CN115002159B - Community classification and user selection method for sparse group intelligent perception system - Google Patents
Community classification and user selection method for sparse group intelligent perception system Download PDFInfo
- Publication number
- CN115002159B CN115002159B CN202210630549.3A CN202210630549A CN115002159B CN 115002159 B CN115002159 B CN 115002159B CN 202210630549 A CN202210630549 A CN 202210630549A CN 115002159 B CN115002159 B CN 115002159B
- Authority
- CN
- China
- Prior art keywords
- community
- users
- sensing
- perception
- task
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000010187 selection method Methods 0.000 title claims abstract description 8
- 230000008447 perception Effects 0.000 title abstract description 34
- 239000011159 matrix material Substances 0.000 claims abstract description 27
- 238000000034 method Methods 0.000 claims abstract description 16
- 230000007115 recruitment Effects 0.000 claims abstract description 13
- 238000001514 detection method Methods 0.000 abstract description 7
- 230000006870 function Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000013507 mapping Methods 0.000 description 5
- 238000000354 decomposition reaction Methods 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 230000002776 aggregation Effects 0.000 description 3
- 238000004220 aggregation Methods 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000013480 data collection Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/906—Clustering; Classification
-
- 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
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/20—Services signaling; Auxiliary data signalling, i.e. transmitting data via a non-traffic channel
- H04W4/21—Services signaling; Auxiliary data signalling, i.e. transmitting data via a non-traffic channel for social networking applications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Medical Informatics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
In sparse mobile crowd sensing, one key issue is user recruitment, i.e., an organizer wishes to recruit a limited number of users under budget constraints, through which data reasoning is performed on data collected by the users in sub-regions to obtain a complete sensing map. However, due to the variability of user mobility, it is not possible to predict exactly which sub-regions will be covered by the user and are more valuable. Aiming at the problem, the invention provides a community classification and user selection method for a sparse group intelligent perception system. Firstly, considering the social relationship of users to directly recruit high-quality users, adopting a non-negative matrix factorization method of a class depth automatic encoder to carry out community detection, and classifying perceived users; then, based on matching of the position attribute of the perception task and the community center, a recruited community is determined, and then users are recruited from the community according to the limitation of the perception cost; and finally, reconstructing a perception map based on the obtained perception data.
Description
Technical Field
The invention belongs to the field of sparse mobile crowd sensing, and particularly relates to a community classification and user selection method for a sparse crowd sensing system.
Background
In recent years, with the popularity of smart phones equipped with rich sensors, mobile crowd sensing (Mobile Crowd Sensing, MCS) has become a promising paradigm for facilitating urban sensing applications, such as traffic data acquisition, air quality detection, and noise detection. In order to obtain high quality perception results, the strategy adopted in mobile crowd sensing is to recruit enough participants as much as possible to improve regional coverage, but this strategy may lead to high sensing costs, so sparse mobile crowd sensing is proposed as a new solution, only a small part of urban partitions are perceived by the participants, while the data of the rest of the partitions are inferred based on the perceived data, and this MCS paradigm is called sparse MCS due to the sparsity of the perceived sub-regions.
Sparse mobile crowd sensing (spark MCS) is a new paradigm for large scale fine-grained urban monitoring applications that collects sensory data from relatively few areas and infers data for uncovered areas. Sparse mobile population sensing becomes an effective paradigm for reducing the perceived cost of large-scale regional overall state monitoring by recruiting only a small number of users to perceive data from selected sub-regions (i.e., cells). Since sensed data from different cells (sub-regions) of the target sensing area may result in different levels of inferred data quality, cell selection (i.e., which cells of the target area to select to collect sensed data from participants) is a critical issue that will affect the total amount of data that needs to be collected to ensure a particular quality (i.e., data collection cost).
In sparse MCSs, one key issue is user recruitment, i.e., an organizer wants to recruit a limited number of users under budget constraints, who can collect data from several useful sub-regions, which are key to data reasoning, to achieve the highest data accuracy for the sensing service. However, due to the variability of user mobility, it is not possible to predict exactly which sub-regions will be covered by the user. Furthermore, it is difficult to predict which sub-regions are more conducive to complex data reasoning without knowing the true values of the sub-regions. Thus, in sparse MCSs, recruiting the most efficient users is a challenging task. It should be noted that the perceptual data often has more complex and variable spatiotemporal correlations, and it is difficult to measure the importance of each perceptual area, resulting in more complex perceptual area selection problems. In addition, the main focus of the current MCS is on the task itself, and the task itself fuses multidimensional environment information and social attributes, but social information of users is rarely considered, and matching degree of the task and the nodes is ignored. Based on this, the present invention contemplates direct recruitment of high quality users in conjunction with the social relationship of the users.
The social relationship of users is stable and has certain dependency, and certain aggregation phenomenon is formed in the network, so that different communities are formed. As shown in fig. 1, community classification is performed according to the friend relationship of users, the classified users are divided into 4 different communities in a virtual space, users in the same community have the same interests, the users in the communities do not show the characteristic of aggregation in geographic positions, but are distributed in different subareas, when data acquisition is performed, the users in the different subareas in the communities upload the data of the area, and meanwhile, the same user can reach different subareas in different recruitment periods to perform task perception, so that the users after community classification can spread over more subareas and can provide sensing data with higher quality. When task allocation is performed, the recruited community users are decided according to the position attribute of the perceived task. Processing the users in the communities, extracting the position center nodes of the communities, matching the perceived task position attribute with the partitioned community center positions, thereby determining which community is recruited, and selecting and recruiting the users in the selected communities according to the perceived cost limit.
Disclosure of Invention
Aiming at the existing research method and the existing problems, the invention provides a sparse crowd sensing user recruitment method based on non-negative matrix factorization of a class-depth automatic encoder. As can be seen from the social network, the social attributes of the users in reality form a community structure through aggregation and refinement, namely, users in the unified community are closely connected with each other, a certain trust relationship is established between the users, and therefore the social relationship is relatively fixed. In contrast, user social attributes of different communities, including hobbies, range of activities, contact, etc., vary widely. Different users can be divided into different communities according to the behavior rules and the interaction frequency. The sparse MCS user recruitment problem is divided into three phases, a community classification phase, a perceived task matching phase, and perceived map reconstruction as shown in fig. 2.
And (5) community classification. The community classification of the main research of the invention is to divide the perceived users into different communities through social relations. The invention adopts a non-negative matrix factorization method of the class depth automatic encoder to process the friend relations among users to obtain community classification results of the users, thereby being convenient for the next stage of perception task matching.
Perception task matching. The community classification results are screened for users to a certain extent, and users in the same community have the same interests and preferences due to the friend relationship of the users, and have the diversity in task selection. And searching a matched community according to the position characteristic value of the task when the task is perceived, and finally selecting a proper user in the community for recruitment to finish the perception task.
And (5) sensing map reconstruction. Thanks to the completion of the first two phases, the selected users arrive at the subareas in the recruitment period to collect the perception data.
Compared with the prior art, the invention has the beneficial effects that: (1) The method comprises the steps of providing a new idea of sparse mobile crowd sensing user recruitment, carrying out community detection on users in consideration of social relations of the users, reasonably dividing the users into different communities, and recruiting the users based on the classified communities. (2) The perception task is designed to be matched with the community classification center node, communities matched with the perception task are selected, the most suitable user group in the communities is selected for recruitment, a task publisher is helped to obtain high-quality perception data, and the task completion rate is improved. (3) The perceived map is reconstructed through the perceived data provided by the user, so that the number of recruited users is saved, and the perceived cost is reduced.
Drawings
Fig. 1 is a schematic diagram of user recruitment.
Fig. 2 is an overall flow chart of the present invention.
FIG. 3 is a diagram of user partitioning prior to community classification.
Fig. 4 is a view of user division after community classification.
FIG. 5 is a schematic diagram of perceived task matching.
Fig. 6 is a schematic diagram of perceived map reconstruction.
Detailed Description
The invention adopts a non-negative matrix factorization method of a class depth encoder to carry out community classification, as shown in figure 3, users are scattered and disordered before community classification, and after the community detection algorithm is adopted to carry out classification, users with the same interest preference are aggregated in the same class of communities through social relationship processing, as shown in figure 4.
The existing community division algorithm has the problems of single characteristic factors for dividing communities, lack of specific quantification of social relations and the like, and in consideration of complex and diverse topological structures of real world networks, mapping between an original network and a community member space is likely to contain quite complex hierarchical information, which cannot be explained by the traditional shallow NMF-based method. Aiming at the problems, the method adopts a class depth automatic coding non-negative matrix factorization method to classify communities.
NMF-based community detection algorithms typically use a adjacency matrix, which is a social relationship between usersBDecomposing into mapping matrixXAnd community member matrixYThe specific decomposition formula is as follows:B≈XY. The formula indicates that the mapping matrix can be learned from social relationships between usersXAnd community membership matrixYHowever, the community network structure in explicit life is generally very complex, and the mapping matrixXContains a relatively complex hierarchy and therefore uses depth NMF to map the matrixXAnd then performing matrix decomposition to learn the hidden hierarchical structure.
Social network adjacency matrixBIs divided intop+1The following matrices:B≈X 1 X 2 …X p Y p . Wherein,,Y p ∈R k n× ，X i ∈R ri-1 ri× (1≤i≤p)at the same timen=r 0 ≥r 1 ≥…≥r p-1 ≥r p =k. At the same time can alsoBThe concrete decomposition is as follows:Y p-1 ≈X p Y p ，…，Y 2 ≈ X 3 …X p Y p ，Y 1 ≈X 2 …X p Y p . By the above formula, each layer is obtainedY i Further get betterY p Constructing the following objective function, calculatingX 1 ,X 2 ,…,X p ,Y p Equal matrix:min Xi,YpLD =||B-X 1 X 2 …X p Y p || F 2 ，s.t.Y p ≥0，X i ≥0，wherein the method comprises the steps ofi= 1,2,…,p。
Similar to the depth auto encoder, the encoder component attempts to convert the original network into community membership space using implicit low-dimensional hidden attributes captured at the middle layer. Each middle tier accounts for similarities between nodes at different granularity levels. The decoder element is symmetrical to the encoder element. It attempts to reconstruct the original network from the community membership space by means of the hierarchical mapping learned in the encoder component. Unlike conventional NMF-based community detection methods that consider only the loss function of the decoder component, DANMF integrates the encoder component and the decoder component into one unified loss function.
The encoder component converts the network into community membership space. The decoder component reconstructs the network from the community membership space. Objective function of encoder component:min Xi,YpLE =||Y p -X p T …X 2 T X 1 T B|| F 2 ，s.t.Y p ≥0，X i ≥0，wherein the method comprises the steps ofi=1,2,…,p. Unified loss function:min Xi,YpL =L D +L E +wL reg =||B-X 1 X 2 …X p Y p || F 2 +||Y p -X p T …X 2 T X 1 T B|| F 2 +wtr (Y p LY p T )，s.t.Y p ≥0，X i ≥0，wherein the method comprises the steps ofi=1,2,…,p. Wherein,,win order for the parameters to be regularized,Lis a laplace matrix. Matrix of communitiesX i ，Feature matrixY i ，Community membership matrixY p And carrying out iterative updating according to the updating rule until convergence.
A schematic diagram of the community matching based on the position characteristics of the perception task is shown in fig. 5. After users are reasonably divided into different communities, calculating the matching degree of the perception task and the community behavior pattern characteristic value, and selecting the users from the matched communities to recruit according to the result. A perceived task may be defined as a tuple (task content, location), where task content refers to various perceived operations such as air quality monitoring, temperature monitoring, perceived noise, taking a picture, taking a video, etc. The location is then the center coordinates and perceived radius of the perceived task area.
The invention considers matching of the perception task and the community according to the position characteristics of the perception task. The method comprises the steps of firstly determining distribution of a perception task according to the position attribute of the perception task, matching the position attribute of the perception task with the center of a divided community, judging whether the perception task is contained in the radius range of the community, and secondly selecting classified users in the community to carry out the perception task.
Firstly, calculating the distance between users according to the historical track data of the users through longitude and latitude, wherein the distance between users is calculated by the longitude and latitudeu 1 Longitude and latitude coordinates of (a) are(a 1, b 1 )User u 2 Longitude and latitude coordinates of (a) are(a 2 , b 2 )Distance, distancedThe calculation is as follows:haversin(d/R)= haversin(b 2 -b 1 )+cos(b 1 )cos(b 2 )haversin(a 2 -a 1 ). Wherein,,Rfor the earth radius, average 6371km was taken. Selecting one user from the user set divided into the same communities, calculating the distance between the user and the rest of the users, sequencing the obtained distances, screening out the users exceeding the threshold value xi by the comparison result, and not incorporating the node when calculating the community center.
And secondly, calculating a community center point. Considering longitude and latitude coordinates of a user, firstly converting the longitude and latitude coordinates from degrees to radians:a 1 =a 1 ×PI/180，b 1 =b 1 ×PI/180. And converting the longitude and latitude coordinates into Cartesian coordinates of the first position:X 1 =cos(a 1 )×cos(b 1 )，Y 1 = cos(a 1 )×sin(b 1 )，Z 1 =sin(a 1 ). Setting the position weights of the users as1And converting the coordinates of the rest users according to the formula. The total weight of all positions isn，x=(x 1 +x 2 +…+x n )/n，y=(y 1 +y 2 +…+y n )/n，z=(z 1 +z 2 +…+z n )/n. Obtain an averagex, y, z. Will averagex, y, zThe coordinates are converted into longitude and latitude:lon=atan2(y, x)，hyp=sqrt(x×x+y×y)，lat=atan2(z, hyp). Finally, the warp yarn is passed throughThe latitude is converted into degrees:lat=(lat×180)/PI，lon=(lon×180)/PI. The obtained coordinates are the coordinates of the community center point.
Determining a perception task based on the perception data collected as neededh j Location and perceived radiusrAnd determining the perception range as a characteristic value of the perception task. First comparing a plurality of community center points and perception tasksh j Selecting communities with small distances as alternatives according to the distances of the positions; and comparing whether the radius of the perception range is within the community perception range, if the radius is larger than the radius of the community perception range, deleting the community from the candidate communities, and finally, selecting the remained communities as optimal selection, and selecting users in the communities under budget constraint.
The present invention reconstructs a perceived map based on compressed perception, as shown in fig. 6. And recruiting users of partial subareas in sparse mobile crowd sensing, and deducing the data of the rest subareas according to the data provided by the recruited users to obtain an deducing matrix which is close to the real environment matrix. The algorithm in which data inference is performed is typically represented as compressed sensing. The data is inferred by the compressed sensing method, and the following is an introduction to compressed sensing.
Consider partially collected perceptual matricesF’Compressed sensing reconstructs a complete inferred sensing matrix based on low rank attributesF * ：minrank(F * )，s.t., F * 。S=F’. Wherein,,Sthe matrix is collected for the sub-region.S[i, j]Indicating whether the corresponding region in the real environment matrix has been selected for data collection, if at periodic intervalsjIn selecting a sub-regioniTo sense, thenS [i, j]=1Otherwise, it isS[i, j]=0. Because of its non-convexity it is difficult to solve this problem directly, based on singular value decomposition, e.gF * =LR T And compressed sensing theory, an optimization problem is established, and the minimized rank is converted into the minimized rankLAndRthe formula is as follows:minw(||L|| F 2 +||R|| F 2 )+||LR T 。S-F’|| F 2 。wherein,,wa trade-off is made in rank minimization and accuracy fitness. To obtain the optimumF * Iterative estimation using alternating least squaresLAndR(F * = LR T )。
in order to better utilize the spatiotemporal correlation in the perceptual data, an explicit spatiotemporal correlation is introduced into the compressed perceptual, the optimization function is as follows:minw r (||L|| F 2 +||R|| F 2 )+ ||LR T 。S-F’|| F 2 +w s ||G(LR T )|| F 2 +w t ||(LR T ) H T || F 2 . Wherein,,GandHthe spatial and temporal constraint matrices are respectively,Gthe matrix controls the correlation of the different sub-areas at the same time,Hthe matrix controls the data correlation at different moments in the same sensing region,w r, w s, w t to balance the weights of the different elements in the optimization problem.
The above-mentioned embodiments of the present invention are not limited by the above-mentioned embodiments, but any other changes, modifications, substitutions, combinations, and simplifications without departing from the spirit and principles of the present invention should be made in the following claims, and all such modifications, substitutions and combinations are to be regarded as equivalent arrangements.
Claims (4)
1. A community classification and user selection method for a sparse crowd sensing system is characterized by comprising a community classification module, a sensing task matching module and a sensing map reconstruction module.
2. The community classification and user selection method for a sparse crowd sensing system according to claim 1, wherein the sensing users are divided by social relations and aggregated into different communities by a community classification module by using a non-negative matrix factorization method of a class depth automatic encoder.
3. The community classification and user selection method for the sparse crowd sensing system according to claim 1, wherein the sensing task is distributed through a sensing task matching module, a matched community is found through the characteristic value of the sensing task, and finally a proper user in the community is selected for recruitment to complete the sensing task.
4. The community classification and user selection method for a sparse crowd sensing system according to claim 1, wherein a complete sensing map is obtained through a sensing map reconstruction module, sensing data is collected by utilizing the selected sensing users reaching sub-areas in a recruitment period, and the complete sensing map is deduced through a compressed sensing algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210630549.3A CN115002159B (en) | 2022-06-06 | 2022-06-06 | Community classification and user selection method for sparse group intelligent perception system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210630549.3A CN115002159B (en) | 2022-06-06 | 2022-06-06 | Community classification and user selection method for sparse group intelligent perception system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115002159A CN115002159A (en) | 2022-09-02 |
| CN115002159B true CN115002159B (en) | 2023-10-03 |
Family
ID=83033221
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210630549.3A Active CN115002159B (en) | 2022-06-06 | 2022-06-06 | Community classification and user selection method for sparse group intelligent perception system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115002159B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115964626B (en) * | 2022-10-27 | 2024-11-05 | 河南大学 | Community detection method based on dynamic multi-scale feature fusion network |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103533078A (en) * | 2013-10-24 | 2014-01-22 | 无锡赛思汇智科技有限公司 | Method and system for generating map |
| CN107563616A (en) * | 2017-08-17 | 2018-01-09 | 上海大学 | A kind of user task distribution and the quorum-sensing system system and method for incentives strategy |
| CN109347924A (en) * | 2018-09-20 | 2019-02-15 | 西北大学 | A kind of recommended method based on intelligent perception |
| CN110061863A (en) * | 2019-03-25 | 2019-07-26 | 浙江理工大学 | A kind of distributed multi-task intelligent perception method based on fairness in sparse network |
| CN113191023A (en) * | 2021-05-28 | 2021-07-30 | 中国人民解放军国防科技大学 | Crowd-sourcing-aware task allocation and user recruitment model cross-validation method and system |
| CN114745465A (en) * | 2022-03-24 | 2022-07-12 | 马斌斌 | Interactive noise self-prior sensing analysis system for smart phone |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140343984A1 (en) * | 2013-03-14 | 2014-11-20 | University Of Southern California | Spatial crowdsourcing with trustworthy query answering |
-
2022
- 2022-06-06 CN CN202210630549.3A patent/CN115002159B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103533078A (en) * | 2013-10-24 | 2014-01-22 | 无锡赛思汇智科技有限公司 | Method and system for generating map |
| CN107563616A (en) * | 2017-08-17 | 2018-01-09 | 上海大学 | A kind of user task distribution and the quorum-sensing system system and method for incentives strategy |
| CN109347924A (en) * | 2018-09-20 | 2019-02-15 | 西北大学 | A kind of recommended method based on intelligent perception |
| CN110061863A (en) * | 2019-03-25 | 2019-07-26 | 浙江理工大学 | A kind of distributed multi-task intelligent perception method based on fairness in sparse network |
| CN113191023A (en) * | 2021-05-28 | 2021-07-30 | 中国人民解放军国防科技大学 | Crowd-sourcing-aware task allocation and user recruitment model cross-validation method and system |
| CN114745465A (en) * | 2022-03-24 | 2022-07-12 | 马斌斌 | Interactive noise self-prior sensing analysis system for smart phone |
Non-Patent Citations (1)
| Title |
|---|
| 向罗勇 ; 陈文 ; 张陆洋 ; .联合区域热度和社交属性感知的移动群智感知参与者选择机制.电信科学.2020,(第02期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115002159A (en) | 2022-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Liu et al. | Big-data-based intelligent spectrum sensing for heterogeneous spectrum communications in 5G | |
| Hu et al. | Federated region-learning: An edge computing based framework for urban environment sensing | |
| CN111756848B (en) | QoS optimization method based on federal learning and mobile perception under mobile edge environment | |
| Jiang et al. | Deep learning-based edge caching in fog radio access networks | |
| CN103699678B (en) | A kind of hierarchy clustering method based on multistage stratified sampling and system | |
| CN106528804B (en) | A kind of tenant group method based on fuzzy clustering | |
| CN108804551A (en) | It is a kind of to take into account diversity and personalized space point of interest recommendation method | |
| CN106777093A (en) | Skyline inquiry systems based on space time series data stream application | |
| CN115002159B (en) | Community classification and user selection method for sparse group intelligent perception system | |
| Shi et al. | QoS prediction for smart service management and recommendation based on the location of mobile users | |
| CN113472420B (en) | Satellite network cache placement method based on regional user interest perception | |
| CN113344464A (en) | Task and user potential relation mining method facing mobile crowd sensing system | |
| CN112487281A (en) | Stereo garage recommendation method | |
| CN108776863A (en) | One kind being based on the maximized intelligent perception motivational techniques of user base number | |
| CN112508408B (en) | Mapping model construction method of radio resource management index under edge calculation | |
| CN116127190A (en) | Digital earth resource recommendation system and method | |
| Xu et al. | Deep convolutional recurrent model for region recommendation with spatial and temporal contexts | |
| CN114430530B (en) | Space division method, apparatus, device, medium, and program product | |
| Chen et al. | A 100-m gridded population dataset of China’s seventh census using ensemble learning and geospatial big data | |
| Zeng et al. | LGSA: A next POI prediction method by using local and global interest with spatiotemporal awareness | |
| CN114995964A (en) | Combination service reconstruction method, device, equipment and computer readable medium | |
| CN110837540A (en) | Method and system for processing spatial position data | |
| Wang et al. | Deep Learning Based Traffic Prediction in Mobile Network-A Survey | |
| Mo et al. | Time distribution based diversified point of interest recommendation | |
| CN114722904A (en) | Sparse crowd sensing-oriented participant optimization selection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |