Crowd cluster data in the USA for analysis of human response to COVID-19 events and policies

Virtual Circular Geofences for Points and Regions of Interests with Spatial Context

Location Based Service (LBS) is gaining popularity on smart phones. One fundamental LBS is range search, which returns all Point of Interests (POIs) within a user-specified range. However, people also leave their location privacy at risks when using LBS like range search. How can a user invoke such service without revealing his location is an interesting, yet challenging problem to solve. Most existing approaches blur a user's location into a cloaked region, so that LBS cannot figure out the exact location of the requesting user. However, this would make the returning results inaccurate, containing some out-of-range POIs. To this end, we propose PAPERS, a new method to provide location privacy for users of range search. PAPERS leverage homomorphic encryption to let the user encrypt her location, and the LBS server can compute distances on ciphertext. In this way, the returning results by LBS are exactly the POIs within the specified range, while LBS learns nothing about user's real location. We implement a prototype of PAPERS, and evaluate it with real POI set of a large-scale production LBS. Experimental results show that PAPERS can achieve the goal of privacy protection, with reasonable overhead in response time and communication cost.

Studying carbon emissions (CE) at different administrative scales will help facilitate crafting tailored emission reduction policies for China’s regions, which is vital for achieving the dual carbon goals. However, previous studies focused on a single administrative scale, lacking multiscale research. This paper combined energy consumption data with nighttime light data and adopted a spatial autocorrelation, variation coefficient (VC), and decoupling model to study the spatiotemporal dynamics and decoupling effect of CE at the three administrative scales of provinces, prefecture-level cities, and counties in China from 2000 to 2020. The results were as follows: (1) The VC of CE showed different trends at different scales, with its coefficient size successively ranked at the county, prefecture, and province levels. (2) CE at different scales showed positive spatial autocorrelation and the significance was strongest at the county level. (3) The decoupling trend between CE and economic growth has generally shown a positive development across different spatial scales. The average elasticity decoupling index at the provincial, prefectural, and county levels has decreased overall, from 0.88, 1.82, and 3.74 to 0.19, 1.36, and 3.05, respectively. However, the characteristics of these changes differ. The CV of the elasticity decoupling index increased at the provincial and prefectural levels, rising from 1.161 to 1.563 to 1.419 and 2.669, respectively, while it decreased slightly at the county level, from 3.862 to 3.765. (4) The dominant type of decoupling at the provincial level had changed from expansive negative decoupling (ED) to strong decoupling (SD). Meanwhile, at the prefectural and county levels, the ED was still dominant, but the number of SD had increased significantly, rising from 75 to 138 at the prefectural level and from 748 to 1160 at the county level. This study demonstrates China’s carbon emissions sensitivity to scale, emphasizing the importance of adapting emission reduction measures to local conditions.

With the continuous progress of mobile Internet technology and GPS positioning technology of mobile devices, Social Network and Location Based Services (LBS) are gradually converging to form Location Based Social Network (LBSN). POI (Point of Interest) recommendation systems face the problems of variable user interests, very sparse user and POI check-in matrices, and nonlinear interaction modeling. To address the above problems, a Graph-enhanced Attention Graph Neural Network model is proposed for POI recommendation (POI-GAGN in short). POI-GAGN mines user/POI node representations on user-POI interaction graph, user-user social interaction graph, and POI-POI association interaction graph through interaction node feature extraction module, learns POI attribute information representations through text feature extraction module, and extracts short-term preference representations of users through short-term preference extraction module. A graph-enhanced attention mechanism is designed to correlates node representations, attribute information representations of POI, and short-term preferences of users with each other to achieve better information fusion. Finally, we conduct sufficient experiments on two real datasets to prove that the recommendation effect of POI-GAGN is better than other current advanced POI recommendation methods, and POI-GAGN can better overcome the problems of data sparsity and cold start in recommendations.

This paper analyzes the application of point of interest (POI) data in the location based service (LBS), and reveals the problems of POI data update. Referring to the share intentions of map communities, the paper proposes a user participating POI update model. The model reveals that the millions of mobile handlers can be the POI data sensors, and LBS is the platform of the model. A basic LBS platform is constructed in order to verify this model, and a prototype system of POI query and update based on LBS is implemented. The function modules and technical details of this system are discussed. Result of the system proves the model.

Location-based services (LBSs) are becoming an increasingly important component in our social and business life. All existing LBS providers support the nearest place searching via a single point of interest (POI) query. That is, in one query, a user is allowed to search for only one type of service. However, in real life, people usually need to search multiple different types of services and hope that their locations are as close as possible for convenience. For example, one user would like to search for a restaurant with a KTV nearby. To support this application scenario, we propose a novel LBS termed “ $N$ -in-One,” which is the first scheme to extend the function of single-POI LBS to multiple-POI LBS such that a single query can be employed to request multiple POIs that are geographically close. Providing “ $N$ -in-One” is challenging because: 1) serving a “ $N$ -in-One” query is not equivalent to serving $N$ queries independently due to the distance correlation among the $N$ POIs; and 2) the cask effect is getting worse in the service area mode of “ $N$ -in-one” as most of the returned results may be rendered useless when some hot POIs are blocked. To overcome these challenges, we propose several algorithms using computational geometry techniques to identify the best $K$ POIs that are geographically close and the service area (denoted by a given-sized rectangle) that can cover as many the best $Q$ clusters as possible while reducing the cask effect in the service area mode. Extensive simulations based on both synthetic and real world data demonstrate the effectiveness of the proposed algorithms.

Editor's evaluation: Disentangling the rhythms of human activity in the built environment for airborne transmission risk: An analysis of large-scale mobility data

Decision letter: Disentangling the rhythms of human activity in the built environment for airborne transmission risk: An analysis of large-scale mobility data

With the ubiquity of GPS-enabled smartphones, Location-Based Service (LBS) as a prominent product of social networks has become an essential part of our daily lives. People can easily socialize and share their check-in data (location, time, text, photo and etc.) via Location-Based Social Networks (LBSN). Through mining the check-in dataset, Point-Of-Interest (POI) recommendation systems assist users in exploring new attractive venues.The primary challenge regarding POI recommendation is to suggest a list of new interesting locations to the query user. Specifically, the excessive sparsity observed in user-location matrices is the main problem. The well known Collaborative Filtering (CF) methods are commonly used with other factors like geographical, social, context-oriented (e.g. text contents) and temporal effects to promote the effectiveness of the location recommendation systems. Despite the vital role of temporal influence, an insufficient amount of research has been devoted to considering the time factor in location recommendation. While we have an insufficient number of records regarding a user’s check-in at a particular location, predicting the time of the visit seems more problematic. We have dedicated part of our research to study various aspects of the temporal influence in the recommendation. Additionally, we use Twitter textual contents to extract a set of spatial phrases associated with each region. Such an act enriches the textual contents of the local POIs. This process enhances location recommendation systems, as it facilitates textual similarity among POI tags and the tweet history of the query user.In short, we aim to address the problems involved with both aspects of Location Inference and Recommendation in social networks. Our research in this thesis has four parts. Firstly, we define a problem which merely considers a single temporal aspect to enhance the performance of a location recommendation task. Subsequently, we develop a probabilistic model which detects a user’s temporal orientation based on visibility weights of POIs visited by her during weekday/weekend cycles. While this method is limited to a single temporal scale, the idea can be adapted to other time-related periodic cycles (e.g. daily home-work return trips). Secondly, we argue that the majority of existing methods merely concentrate on a limited number of temporal scales and neglect others. We propose a probabilistic generative model, named after Multi-aspect Time-related Influence (MATI) which employs the user’s check-in log to detect her temporal mobility pattern under various scales (e.g. minute, hour, day and so on). It then performs recommendation using multi-aspect temporal correlations between the query user and proposed locations. Thirdly, we further study the role of the time factor in recommendation models. We define a new problem to jointly model a pair of heterogeneous time-related effects (recency and the subset feature) in location recommendation. To address the challenges, we propose a generative model which computes the probability of the query user visiting a proposed location based on various homogeneous subset attributes. At the same time, the model calculates how likely the newly visited venues will obtain a higher rank compared to others. The model finally performs a POI recommendation through combining the effects learned from both homogeneous and heterogeneous temporal influences. Fourthly, we take textual contents into consideration to tackle the data sparsity problem in location recommendation systems. We propose an approach to detect focal spatial phrases associated with each specific scalable geographical area. For this task, we process GPS-enabled tweets. The main problem here is that Twitter messages are lexically varied and contain limited information. Our model calculates stickiness threshold to exploit the most probable segments out of the tweet contents. We employ a probabilistic model to measure how strongly each of the spatial keywords can be linked to the predefined region.

A Choice Model with a Diverging Choice Set for POI Data Analysis

Recent developments in mobile techniques have enabled a great variety of Location Based Services (LBS). A high positioning accuracy is a fundamental requirement for precision LBS applications, e.g., precise LBS marketing in shopping malls or indoor emergency evacuation services with mobile devices. However, most offerable commercial positioning systems, such as GPS/GNSS and RF-based systems, can not provide positioning accuracy within one meter. In this work, a new positioning approach is proposed for mobile devices, Called Photo Positioning, it can provide a high accuracy positioning service. The ”positioning” here means to find the location where a photo was taken by investigating the geometric relations between the images of points of interest (POI) in the photo and their location in the real world based on the principle of photo imaging. To implement a photo positioning system, three major components are needed, including a POI database, a method to recognize and locate POIs in a photo and an algorithm to calculate the position where the photo was taken from the POI information. A positioning algorithm based on the geometric similarity in photo imaging is presented in this work, and a prototype system is developed for Android smartphone platforms. Our experimental results shows that the average positioning error of the proposed photo positioning approach can be as low as 74.34 cm.

We would like to welcome you to the first international workshop on mobile location-based service or MLBS 2011! The number of "smart" wireless devices such as mobile phones and iPad-like computers has been rapidly growing. Being able to keep track of locations of moving devices can enhance a number of applications. Location-Based Service (LBS) is quickly becoming the next ubiquitous technology for a wide range of mobile applications, such as location positioning, location navigation, location-aware search, social networks, and ads, just to name a few. LBS was first deployed in the turn of the century by Palm VII, Swisscom, Vodafone, and DoCoMo [1]. These first wave of deployment performed location positioning based on the locations of the nearest cell towers. The accuracy of such approach ranges from one hundred to a few thousand meters, depending on the density of cell towers. In 2004, Global Position System (GPS) was tested successfully to work with a mobile phone by Quadcomm [2]. GPS is now available on most smart phones. GPS can achieve outdoor location positioning with approximately ten-meter accuracy. Its major shortcomings are high power consumption, long TTFF (time to the first fix), and unavailability in urban tunnels and indoor. Some remedies such as AGPS and hybrid cells have recently been researched, experimented, and deployed. Indoor positioning and indoor navigation (IPIN) is a capability that is in high demand in Asia, where most commerce activities take place in high-rise buildings. A shopper may want to find nearby stores selling a particular merchandise, a store may want to deliver coupons to nearby shoppers, and a user may want to locate nearby friends, to name just a few. Existing cell-based, GPS-based and hybrid technologies cannot perform accurate indoor location positioning. Solutions using WIFI signals can achieve sub-five-meter accuracy, both indoor and outdoor. However, access-point density must be high. Furthermore, WIFI-based positioning suffers from both time-consuming site survey (to survey their locations and signal strengths) and signal disturbance [3]. New signal sources such as radar, sonar and inertial navigation systems are now being actively investigated [4]. These devices also suffer from their limitations. A good solution is likely to be one that combines signals from complementary sources. Besides hardware solutions, data-driven software solutions can assist location positioning. For instance, maps and POIs (points of interest) can be mined from users' traveling history. Google recently experiments with a transit alert system based on user moving patterns [5]. It was observed that in a city like Zurich, where transits are always on time and traffic is light, predicting bus arrival time is trivial. In a city like Taipei and Hangzhou, where every bus is equipped with a GPS, bus arrival time can be predicted fairly accurately by taking traffic congestion into consideration. In a city like Bangalore and Beijing, where limited signals are available with transits (to-date) and traffic can be insanely congested, the system can use the position of a user on a bus (assuming the route of the bus is known) to predict the bus's location. We believe that user moving patterns can provide useful information for applications such as information ranking and ad matching. While the opportunities are exciting, preserving user privacy is an issue that must be thoroughly addressed when historical data of users are processed and mined [6]. Another important research area of LBS is data storage and query processing. As location-cognizant devices become ubiquitous, the volume of such data poses unprecedented challenges for LBS providers. Fresh data must be stored and indexed as they arrive, and historical data archived before main memory is full. Spatial query processing of LBS must be both fast and accurate (location updates are reflected in real-time). Traditional spatial data structures such as Quad-trees may not be able to handle such high-frequency current updates and queries. Furthermore, user privacy, including both location privacy and query privacy, must be preserved. In summary, there are at least six key areas of research and development to enhance LBS: Location signal acquisition and processing,Signal fusion,Spatial query processing,Privacy-preserved data mining,Power-conserving algorithms, andLBS-enabled system evaluation. We are excited to organize this first international workshop on mobile location-based service (MLBS). The workshop accepted twelve papers covering the aforementioned areas. We believe MLBS will be an active research direction in the next decade because of the momentum of smart mobile devices, the set of exciting problems to be tackled, and applications to be developed.

ABSTRACTWith the popularity of mobile devices and smartphones, we have witnessed rapid growth in mobile applications and services, especially in location-based services (LBS). According to a mobile marketing survey, maps/location searches are among the most utilized services on smartphones. Points of interest (POIs), such as stores, shops, gas stations, parking lots, and bus stops, are particularly important for maps/location searches. Existing map services such as Google Maps and Wikimapia are constructed manually either professionally or with crowd sourcing. However, manual annotation is costly and limited in current POI search services. With the abundance of information on the Web, many store POIs can be extracted from the Web. In this paper, we focus on automatically constructing a POI database to enable store POI map searches. We propose techniques that are required to construct a POI database, including focused crawling, information extraction, and information retrieval techniques. We first crawl Yellow Page web sites to obtain vocabularies of store names. These vocabularies are then investigated with search engines to obtain sentences containing these store names from search snippets in order to train a store name recognition model. To extract POIs scattered across the Web, we propose a query-based crawler to find address-bearing pages that might be used to extract addresses and store names. We crawled 1.25 million distinct POI pairs scattered across the Web and implemented a POI search service via Apache Lucent’s search platform, called Solr. The experimental results demonstrate that the proposed geographical information retrieval model outperforms Wikimapia and a commercial app called ‘What’s the Number?’

Context has been recognized as an important factor to consider in personalized recommender systems. Particularly in location-based services (LBSs), a fundamental task is to recommend to a mobile user where he/she could be interested to visit next at the right time. Additionally, location-based social networks (LBSNs) allow users to share location-embedded information with friends who often co-occur in the same or nearby points-of-interest (POIs) or share similar POI visiting histories, due to the social homophily theory and Tobler's first law of geography. So, both the time information and LBSN friendship relations should be utilized for POI recommendation. Tensor completion has recently gained some attention in time-aware recommender systems. The problem decomposes a user-item-time tensor into low-rank embedding matrices of users, items and times using its observed entries, so that the underlying low-rank subspace structure can be tracked to fill the missing entries for time-aware recommendation. However, these tensor completion methods ignore the social-spatial context information available in LBSNs, which is important for POI recommendation since people tend to share their preferences with their friends, and near things are more related than distant things. In this paper, we utilize the side information of social networks and POI locations to enhance the tensor completion model paradigm for more effective time-aware POI recommendation. Specifically, we propose a regularization loss head based on a novel social Hausdorff distance function to optimize the reconstructed tensor. We also quantify the popularity of different POIs with location entropy to prevent very popular POIs from being over-represented hence suppressing the appearance of other more diverse POIs. To address the sensitivity of negative sampling, we train the model on the whole data by treating all unlabeled entries in the observed tensor as negative, and rewriting the loss function in a smart way to reduce the computational cost. Through extensive experiments on real datasets, we demonstrate the superiority of our model over state-of-the-art tensor completion methods.

This work developed an open accessed Mobile Digital Culture Heritage (M-DCH) platform called Demodulating and Encoding Heritage (DEH) that allows users to download M-DCH contents and upload M-DEH contents. For the M-DCH content concern, DEH interprets M-DEH from the viewpoint of point, line, area and story/site. Point of Interest (POI) is the basic unit of M-DCH and can contain text plus image/pictures, video or audio. A Line of Interest (LOI) is a sequence of POIs in which a visiting sequence of the composed POIs is enforced to have the temporal or spatial view and interpretation of the contained POIs. An Area of Interest (AOI) is a set of POIs in which the composed POIs are related, but the visiting of them can be random. A Story/Site of Interest (SOI) can contain many units, in which each unit can be a POI, a LOI or an AOI, to describe a complete story or a site of culture heritage. For the technical concern, DEH defines a new scenario of the grouping function. A group in DEH is composed of a group creator/leader and some members. In addition to the regular group functions i.e., sharing the content to group members, which belongs to the read action, defined in common social network, e.g., FB, Line, and WeChat, DEH's grouping function allows the group creator/leader to have the privilege of modifying contents created by group members, which belongs to the write action. This function is especially needed for content creation and sharing that is instructed/advised/guided in the teacher-student way. A set of metadata elements was defined for POI, LOI, AOI and SOI respectively such that the system can conveniently manipulate the M-DEH contents. The DEH platform was implemented by utilizing many tools for web and mobile APPs, e.g., Django for web API and Node.js for mobile API, based on the techniques of Location Based Service (LBS) and Geographical Information System (GIS).