微軟亞洲研究院基於GPS數據展開的研究工做,取得了另學術界矚目的成就。從2008年開始每一年都在頂級的計算機類會議上有文章發出,掀起了研究GPS數據智能化處理的熱潮。 git
他們的工做由謝幸研究員和鄭宇研究員主導。實驗數據採集主要有兩個工程: web
一、Geolife工程使用的,170多個志願者4年左右的GPS軌跡; 算法
二、北京市2萬多出租車約3個月的行駛軌跡。 數據庫
參見:http://research.microsoft.com/en-us/projects/geolife/ 網絡
我的將近年來該小組發表的論文作了下整理,概括其研究脈絡以下(我的愚見,僅供拍磚,前面序號按照文章發表時間倒序排列,時間排列文獻全表見最後的「參考文獻」部分): 數據結構
階段一: GPS+webGIS:(GPS數據在GIS地圖上的靈活展現和使用) app
4一、(生成用戶日誌,幫助用戶回憶),Searching Your Life on Web Maps,SIGIR2008 性能
4二、(GeoLife1.0,在地圖上管理本身的GPS日誌)GeoLife:Managing and Understanding Your Past Life over Maps,MDM08 學習
階段二:GPS數據的簡單挖掘(出行方式,用戶軌跡的類似性,關聯規則,異常行爲,羣組發現) ui
40、(理解用戶行爲),Understanding User Behavior Geospatially,計算機通訊2008
3六、(出行模式:步行,公交,駕駛)Learning Transportation Mode from Raw GPS Data for GeographicApplication on the Web,www2008
3七、(理解移動特性),Understanding mobility based on GPS data,ubicom2008
3八、(基於軌跡挖掘用戶類似性)Mining user similarity based on location history,GIS2008
3三、(挖掘有趣位置及旅行順序)Mining Interesting Locations and Travel Sequences From GPSTrajectories,WWW 2009
3四、(挖掘我的生活模式)Mining Individual Life Pattern Based on Location History,MDM 2009
3五、(GeoLife生活模式),GeoLife2.0: A Location-Based Social Networking Service,MDM2009
30、(挖掘位置間的關聯),Mining Correlation Between Locations Using Human Location History,GIS 2009
2九、(基於GPS軌跡發現用戶交通模式:行走、駕車、公交、地鐵),Understanding transportation modes based on GPS data for Webapplications,ACMTransaction on the Web,2010
2三、(學習位置間的關聯)Learning Location Correlation from GPS trajectories,MDM2010
九、(交通異常,前一個版本)Discovering Spatio-Temporal Causal Interactions in Traffic DataStreams,sigkdd2011
一、羣組發現(提出了一個羣組發現加速的數據結構):A Framework of Traveling Companion Discovery on Trajectory DataStreams,ICDE2012
階段三:結合外部信息,提供基於雲的LBS服務
2七、(計算機學會刊文)Enabling smart location-based services by mining GPS traces,Communication of China ComputerFederation2010
2八、(基於雲的位置服務)Location-Based Services on the Cloud,Journal on Digital Mobile Era 2010
十、(基於雲的,融合多種信息,探尋用戶習慣的發現)Driving with Knowledge from the Physical World,SIGKDD 2011
階段四:同協同過濾等算法結合實現位置、活動的推薦
2四、(位置和活動推薦)Collaborative Location and Activity Recommendations With GPS HistoryData,WWW 2010
2二、(基於用戶的推薦,旅遊景點和行程)Collaborative Filtering Meets Mobile Recommendation: A User-centeredApproach, AAAI2010
20、(前期版本,智能的行程推薦)Smart Itinerary Recommendation based on User-Generated GPSTrajectories,UIC2010
2一、(GeoLife,智能生活,雙層網絡),GeoLife: A Collaborative Social Networking Service among User,location and trajectory,IEEE Data(base) Engineering Bulletin
1四、(旅遊推薦,較全面的工做),Learning travel recommendations from user-generated GPS traces, ACM Transaction onIntelligent Systems and Technology,2011
1五、(根據單個軌跡推薦位置和朋友),Recommending friends and locations based on individual locationhistory,2011
十二、(我的行程推薦,最第一版本),Social Itinerary Recommendation from User-generated Digital Trails,Personal and Ubiquitous Computing
二、熱點旅遊位置推薦、周邊活動發現(第二次發表:協同過濾):Towards Mobile Intelligence: Learning from GPS History Data forCollaborative Recommendation, Artificial Intelligence Journal,2012-04-29
階段五:引入語義軌跡,泛華推薦和類似性算法
1九、(利用語義軌跡,發現類似用戶),Finding Similar Users Using Category-Based Location History,GIS2010
三、(語義軌跡歷史)基於GPS推到用戶的社會關係:InferringSocial Ties between Users with Human Location History,Journal of Ambient Intelligence andHumanized Computing
階段六:根據出租車數據,實現智能化駕駛
1六、(根據出租車歷史,指導行駛方向)T-Drive: Driving Directions Based on Taxi Trajectories,GIS,2010
1七、(出租司機,靈活駕駛)Drive smartly as a taxi driver,UIC2010
1三、(推到出租車行駛情況),Inferring Taxi Status Using GPS Trajectories,2011
六、出租車數據探測城市的異常交通:On Mining Anomalous Patterns in Road Traffic Streams:2011International Conference on Advanced Data Mining and Applications
七、(基於出租數據的城市規劃,計算)Urban Computing with Taxicabs,
八、(尋找下一個旅客)Where to Find My Next Passenger?Ubicom2011
四、基於時間關係圖發現出租車行駛路線:T-Drive: Enhancing Driving Directions with Taxi Drivers'Intelligence,IEEETransactions on Knowledge and Data Engineering (TKDE)
(貫穿)軌跡的預處理類工做:
3九、(靈活的時空索引),A Flexible Spatio-temporal Indexing Scheme for Large-scale GPS TrackRetrieval,2008
3二、(基於客觀需求的,軌跡簡化策略,方向變化的位置信息量越大)Trajectory Simplification Method for Location-Based SocialNetworking Services,SIGSPATIAL GIS workshop on location-based social networks,2009
3一、(稀疏採樣軌跡的地圖匹配)Map-Matching for Low-Sampling-Rate GPS Trajectories,GIS2009
2六、(Top-k區域查詢)AnsweringTop-k Similar Region Queries,DASFAA 2010
2五、(基於位置點,搜索軌跡)Searching Trajectories by Locations: An Efficiency Study,SIGMOD
1八、(發現數據庫中的類似記錄),Detecting Nearly Duplicated Records in Location Datasets,GIS2010
十一、(給定一組點,檢索k個最近鄰的點)Retrievingk-Nearest Neighboring Trajectories by a Set of Point Locations,SSTD2011
五、低採樣軌跡的不肯定性處理(比地圖匹配性能更好):Reducing Uncertainty of Low-Sampling-Rate Trajectories, ICDE 2012
一、羣組發現(提出了一個羣組發現加速的數據結構):A Framework of Traveling Companion Discovery on Trajectory DataStreams,ICDE2012
二、熱點旅遊位置推薦、周邊活動發現(第二次發表:協同過濾):Towards Mobile Intelligence: Learning from GPS History Data forCollaborative Recommendation, Artificial Intelligence Journal,2012-04-29
三、(語義軌跡歷史)基於GPS推到用戶的社會關係:InferringSocial Ties between Users with Human Location History,Journal of Ambient Intelligence andHumanized Computing
四、基於時間關係圖發現出租車行駛路線:T-Drive: Enhancing Driving Directions with Taxi Drivers'Intelligence,IEEETransactions on Knowledge and Data Engineering (TKDE)
五、低採樣軌跡的不肯定性處理(比地圖匹配性能更好):Reducing Uncertainty of Low-Sampling-Rate Trajectories, ICDE 2012
六、出租車數據探測城市的異常交通:On Mining Anomalous Patterns in Road Traffic Streams:2011International Conference on Advanced Data Mining and Applications
七、(基於出租數據的城市規劃,計算)Urban Computing with Taxicabs,
八、(尋找下一個旅客)Where to Find My Next Passenger?Ubicom2011
九、(交通異常,前一個版本)Discovering Spatio-Temporal Causal Interactions in Traffic DataStreams,sigkdd2011
十、(基於雲的,融合多種信息,探尋用戶習慣的發現)Driving with Knowledge from the Physical World,SIGKDD 2011
十一、(給定一組點,檢索k個最近鄰的點)Retrievingk-Nearest Neighboring Trajectories by a Set of Point Locations,SSTD2011
十二、(我的行程推薦,最第一版本),Social Itinerary Recommendation from User-generated Digital Trails,Personal and Ubiquitous Computing
1三、(推到出租車行駛情況),Inferring Taxi Status Using GPS Trajectories,2011
1四、(旅遊推薦,較全面的工做),Learning travel recommendations from user-generated GPS traces, ACM Transaction onIntelligent Systems and Technology,2011
1五、(根據單個軌跡推薦位置和朋友),Recommending friends and locations based on individual locationhistory,2011
1六、(根據出租車歷史,指導行駛方向)T-Drive: Driving Directions Based on Taxi Trajectories,GIS,2010
1七、(出租司機,靈活駕駛)Drive smartly as a taxi driver,UIC2010
1八、(發現數據庫中的類似記錄),Detecting Nearly Duplicated Records in Location Datasets,GIS2010
1九、(利用語義軌跡,發現類似用戶),Finding Similar Users Using Category-Based Location History,GIS2010
20、(前期版本,智能的造成推薦)Smart Itinerary Recommendation based on User-Generated GPSTrajectories,UIC2010
2一、(GeoLife,智能生活,雙層網絡),GeoLife: A Collaborative Social Networking Service among User,location and trajectory,IEEE Data(base) Engineering Bulletin
2二、(基於用戶的推薦,旅遊景點和造成)Collaborative Filtering Meets Mobile Recommendation: A User-centeredApproach, AAAI2010
2三、(學習位置間的關聯)Learning Location Correlation from GPS trajectories,MDM2010
2四、(位置和活動推薦)Collaborative Location and Activity Recommendations With GPS HistoryData,WWW 2010
2五、(基於位置點,搜索軌跡)Searching Trajectories by Locations: An Efficiency Study,SIGMOD 2010
2六、(Top-k區域查詢)AnsweringTop-k Similar Region Queries,DASFAA 2010
2七、(計算機學會刊文)Enabling smart location-based services by mining GPS traces,Communication of China ComputerFederation2010
2八、(基於雲的位置服務)Location-Based Services on the Cloud,Journal on Digital Mobile Era 2010
2九、(基於GPS軌跡發現用戶交通模式:行走、駕車、公交、地鐵),Understanding transportation modes based on GPS data for Webapplications,ACMTransaction on the Web,2010
30、(挖掘位置間的關聯),Mining Correlation Between Locations Using Human Location History,GIS 2009
3一、(稀疏採樣軌跡的地圖匹配)Map-Matching for Low-Sampling-Rate GPS Trajectories,GIS2009
3二、(基於客觀需求的,軌跡簡化策略,方向變化的位置信息量越大)Trajectory Simplification Method for Location-Based SocialNetworking Services,SIGSPATIAL GIS workshop on location-based social networks,2009
3三、(挖掘有趣位置及旅行順序)Mining Interesting Locations and Travel Sequences From GPSTrajectories,WWW 2009
3四、(挖掘我的生活模式)Mining Individual Life Pattern Based on Location History,MDM 2009
3五、(GeoLife生活模式),GeoLife2.0: A Location-Based Social Networking Service,MDM2009
3六、(出行模式:步行,公交,駕駛)Learning Transportation Mode from Raw GPS Data for GeographicApplication on the Web,www2008
3七、(理解移動特性),Understanding mobility based on GPS data,ubicom2008
3八、(基於軌跡挖掘用戶類似性)Mining user similarity based on location history,GIS2008
3九、(靈活的時空索引),A Flexible Spatio-temporal Indexing Scheme for Large-scale GPS TrackRetrieval,2008
40、(理解用戶行爲),Understanding User Behavior Geospatially,計算機通訊2008
4一、(生成用戶日誌,幫助用戶回憶),Searching Your Life on Web Maps,SIGIR2008
4二、(GeoLife1.0,在地圖上管理本身的GPS日誌)GeoLife:Managing and Understanding Your Past Life over Maps,MDM08