Moving objects databases (MODs) have been extensively studied due to their wide variety of applications including traffic management, tourist service and mobile commerce. However, queries in natural languages are still not supported in MODs. Since most users are not familiar with structured query languages, it is essentially important to bridge the gap between natural languages and the underlying MODs system commands. Motivated by this, we design a natural language interface for moving objects, named NALMO. In general, we use semantic parsing in combination with a location knowledge base and domain-specific rules to interpret natural language queries. We design a corpus of moving objects queries for model training, which is later used to determine the query type. Extracted entities from parsing are mapped through deterministic rules to perform query composition. NALMO is able to well translate moving objects queries into structured (executable) languages. We support four kinds of queries including time interval queries, range queries, nearest neighbor queries and trajectory similarity queries. We develop the system in a prototype system SECONDO and evaluate our approach using 240 natural language queries extracted from popular conference and journal papers in the domain of moving objects. Experimental results show that (i) NALMO achieves accuracy and precision 98.1 and 88.1, respectively, and (ii) the average time cost of translating a query is 1.47s.
{"title":"NALMO: A Natural Language Interface for Moving Objects Databases","authors":"Xieyang Wang, Jianqiu Xu, Hua Lu","doi":"10.1145/3469830.3470894","DOIUrl":"https://doi.org/10.1145/3469830.3470894","url":null,"abstract":"Moving objects databases (MODs) have been extensively studied due to their wide variety of applications including traffic management, tourist service and mobile commerce. However, queries in natural languages are still not supported in MODs. Since most users are not familiar with structured query languages, it is essentially important to bridge the gap between natural languages and the underlying MODs system commands. Motivated by this, we design a natural language interface for moving objects, named NALMO. In general, we use semantic parsing in combination with a location knowledge base and domain-specific rules to interpret natural language queries. We design a corpus of moving objects queries for model training, which is later used to determine the query type. Extracted entities from parsing are mapped through deterministic rules to perform query composition. NALMO is able to well translate moving objects queries into structured (executable) languages. We support four kinds of queries including time interval queries, range queries, nearest neighbor queries and trajectory similarity queries. We develop the system in a prototype system SECONDO and evaluate our approach using 240 natural language queries extracted from popular conference and journal papers in the domain of moving objects. Experimental results show that (i) NALMO achieves accuracy and precision 98.1 and 88.1, respectively, and (ii) the average time cost of translating a query is 1.47s.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115515231","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xu Teng, Adam Corpstein, Joel Holm, Willis Knox, Becker Mathie, Philip R. O. Payne, Ethan Vander Wiel, Prabin Giri, Goce Trajcevski, A. Dotter, J. Andrews, S. Coughlin, Y. Qin, J. G. Serra-Perez, N. Tran, Jaime Roman-Garja, K. Kovlakas, E. Zapartas, S. Bavera, D. Misra, T. Fragos
We present CACSE – a system for Context Aware Clustering of Stellar Evolution – for datasets corresponding to temporal evolution of stars, which are multivariate time series, usually with a large number of attributes (e.g., ≥ 40). Typically, the datasets are obtained by simulation and are relatively large in size (5 ∼ 10 GB per certain interval of values for various initial conditions). Investigating common evolutionary trends in these datasets often depends on the context – i.e., not all the attributes are always of interest, and among the subset of the context-relevant attributes, some may have more impact than others. To enable such context-aware clustering, our CACSE system provides functionalities allowing the domain experts to dynamically select attributes that matter, and assign desired weights/priorities. Our system consists of a PostgreSQL database, Python-based middleware with RESTful and Django framework, and a web-based user interface as frontend. The user interface provides multiple interactive options, including selection of datasets and preferred attributes along with the corresponding weights. Subsequently, the users can select a time instant or a time range to visualize the formed clusters. Thus, CACSE enables a detection of changes in the the set of clusters (i.e., convoys) of stellar evolution tracks. Current version provides two of the most popular clustering algorithms – k-means and DBSCAN.
{"title":"CACSE: Context Aware Clustering of Stellar Evolution","authors":"Xu Teng, Adam Corpstein, Joel Holm, Willis Knox, Becker Mathie, Philip R. O. Payne, Ethan Vander Wiel, Prabin Giri, Goce Trajcevski, A. Dotter, J. Andrews, S. Coughlin, Y. Qin, J. G. Serra-Perez, N. Tran, Jaime Roman-Garja, K. Kovlakas, E. Zapartas, S. Bavera, D. Misra, T. Fragos","doi":"10.1145/3469830.3470916","DOIUrl":"https://doi.org/10.1145/3469830.3470916","url":null,"abstract":"We present CACSE – a system for Context Aware Clustering of Stellar Evolution – for datasets corresponding to temporal evolution of stars, which are multivariate time series, usually with a large number of attributes (e.g., ≥ 40). Typically, the datasets are obtained by simulation and are relatively large in size (5 ∼ 10 GB per certain interval of values for various initial conditions). Investigating common evolutionary trends in these datasets often depends on the context – i.e., not all the attributes are always of interest, and among the subset of the context-relevant attributes, some may have more impact than others. To enable such context-aware clustering, our CACSE system provides functionalities allowing the domain experts to dynamically select attributes that matter, and assign desired weights/priorities. Our system consists of a PostgreSQL database, Python-based middleware with RESTful and Django framework, and a web-based user interface as frontend. The user interface provides multiple interactive options, including selection of datasets and preferred attributes along with the corresponding weights. Subsequently, the users can select a time instant or a time range to visualize the formed clusters. Thus, CACSE enables a detection of changes in the the set of clusters (i.e., convoys) of stellar evolution tracks. Current version provides two of the most popular clustering algorithms – k-means and DBSCAN.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125777264","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hafsa El Hafyani, Mohammad Abboud, Jingwei Zuo, K. Zeitouni, Y. Taher
Wide spread use of sensors and mobile devices along with the new paradigm of Mobile Crowd-Sensing (MCS), allows monitoring air pollution in urban areas. Several measurements are collected, such as Particulate Matters, Nitrogen dioxide, and others. Mining the context of MCS data in such domains is a key factor for identifying the individuals’ exposure to air pollution, but it is challenging due to the lack or the weakness of predictors. We have previously developed a multi-view learning approach which learns the context solely from the sensor measurements. In this demonstration, we propose a visualization tool (COMIC) showing the different recognized contexts using an improved version of our algorithm. We also demonstrate the change points detected by a multi-dimensional CPD model. We leverage real data from a MCS campaign, and compare different methods.
{"title":"Tell Me What Air You Breath, I Tell You Where You Are","authors":"Hafsa El Hafyani, Mohammad Abboud, Jingwei Zuo, K. Zeitouni, Y. Taher","doi":"10.1145/3469830.3470914","DOIUrl":"https://doi.org/10.1145/3469830.3470914","url":null,"abstract":"Wide spread use of sensors and mobile devices along with the new paradigm of Mobile Crowd-Sensing (MCS), allows monitoring air pollution in urban areas. Several measurements are collected, such as Particulate Matters, Nitrogen dioxide, and others. Mining the context of MCS data in such domains is a key factor for identifying the individuals’ exposure to air pollution, but it is challenging due to the lack or the weakness of predictors. We have previously developed a multi-view learning approach which learns the context solely from the sensor measurements. In this demonstration, we propose a visualization tool (COMIC) showing the different recognized contexts using an improved version of our algorithm. We also demonstrate the change points detected by a multi-dimensional CPD model. We leverage real data from a MCS campaign, and compare different methods.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129691428","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Panagiotis Tampakis, Dimitris Spyrellis, C. Doulkeridis, N. Pelekis, Christos Kalyvas, Akrivi Vlachou
Spatial-keyword queries are important for a wide range of applications that retrieve data based on a combination of keyword search and spatial constraints. However, efficient processing of spatial-keyword queries is not a trivial task because the combination of textual and spatial data results in a high-dimensional representation that is challenging to index effectively. To address this problem, in this paper, we propose a novel indexing scheme for efficient support of spatial-keyword range queries. At the heart of our approach lies a carefully-designed mapping of spatio-textual data to a two-dimensional (2D) space that produces compact partitions of spatio-textual data. In turn, the mapped 2D data can be indexed effectively by traditional spatial data structures, such as an R-tree. We propose bounds, theoretically proven for correctness, that lead to the design of a filter-and-refine algorithm that prunes the search space effectively. In this way, our approach for spatial-keyword range queries is readily applicable to any database system that provides spatial support. In our experimental evaluation, we demonstrate how our algorithm can be implemented over PostgreSQL and exploit its underlying spatial index provided by PostGIS, in order to process spatial-keyword range queries efficiently. Moreover, we show that our solution outperforms different competitor approaches.
{"title":"A Novel Indexing Method for Spatial-Keyword Range Queries","authors":"Panagiotis Tampakis, Dimitris Spyrellis, C. Doulkeridis, N. Pelekis, Christos Kalyvas, Akrivi Vlachou","doi":"10.1145/3469830.3470897","DOIUrl":"https://doi.org/10.1145/3469830.3470897","url":null,"abstract":"Spatial-keyword queries are important for a wide range of applications that retrieve data based on a combination of keyword search and spatial constraints. However, efficient processing of spatial-keyword queries is not a trivial task because the combination of textual and spatial data results in a high-dimensional representation that is challenging to index effectively. To address this problem, in this paper, we propose a novel indexing scheme for efficient support of spatial-keyword range queries. At the heart of our approach lies a carefully-designed mapping of spatio-textual data to a two-dimensional (2D) space that produces compact partitions of spatio-textual data. In turn, the mapped 2D data can be indexed effectively by traditional spatial data structures, such as an R-tree. We propose bounds, theoretically proven for correctness, that lead to the design of a filter-and-refine algorithm that prunes the search space effectively. In this way, our approach for spatial-keyword range queries is readily applicable to any database system that provides spatial support. In our experimental evaluation, we demonstrate how our algorithm can be implemented over PostgreSQL and exploit its underlying spatial index provided by PostGIS, in order to process spatial-keyword range queries efficiently. Moreover, we show that our solution outperforms different competitor approaches.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"91 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129252143","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Linas Petkevičius, Simonas Šaltenis, A. Civilis, K. Torp
The continued spread of electric vehicles raises new challenges for the supporting digital infrastructure. For example, long-distance route planning for such vehicles relies on the prediction of both the expected travel time as well as energy use. We envision a two-tier architecture to produce such predictions. First, a routing and travel-time-prediction subsystem generates a suggested route and predicts how the speed will vary along the route. Next, the expected energy use is predicted from the speed profile and other contextual characteristics, such as weather information and slope. To this end, the paper proposes deep-learning models that are built from EV tracking data. First, as the speed profile of a route is one of the main predictors for energy use, different simple ways to build speed profiles are explored. Next, eight different deep-learning models for energy-use prediction are proposed. Four of the models are probabilistic in that they predict not a single-point estimate but parameters of a probability distribution of energy use on the route. This is particularly relevant when predicting EV energy use, which is highly sensitive to many input characteristics and, thus, can hardly be predicted precisely. Extensive experiments with two real-world EV tracking datasets validate the proposed methods. The code for this research has been made available on GitHub.
{"title":"Probabilistic Deep Learning for Electric-Vehicle Energy-Use Prediction","authors":"Linas Petkevičius, Simonas Šaltenis, A. Civilis, K. Torp","doi":"10.1145/3469830.3470915","DOIUrl":"https://doi.org/10.1145/3469830.3470915","url":null,"abstract":"The continued spread of electric vehicles raises new challenges for the supporting digital infrastructure. For example, long-distance route planning for such vehicles relies on the prediction of both the expected travel time as well as energy use. We envision a two-tier architecture to produce such predictions. First, a routing and travel-time-prediction subsystem generates a suggested route and predicts how the speed will vary along the route. Next, the expected energy use is predicted from the speed profile and other contextual characteristics, such as weather information and slope. To this end, the paper proposes deep-learning models that are built from EV tracking data. First, as the speed profile of a route is one of the main predictors for energy use, different simple ways to build speed profiles are explored. Next, eight different deep-learning models for energy-use prediction are proposed. Four of the models are probabilistic in that they predict not a single-point estimate but parameters of a probability distribution of energy use on the route. This is particularly relevant when predicting EV energy use, which is highly sensitive to many input characteristics and, thus, can hardly be predicted precisely. Extensive experiments with two real-world EV tracking datasets validate the proposed methods. The code for this research has been made available on GitHub.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"111 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124231371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
T. Petersen, M. A. Suryani, C. Beth, Hardik Patel, K. Wallmann, M. Renz
Quantitative information derived from scientific documents provides an important source of data for studies in almost all domains, however, manual extraction of this information is very time consuming. In this paper we will introduce a system Geo-Quantities that supports the automatic extraction of quantitative, spatial and temporal information of a given measurement entity from scientific literature using text mining techniques. The difficulty of automatic measurement recognition is mainly caused by the diverse expressions in the papers. Geo-Quantities offers an interactive interface for the visualization of extracted user-defined information, in particular spatial and temporal context. In our demonstration, we will showcase the capabilities of our system by retrieving measurements such as “mass accumulation rates” and “sedimentation rates” from scientific publications in the field of marine geology, which could have high impact in studies for building global mass accumulation rate maps. For training and evaluation of Geo-Quantities we use a corpus of domain-relevant papers.
{"title":"Geo-Quantities: A Framework for Automatic Extraction of Measurements and Spatial Context from Scientific Documents","authors":"T. Petersen, M. A. Suryani, C. Beth, Hardik Patel, K. Wallmann, M. Renz","doi":"10.1145/3469830.3470911","DOIUrl":"https://doi.org/10.1145/3469830.3470911","url":null,"abstract":"Quantitative information derived from scientific documents provides an important source of data for studies in almost all domains, however, manual extraction of this information is very time consuming. In this paper we will introduce a system Geo-Quantities that supports the automatic extraction of quantitative, spatial and temporal information of a given measurement entity from scientific literature using text mining techniques. The difficulty of automatic measurement recognition is mainly caused by the diverse expressions in the papers. Geo-Quantities offers an interactive interface for the visualization of extracted user-defined information, in particular spatial and temporal context. In our demonstration, we will showcase the capabilities of our system by retrieving measurements such as “mass accumulation rates” and “sedimentation rates” from scientific publications in the field of marine geology, which could have high impact in studies for building global mass accumulation rate maps. For training and evaluation of Geo-Quantities we use a corpus of domain-relevant papers.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"119 9","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114112890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Earth observation data is collected by ever-expanding fleets of satellites including Landsat1-8, Sentinel1 & Sentinel2, SPOT1-7 and WorldView1-3. These satellites generate at spatial resolutions (pixel size) from 30m to 31cm and provide revisit rates of as frequent as every 5 days. This allows us not only to look at high-resolution images of every corner of the Earth, but also to track events and observe change over time. During the past 5 years, medium spatial resolution satellite data (30 − 10m pixels) have developed very high temporal revisit frequencies of 5-16 days and spatial-temporal structures have been developed to manage these vast data sets. However, high resolution satellite images and rapidly increasing revisit rates create major data management and mining challenges. This work discusses six challenges of integrating observations at different times, from different sensors, at different spatial resolutions and different temporal frequencies into a unified Earth Observation Data Cube, that is, a tensor of location, time, and spectral bands. Challenges include creating a unified data cube from heterogeneous sensors, scaling geo-registration (mapping pixel between images), accounting for uncertainty across observations, imputing missing observations, broad area event detection, and ultimately, predicting the future state of our planet. With such a unified Earth Observation Data Cube in place, we describe potential application areas such as detecting anthropogenic land cover change, early warning of natural hazards, tracing movement of animals, finding missing airplanes, and rapid detection of forest fires.
{"title":"Mining High Resolution Earth Observation Data Cubes","authors":"Andreas Zuefle, K. Wessels, D. Pfoser","doi":"10.1145/3469830.3470917","DOIUrl":"https://doi.org/10.1145/3469830.3470917","url":null,"abstract":"Earth observation data is collected by ever-expanding fleets of satellites including Landsat1-8, Sentinel1 & Sentinel2, SPOT1-7 and WorldView1-3. These satellites generate at spatial resolutions (pixel size) from 30m to 31cm and provide revisit rates of as frequent as every 5 days. This allows us not only to look at high-resolution images of every corner of the Earth, but also to track events and observe change over time. During the past 5 years, medium spatial resolution satellite data (30 − 10m pixels) have developed very high temporal revisit frequencies of 5-16 days and spatial-temporal structures have been developed to manage these vast data sets. However, high resolution satellite images and rapidly increasing revisit rates create major data management and mining challenges. This work discusses six challenges of integrating observations at different times, from different sensors, at different spatial resolutions and different temporal frequencies into a unified Earth Observation Data Cube, that is, a tensor of location, time, and spectral bands. Challenges include creating a unified data cube from heterogeneous sensors, scaling geo-registration (mapping pixel between images), accounting for uncertainty across observations, imputing missing observations, broad area event detection, and ultimately, predicting the future state of our planet. With such a unified Earth Observation Data Cube in place, we describe potential application areas such as detecting anthropogenic land cover change, early warning of natural hazards, tracing movement of animals, finding missing airplanes, and rapid detection of forest fires.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"2010 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125634000","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Teddy Cunningham, Graham Cormode, H. Ferhatosmanoğlu
Sharing sensitive data is vital in enabling many modern data analysis and machine learning tasks. However, current methods for data release are insufficiently accurate or granular to provide meaningful utility, and they carry a high risk of deanonymization or membership inference attacks. In this paper, we propose a differentially private synthetic data generation solution with a focus on the compelling domain of location data. We present two methods with high practical utility for generating synthetic location data from real locations, both of which protect the existence and true location of each individual in the original dataset. Our first, partitioning-based approach introduces a novel method for privately generating point data using kernel density estimation, in addition to employing private adaptations of classic statistical techniques, such as clustering, for private partitioning. Our second, network-based approach incorporates public geographic information, such as the road network of a city, to constrain the bounds of synthetic data points and hence improve the accuracy of the synthetic data. Both methods satisfy the requirements of differential privacy, while also enabling accurate generation of synthetic data that aims to preserve the distribution of the real locations. We conduct experiments using three large-scale location datasets to show that the proposed solutions generate synthetic location data with high utility and strong similarity to the real datasets. We highlight some practical applications for our work by applying our synthetic data to a range of location analytics queries, and we demonstrate that our synthetic data produces near-identical answers to the same queries compared to when real data is used. Our results show that the proposed approaches are practical solutions for sharing and analyzing sensitive location data privately.
{"title":"Privacy-Preserving Synthetic Location Data in the Real World","authors":"Teddy Cunningham, Graham Cormode, H. Ferhatosmanoğlu","doi":"10.1145/3469830.3470893","DOIUrl":"https://doi.org/10.1145/3469830.3470893","url":null,"abstract":"Sharing sensitive data is vital in enabling many modern data analysis and machine learning tasks. However, current methods for data release are insufficiently accurate or granular to provide meaningful utility, and they carry a high risk of deanonymization or membership inference attacks. In this paper, we propose a differentially private synthetic data generation solution with a focus on the compelling domain of location data. We present two methods with high practical utility for generating synthetic location data from real locations, both of which protect the existence and true location of each individual in the original dataset. Our first, partitioning-based approach introduces a novel method for privately generating point data using kernel density estimation, in addition to employing private adaptations of classic statistical techniques, such as clustering, for private partitioning. Our second, network-based approach incorporates public geographic information, such as the road network of a city, to constrain the bounds of synthetic data points and hence improve the accuracy of the synthetic data. Both methods satisfy the requirements of differential privacy, while also enabling accurate generation of synthetic data that aims to preserve the distribution of the real locations. We conduct experiments using three large-scale location datasets to show that the proposed solutions generate synthetic location data with high utility and strong similarity to the real datasets. We highlight some practical applications for our work by applying our synthetic data to a range of location analytics queries, and we demonstrate that our synthetic data produces near-identical answers to the same queries compared to when real data is used. Our results show that the proposed approaches are practical solutions for sharing and analyzing sensitive location data privately.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-08-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132590728","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Online gaming is growing faster than ever before, with increasing challenges of providing better user experience. Recommender systems (RS) for online games face unique challenges since they must fulfill players’ distinct desires, at different user levels, based on their action sequences of various action types. Although many sequential RS already exist, they are mainly single-sequence, single-task, and single-user-level. In this paper, we introduce a new sequential recommendation model for multiple sequences, multiple tasks, and multiple user levels (abbreviated as M3Rec) in Tencent Games platform, which can fully utilize complex data in online games. We leverage Graph Neural Network and multi-task learning to design M3Rec in order to model the complex information in the heterogeneous sequential recommendation scenario of Tencent Games. We verify the effectiveness of M3Rec on three online games of Tencent Games platform, in both offline and online evaluations. The results show that M3Rec successfully addresses the challenges of recommendation in online games, and it generates superior recommendations compared with state-of-the-art sequential recommendation approaches.
{"title":"Sequential Recommendation in Online Games with Multiple Sequences, Tasks and User Levels","authors":"Si Chen, Yuqiu Qian, Hui Li, Chen Lin","doi":"10.1145/3469830.3470906","DOIUrl":"https://doi.org/10.1145/3469830.3470906","url":null,"abstract":"Online gaming is growing faster than ever before, with increasing challenges of providing better user experience. Recommender systems (RS) for online games face unique challenges since they must fulfill players’ distinct desires, at different user levels, based on their action sequences of various action types. Although many sequential RS already exist, they are mainly single-sequence, single-task, and single-user-level. In this paper, we introduce a new sequential recommendation model for multiple sequences, multiple tasks, and multiple user levels (abbreviated as M3Rec) in Tencent Games platform, which can fully utilize complex data in online games. We leverage Graph Neural Network and multi-task learning to design M3Rec in order to model the complex information in the heterogeneous sequential recommendation scenario of Tencent Games. We verify the effectiveness of M3Rec on three online games of Tencent Games platform, in both offline and online evaluations. The results show that M3Rec successfully addresses the challenges of recommendation in online games, and it generates superior recommendations compared with state-of-the-art sequential recommendation approaches.","PeriodicalId":206910,"journal":{"name":"17th International Symposium on Spatial and Temporal Databases","volume":"218 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131833186","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：