by all mobile terminals around Wi-Fi sniffer acquisition Wi-Fi sniffer wireless
Probe request.
The embodiment of the present invention is suitable for server, which is used to construct the social networks between mobile terminal user,
The server is also associated with 3 or 3 or more Wi-Fi (WIreless-FIdelity, Wireless Fidelity) sniffer, for obtaining
It is taking that mobile terminal sends, for detect Wi-Fi Hotspot and request access Wi-Fi Hotspot data (for the ease of subsequent descriptions,
The data are known as wireless exploration claim frame).In embodiments of the present invention, it first passes through Wi-Fi sniffer and obtains Wi-Fi sniffer
The wireless exploration claim frame that all mobile terminals of surrounding are sent, wherein wireless exploration claim frame includes the MAC of mobile terminal
(Media Access Control, media access control) address, the sending time of wireless exploration claim frame and wireless exploration
The signal strength of claim frame.
In step s 102, corresponding according to all MAC Address in wireless exploration claim frame acquisition wireless exploration claim frame
Contextual information.
In embodiments of the present invention, all MAC Address pair in wireless exploration claim frame are obtained according to wireless exploration claim frame
The contextual information answered, wherein contextual information includes the distance between MAC Address dynamic terminal corresponding with other MAC Address.
As illustratively, in the wireless exploration claim frame of acquisition, MAC Address has A, B, C respectively, then the contextual information of MAC Address A is
(B, dAB), (C, dAC), and so on, wherein dAB、dACRespectively MAC Address A, B corresponds to the distance between mobile terminal and MAC
A, C correspond to the distance between mobile terminal for address.
Preferably, corresponding according to all MAC Address in wireless exploration claim frame acquisition wireless exploration claim frame in acquisition
When contextual information, first movement terminal and the second mobile terminal in preset time period are obtained according to wireless exploration claim frame and issued
Wireless exploration claim frame real-time signal strength, it is strong by the real-time signal strength and preset signal of wireless exploration claim frame
The signal strength average value that Mean Value Formulas calculates wireless exploration claim frame within a preset period of time is spent, it is average according to signal strength
Value obtains the first relative position of first movement terminal within a preset period of time and Wi-Fi sniffer, the second mobile terminal respectively
With the second relative position of Wi-Fi sniffer, finally obtained according to the first relative position and the second relative position in preset time
The second mobile terminal is at a distance from first movement terminal in section, to realize the spacing to two mobile terminals under off-line state
From accurately calculate, and then increase the scope of application of model training.As illustratively, n is got in n Wi-Fi sniffer
It is a at a distance from mobile terminal after, it may be determined that the relative position of the mobile terminal and n Wi-Fi sniffer determines the shifting
The coordinate of dynamic terminal after the coordinate that other mobile terminals have similarly been determined, can be calculated according to the coordinate of two mobile terminals
The distance between two mobile terminals out, wherein n is not less than 3.
When obtaining the signal strength average value of wireless exploration claim frame, the second mobile mobile terminal is considered as in window
The region that size is W is mobile, and S is the signal strength average value of wireless exploration claim frame when moving in the window, and s is wireless visits
Therefore the signal strength real value for surveying claim frame further, passes through formulaCalculate wireless exploration claim frame
Signal strength average value, thus improve positioning the second mobile terminal accuracy.The window size W of setting should not mistake
Greatly, also unsuitable too small, window is excessive, will affect real-time, and window is too small, then influences stability, it is preferable that sets window size
20 are set to, to improve the accuracy of signal strength average value.
Preferably, corresponding up and down according to all MAC Address in wireless exploration claim frame acquisition wireless exploration claim frame
Before literary information, according in wireless exploration claim frame sending time and MAC Address content to wireless exploration claim frame carry out
Screening, to reduce invalid training data, accelerates subsequent model training process, and then improve model training efficiency.
Further, when being screened to wireless exploration claim frame, first according to preceding 3 bytes of MAC Address to all MAC Address into
Row filtering, thus filter out the MAC Address and corresponding contextual information of non-moving cell phone type mobile terminal, i.e. the MAC
All wireless exploration claim frames that location is sent, the sending time and transmission times further according to wireless exploration claim frame are to all second
The MAC Address of mobile terminal carries out address filtering, so that filtering out transmission times in preset period of time is lower than preset times threshold value
The wireless exploration claim frame that mobile terminal is sent.
In step s 103, all MAC Address and corresponding contextual information are input to Skip-Gram neural network mould
Type carries out social networks assertiveness training.
In embodiments of the present invention, word centered on all MAC Address and corresponding contextual information are input to Skip-
Gram neural network model carries out social networks assertiveness training, so that the map contextual information of all mobile terminals be arrived
It in Skip-Gram neural network model, is indicated in the form of higher-dimension term vector, which is Skip-Gram nerve
The higher-dimension term vector of higher dimensional space in network model carries out in specific operation process to Skip-Gram neural network model
When training, the number that each MAC Address in all contextual informations sends wireless exploration claim frame is counted, transmission times is less
Negative sample of the mobile terminal as Skip-Gram neural network model, to accelerate Skip-Gram by negative sampling policy
The training speed of neural network model.
In step S104, closed according to the social activity of the Skip-Gram neural network model after cosine similarity adjusting training
Copula vector to obtain the social modeling around Wi-Fi sniffer, and exports the social networks around Wi-Fi sniffer
Model.
In embodiments of the present invention, after carrying out social networks assertiveness training to Skip-Gram neural network model, on
Context information is mapped in Skip-Gram neural network model, is indicated in the form of higher-dimension term vector, for convenient for subsequent descriptions,
The contextual information that will be mapped in Skip-Gram neural network model is known as social networks term vector, according to cosine similarity
The social networks term vector of Skip-Gram neural network model after adjusting assertiveness training, to obtain around Wi-Fi sniffer
Social modeling, and export the social modeling.
Preferably, in the social networks word according to the Skip-Gram neural network model after cosine similarity adjusting training
After vector, before exporting the social modeling around Wi-Fi sniffer, all mobile terminals in contextual information
The distance between using preset clustering algorithm to all mobile terminals carry out Density Clustering, to obtain around Wi-Fi sniffer
Social modeling verification result, verification result is evaluated by preset evaluation algorithms, to determine verification result
It is whether accurate, to improve the accuracy of building social modeling.Further, which is DBSCAN
(Density-Based Spatial Clustering ofApplications withNoise) density clustering algorithm, thus
Enhance the applicability of social networks verifying.Still further, the evaluation algorithms are F-Measure algorithm, to improve pair
The accuracy of verification result evaluation.
Preferably, in the social networks word according to the Skip-Gram neural network model after cosine similarity adjusting training
After vector, before exporting the social modeling around Wi-Fi sniffer, by preset dimension-reduction algorithm by social networks word
DUAL PROBLEMS OF VECTOR MAPPING is to two-dimensional space, consequently facilitating user carries out visual inspection to social networks, further improves the social pass of building
The accuracy of system.
In embodiments of the present invention, obtain what all mobile terminals around Wi-Fi sniffer were sent by Wi-Fi sniffer
Wireless exploration claim frame, according to the corresponding context of MAC Address all in wireless exploration claim frame acquisition wireless exploration claim frame
All MAC Address and corresponding contextual information are input to Skip-Gram neural network model and carry out social networks table by information
Up to training, according to the social networks term vector of the Skip-Gram neural network model after cosine similarity adjusting training, to obtain
Social modeling around Wi-Fi sniffer, and the social modeling around Wi-Fi sniffer is exported, to pass through shifting
The distance between dynamic terminal simplifies the building process of social modeling, and then improves the accuracy rate of social networks building.
Embodiment two:
The knot of Fig. 2 shows the provided by Embodiment 2 of the present invention social modeling training device based on mobile terminal
Structure, for ease of description, only parts related to embodiments of the present invention are shown, including:
Claim frame acquiring unit 21, for obtaining all mobile terminal hairs around Wi-Fi sniffer by Wi-Fi sniffer
The wireless exploration claim frame sent;
Information acquisition unit 22, for obtaining all MAC Address in wireless exploration claim frame according to wireless exploration claim frame
Corresponding contextual information;
Model training unit 23, for all MAC Address and corresponding contextual information to be input to Skip-Gram nerve
Network model carries out social networks assertiveness training;And
Output unit 24 is adjusted, for according to the Skip-Gram neural network model after cosine similarity adjusting training
Social networks term vector to obtain the social modeling around Wi-Fi sniffer, and exports the society around Wi-Fi sniffer
Hand over relational model.
In embodiments of the present invention, obtain what all mobile terminals around Wi-Fi sniffer were sent by Wi-Fi sniffer
Wireless exploration claim frame, according to the corresponding context of MAC Address all in wireless exploration claim frame acquisition wireless exploration claim frame
All MAC Address and corresponding contextual information are input to Skip-Gram neural network model and carry out social networks table by information
Up to training, according to the social networks term vector of the Skip-Gram neural network model after cosine similarity adjusting training, to obtain
Social modeling around Wi-Fi sniffer, and the social modeling around Wi-Fi sniffer is exported, to pass through shifting
The distance between dynamic terminal simplifies the building process of social modeling, and then improves the accuracy rate of social networks building.
In embodiments of the present invention, each unit of the social modeling training device based on mobile terminal can be by corresponding
Hardware or software unit realize that each unit can be independent soft and hardware unit, also can integrate as a soft and hardware unit,
Herein not to limit the present invention.The specific embodiment of each unit can refer to the description of embodiment one, and details are not described herein.
Embodiment three:
Fig. 3 shows the knot of the social modeling training device based on mobile terminal of the offer of the embodiment of the present invention three
Structure, for ease of description, only parts related to embodiments of the present invention are shown, including:
Claim frame acquiring unit 31, for obtaining all mobile terminal hairs around Wi-Fi sniffer by Wi-Fi sniffer
The wireless exploration claim frame sent;
Information acquisition unit 32, for obtaining all MAC Address in wireless exploration claim frame according to wireless exploration claim frame
Corresponding contextual information;
Model training unit 33, for all MAC Address and corresponding contextual information to be input to Skip-Gram nerve
Network model carries out social networks assertiveness training;And
Output unit 34 is adjusted, for according to the Skip-Gram neural network model after cosine similarity adjusting training
Social networks term vector to obtain the social modeling around Wi-Fi sniffer, and exports the society around Wi-Fi sniffer
Hand over relational model.
Wherein, output unit 34 is adjusted, comprising:
Authentication unit 341 is clustered, is used for the distance between all mobile terminals in contextual information default
Clustering algorithm to all mobile terminals carry out Density Clustering, to obtain testing for the social modeling around Wi-Fi sniffer
Demonstrate,prove result;And
Evaluation unit 342 is verified, for evaluating by preset evaluation algorithms verification result, to determine verifying knot
Whether fruit is accurate.
Information acquisition unit 32, comprising:
Signal acquisition subelement 321, for obtaining first movement terminal in preset time period according to wireless exploration claim frame
The real-time signal strength of the wireless exploration claim frame issued with the second mobile terminal;
Signature computation unit 322, for passing through the real-time signal strength of wireless exploration claim frame and preset signal strength
Mean Value Formulas calculates the signal strength average value of wireless exploration claim frame within a preset period of time;
Position acquisition unit 323, for obtaining first movement within a preset period of time respectively according to signal strength average value
Terminal and the first relative position of Wi-Fi sniffer, the second relative position of the second mobile terminal and Wi-Fi sniffer;And
Distance acquiring unit 324, for being obtained within a preset period of time according to the first relative position and the second relative position
Second mobile terminal is at a distance from first movement terminal.
In embodiments of the present invention, obtain what all mobile terminals around Wi-Fi sniffer were sent by Wi-Fi sniffer
Wireless exploration claim frame, according to the corresponding context of MAC Address all in wireless exploration claim frame acquisition wireless exploration claim frame
All MAC Address and corresponding contextual information are input to Skip-Gram neural network model and carry out social networks table by information
Up to training, according to the social networks term vector of the Skip-Gram neural network model after cosine similarity adjusting training, and exchange
Social modeling after whole is verified, is evaluated, and to obtain the social modeling around Wi-Fi sniffer, and exports Wi-
Social modeling around Fi sniffer, to simplify the building of social modeling by the distance between mobile terminal
Process, and then improve the accuracy rate of social networks building.
In embodiments of the present invention, each unit of the social modeling training device based on mobile terminal can be by corresponding
Hardware or software unit realize that each unit can be independent soft and hardware unit, also can integrate as a soft and hardware unit,
Herein not to limit the present invention.The specific embodiment of each unit can refer to the description of embodiment one, and details are not described herein.
Example IV:
The structure that Fig. 4 shows the server of the offer of the embodiment of the present invention four illustrates only and this hair for ease of description
The relevant part of bright embodiment, including:
The server 4 of the embodiment of the present invention includes processor 41, memory 42 and is stored in memory 42 and can be
The computer program 43 run on processor 41.The processor 41 is realized above-mentioned based on mobile terminal when executing computer program 43
Social modeling training method embodiment in step, such as step S101 to S104 shown in FIG. 1.Alternatively, processor
It is realized when 41 execution computer program 43 each in above-mentioned each social modeling training device embodiment based on mobile terminal
The function of unit, such as the function of unit 31 to 34 shown in unit 21 to 24 and Fig. 3 shown in Fig. 2.
In embodiments of the present invention, when which executes computer program, Wi-Fi is obtained by Wi-Fi sniffer and is smelt
The wireless exploration claim frame that all mobile terminals are sent around device is visited, wireless exploration claim frame is obtained according to wireless exploration claim frame
In the corresponding contextual information of all MAC Address, all MAC Address and corresponding contextual information are input to Skip-Gram
Neural network model carries out social networks assertiveness training, according to the Skip-Gram neural network after cosine similarity adjusting training
The social networks term vector of model to obtain the social modeling around Wi-Fi sniffer, and exports Wi-Fi sniffer week
The social modeling enclosed, to simplify the building process of social modeling by the distance between mobile terminal, in turn
Improve the accuracy rate of social networks building.
The processor realizes that the above-mentioned social modeling training method based on mobile terminal is real when executing computer program
Applying the step in example can refer to the description of embodiment one, and details are not described herein.
Embodiment five:
In embodiments of the present invention, a kind of computer readable storage medium is provided, which deposits
Computer program is contained, which realizes above-mentioned each social networks mould based on mobile terminal when being executed by processor
Step in type training method embodiment, for example, step S101 to S104 shown in FIG. 1.Alternatively, the computer program is processed
Device realizes the function of each unit in above-mentioned each social modeling training device embodiment based on mobile terminal, example when executing
The function of unit 31 to 34 shown in unit 21 to 24 and Fig. 3 as shown in Figure 2.
In embodiments of the present invention, after computer program is executed by processor, Wi-Fi is obtained by Wi-Fi sniffer
The wireless exploration claim frame that all mobile terminals are sent around sniffer obtains wireless exploration request according to wireless exploration claim frame
All MAC Address and corresponding contextual information are input to Skip- by the corresponding contextual information of all MAC Address in frame
Gram neural network model carries out social networks assertiveness training, according to the Skip-Gram nerve after cosine similarity adjusting training
The social networks term vector of network model to obtain the social modeling around Wi-Fi sniffer, and exports Wi-Fi sniff
Social modeling around device, so that the building process of social modeling is simplified by the distance between mobile terminal,
And then improve the accuracy rate of social networks building.
Realize that the above-mentioned social modeling training method based on mobile terminal is real when processor executes computer program
Applying the step in example can refer to the description of embodiment one, and details are not described herein.
The computer readable storage medium of the embodiment of the present invention may include can carry computer program code any
Entity or device, storage medium, for example, the memories such as ROM/RAM, disk, CD, flash memory.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (10)
1. a kind of social modeling training method based on mobile terminal, which is characterized in that the described method includes:
The wireless exploration claim frame that all mobile terminals are sent around the Wi-Fi sniffer, institute are obtained by Wi-Fi sniffer
Stating wireless exploration claim frame includes MAC Address, sending time and signal strength;
According to the corresponding context letter of MAC Address all in the wireless exploration claim frame acquisition wireless exploration claim frame
Breath, the contextual information includes the distance between MAC Address mobile terminal corresponding with other MAC Address;
All MAC Address and the corresponding contextual information are input to Skip-Gram neural network model and carry out society
Hand over relationship expression training;
According to the social networks term vector of the Skip-Gram neural network model after cosine similarity adjusting training, to obtain
Social modeling to around the Wi-Fi sniffer, and export the social modeling around the Wi-Fi sniffer.
2. the method as described in claim 1, which is characterized in that obtain the wireless exploration according to the wireless exploration claim frame
In claim frame the step of all MAC Address corresponding contextual information before, the method also includes:
According in the wireless exploration claim frame sending time and MAC Address content the wireless exploration claim frame is carried out
Screening.
3. the method as described in claim 1, which is characterized in that obtain the wireless exploration according to the wireless exploration claim frame
In claim frame the step of all MAC Address corresponding contextual information, comprising:
The institute that first movement terminal and the second mobile terminal issue in preset time period is obtained according to the wireless exploration claim frame
State the real-time signal strength of wireless exploration claim frame;
It is calculated by the real-time signal strength of the wireless exploration claim frame and preset signal strength Mean Value Formulas described
The signal strength average value of the wireless exploration claim frame in preset time period;
The first movement terminal within a preset period of time is obtained respectively according to the signal strength average value to smell with the Wi-Fi
Visit the first relative position of device, the second relative position of second mobile terminal and the Wi-Fi sniffer;
The second mobile terminal within a preset period of time and the are obtained according to first relative position and second relative position
The distance of one mobile terminal.
4. the method as described in claim 1, which is characterized in that according to the Skip- after cosine similarity adjusting training
After the step of social networks term vector of Gram neural network model, the social networks around the Wi-Fi sniffer are exported
Before the step of model, the method also includes:
Owned using preset clustering algorithm to described according to the distance between all mobile terminals in the contextual information
Mobile terminal carries out Density Clustering, to obtain the verification result of the social modeling around the Wi-Fi sniffer;
The verification result is evaluated by preset evaluation algorithms, it is whether accurate with the determination verification result.
5. the method as described in claim 1, which is characterized in that according to the Skip- after cosine similarity adjusting training
After the step of social networks term vector of Gram neural network model, the social networks around the Wi-Fi sniffer are exported
Before the step of model, the method also includes:
The social networks term vector is mapped to two-dimensional space by preset dimension-reduction algorithm.
6. a kind of social modeling training device based on mobile terminal, which is characterized in that described device includes:
Claim frame acquiring unit is sent for obtaining all mobile terminals around the Wi-Fi sniffer by Wi-Fi sniffer
Wireless exploration claim frame;
Information acquisition unit, for according to all MAC in the wireless exploration claim frame acquisition wireless exploration claim frame
The corresponding contextual information in location;
Model training unit, for all MAC Address and the corresponding contextual information to be input to Skip-Gram mind
Social networks assertiveness training is carried out through network model;And
Output unit is adjusted, for the society according to the Skip-Gram neural network model after cosine similarity adjusting training
Friendship relationship term vector to obtain the social modeling around the Wi-Fi sniffer, and exports the Wi-Fi sniffer week
The social modeling enclosed.
7. device as claimed in claim 6, which is characterized in that the information acquisition unit includes:
Signal acquisition subelement, for obtaining first movement terminal and the in preset time period according to the wireless exploration claim frame
The real-time signal strength for the wireless exploration claim frame that two mobile terminals issue;
Signature computation unit, for by the wireless exploration claim frame real-time signal strength and preset signal strength it is average
It is worth the signal strength average value that formula calculates the wireless exploration claim frame in the preset time period;
Position acquisition unit, for obtaining the first movement within a preset period of time respectively according to the signal strength average value
Terminal and the first relative position of the Wi-Fi sniffer, the second phase of second mobile terminal and the Wi-Fi sniffer
To position;And
Distance acquiring unit, for being obtained within a preset period of time according to first relative position and second relative position
Second mobile terminal is at a distance from first movement terminal.
8. device as claimed in claim 6, which is characterized in that the adjustment output unit includes:
Authentication unit is clustered, for gathering according to the distance between all mobile terminals in the contextual information using preset
Class algorithm carries out Density Clustering to all mobile terminals, to obtain the social modeling around the Wi-Fi sniffer
Verification result;And
Evaluation unit is verified, for evaluating by preset evaluation algorithms the verification result, with the determination verifying
As a result whether accurate.
9. a kind of server, including memory, processor and storage can transport in the memory and on the processor
Capable computer program, which is characterized in that the processor realizes such as claim 1 to 5 when executing the computer program
The step of the method.
10. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In realization is such as the step of claim 1 to 5 the method when the computer program is executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811283737.3A CN109639452A (en) | 2018-10-31 | 2018-10-31 | Social modeling training method, device, server and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811283737.3A CN109639452A (en) | 2018-10-31 | 2018-10-31 | Social modeling training method, device, server and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109639452A true CN109639452A (en) | 2019-04-16 |
Family
ID=66066958
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811283737.3A Pending CN109639452A (en) | 2018-10-31 | 2018-10-31 | Social modeling training method, device, server and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109639452A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112508408A (en) * | 2020-12-10 | 2021-03-16 | 北京科技大学 | Mapping model construction method of wireless resource management index under edge calculation |
CN114339588A (en) * | 2020-09-25 | 2022-04-12 | 纬创资通股份有限公司 | Social distance judgment system and social distance judgment method |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105117947A (en) * | 2015-09-07 | 2015-12-02 | 丹阳伦图电子技术有限公司 | Wi-Fi technology-based ESL (electronic shelf label) system |
CN107111859A (en) * | 2014-10-22 | 2017-08-29 | 脸谱公司 | For the social activity scoring of network element |
CN107239444A (en) * | 2017-05-26 | 2017-10-10 | 华中科技大学 | A kind of term vector training method and system for merging part of speech and positional information |
US20180060755A1 (en) * | 2016-09-01 | 2018-03-01 | Facebook, Inc. | Systems and methods for recommending pages |
CN107818164A (en) * | 2017-11-02 | 2018-03-20 | 东北师范大学 | A kind of intelligent answer method and its system |
CN107832306A (en) * | 2017-11-28 | 2018-03-23 | 武汉大学 | A kind of similar entities method for digging based on Doc2vec |
CN107944498A (en) * | 2017-12-06 | 2018-04-20 | 河海大学 | A kind of indoor people's swarm clustering method based on multi-tag |
CN108521632A (en) * | 2018-04-04 | 2018-09-11 | 中山大学 | Real-time social networks based on Wi-Fi find system and implementation method |
CN108628834A (en) * | 2018-05-14 | 2018-10-09 | 国家计算机网络与信息安全管理中心 | A kind of word lists dendrography learning method based on syntax dependence |
-
2018
- 2018-10-31 CN CN201811283737.3A patent/CN109639452A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107111859A (en) * | 2014-10-22 | 2017-08-29 | 脸谱公司 | For the social activity scoring of network element |
CN105117947A (en) * | 2015-09-07 | 2015-12-02 | 丹阳伦图电子技术有限公司 | Wi-Fi technology-based ESL (electronic shelf label) system |
US20180060755A1 (en) * | 2016-09-01 | 2018-03-01 | Facebook, Inc. | Systems and methods for recommending pages |
CN107239444A (en) * | 2017-05-26 | 2017-10-10 | 华中科技大学 | A kind of term vector training method and system for merging part of speech and positional information |
CN107818164A (en) * | 2017-11-02 | 2018-03-20 | 东北师范大学 | A kind of intelligent answer method and its system |
CN107832306A (en) * | 2017-11-28 | 2018-03-23 | 武汉大学 | A kind of similar entities method for digging based on Doc2vec |
CN107944498A (en) * | 2017-12-06 | 2018-04-20 | 河海大学 | A kind of indoor people's swarm clustering method based on multi-tag |
CN108521632A (en) * | 2018-04-04 | 2018-09-11 | 中山大学 | Real-time social networks based on Wi-Fi find system and implementation method |
CN108628834A (en) * | 2018-05-14 | 2018-10-09 | 国家计算机网络与信息安全管理中心 | A kind of word lists dendrography learning method based on syntax dependence |
Non-Patent Citations (1)
Title |
---|
罗成文 等: "SODAR:non-obtrusive off-line social structure reconstruction through passive wireless sensing", 《IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114339588A (en) * | 2020-09-25 | 2022-04-12 | 纬创资通股份有限公司 | Social distance judgment system and social distance judgment method |
CN114339588B (en) * | 2020-09-25 | 2023-11-10 | 纬创资通股份有限公司 | Social distance judging system and social distance judging method |
CN112508408A (en) * | 2020-12-10 | 2021-03-16 | 北京科技大学 | Mapping model construction method of wireless resource management index under edge calculation |
CN112508408B (en) * | 2020-12-10 | 2024-01-05 | 北京科技大学 | Mapping model construction method of radio resource management index under edge calculation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109635872A (en) | Personal identification method, electronic equipment and computer program product | |
CN110097066B (en) | User classification method and device and electronic equipment | |
CN106875422B (en) | Face tracking method and device | |
CN108460101A (en) | Point of interest of the facing position social networks based on geographical location regularization recommends method | |
CN105830081A (en) | Methods and systems of generating application-specific models for the targeted protection of vital applications | |
CN108229268A (en) | Expression Recognition and convolutional neural networks model training method, device and electronic equipment | |
CN111401447B (en) | Artificial intelligence-based flow cheating identification method and device and electronic equipment | |
CN110022531B (en) | Localized differential privacy urban garbage data report and privacy calculation method | |
CN109977651A (en) | Man-machine recognition methods, device and electronic equipment based on sliding trace | |
CN111460294A (en) | Message pushing method and device, computer equipment and storage medium | |
CN110300127A (en) | A kind of network inbreak detection method based on deep learning, device and equipment | |
CN107818301A (en) | Update the method, apparatus and electronic equipment of biometric templates | |
CN110390673A (en) | Cigarette automatic testing method based on deep learning under a kind of monitoring scene | |
CN107622197A (en) | Device identification method and device, weighing computation method and device for equipment identification | |
CN110245714A (en) | Image-recognizing method, device and electronic equipment | |
CN109639452A (en) | Social modeling training method, device, server and storage medium | |
CN109829477A (en) | More attribute physical layer authentication methods, device and server based on heuristic cluster | |
CN107169769A (en) | The brush amount recognition methods of application program, device | |
CN111695046A (en) | User portrait inference method and device based on spatio-temporal mobile data representation learning | |
CN110008402A (en) | A kind of point of interest recommended method of the decentralization matrix decomposition based on social networks | |
Sundsøy | Can mobile usage predict illiteracy in a developing country? | |
CN111813910A (en) | Method, system, terminal device and computer storage medium for updating customer service problem | |
CN105338186A (en) | Context awareness-based Android mobile terminal power management method | |
CN105721467B (en) | Social networks Sybil crowd surveillance method | |
CN107516112A (en) | Object type recognition methods, device, equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190416 |
|
RJ01 | Rejection of invention patent application after publication |