顾及时空语义的疑犯位置时空预测
|
摘要
预测疑犯的社会活动行踪,对案件嫌疑人的排查以及犯罪行为的主动预防具有重大意义。当前研究主要依据疑犯的历史系列作案位置预测其住址或未来犯罪位置,缺少对其复杂社会活动位置的转移过程进行建模,也没有考虑位置数据稀疏性对预测性能产生的影响。为此,提出了融合时空语义的位置时空预测(spatio-temporal semantics location prediction,SSLP)模型。首先,利用疑犯在不同语义时段和语义位置上的分布邻近性提取目标疑犯的相似疑犯群体;其次,结合该群体的轨迹数据和位置语义信息,基于核密度平滑方法估算出涉及未记录位置的转移频次及其时态访问概率;最后,采用贝叶斯模型实现疑犯个体的时空预测。实验结果表明,基于W市2013年1月至6月间158名疑犯的17 539个位置记录数据,SSLP模型在top-k距离偏离度和top-k精确率上优于其他流行方法 40%～50%,对疑犯位置数据稀疏性具有优异的适应能力。
Existing studies have failed to capture the complex social location transition patterns of suspects and lack the capacity to address the issue of data sparsity. Therefore, we propose a location prediction model called SSLP(spatio-temporal semantics location prediction) to enhance the location prediction performance.Firstly, the similar suspect groups of the target suspects are extracted using the distributed proximity of the suspects in different semantic periods and semantic positions.Then, their mobility data are applied to estimate transition frequencies and temporal visiting probabilities for unobserved locations based on a KDE(kernel density estimating) smoothing method. Finally, by a Bayesian-based formula, the spatio-temporal prediction for the individual suspect can be realized. In the experiments with the location recording data set consisting of 158 suspects and their 17 539 location records from January to June 2013 in W city, SSLP model outperforms baseline algorithms by 40%-50%, validating its adaptability for data sparsity problem.
Existing studies have failed to capture the complex social location transition patterns of suspects and lack the capacity to address the issue of data sparsity. Therefore, we propose a location prediction model called SSLP(spatio-temporal semantics location prediction) to enhance the location prediction performance.Firstly, the similar suspect groups of the target suspects are extracted using the distributed proximity of the suspects in different semantic periods and semantic positions.Then, their mobility data are applied to estimate transition frequencies and temporal visiting probabilities for unobserved locations based on a KDE(kernel density estimating) smoothing method. Finally, by a Bayesian-based formula, the spatio-temporal prediction for the individual suspect can be realized. In the experiments with the location recording data set consisting of 158 suspects and their 17 539 location records from January to June 2013 in W city, SSLP model outperforms baseline algorithms by 40%-50%, validating its adaptability for data sparsity problem.
引文
[1]Shiode S,Shiode N,Block R,et al.Space-Time Characteristics of Micro-scale Crime Occurrences:An Application of a Network-Based Space-Time Search Window Technique for Crime Incidents in Chicago[J].International Journal of Geographical Information Science,2015,29(5):697-719
[2]Hammond L.Geographical Profiling in a Novel Context:Prioritising the Search for New Zealand Sex Offenders[J].Psychology,Crime and Law,2014,20(4):358-371
[3]Chen Naichang,Shi Wei,Song Dawei.Prediction of Series Criminals:An Approach Based on Modeling[C].International Conference on Computational and Information Sciences,Chengdu,China,2010
[4]Qian Cheng,WangYubo,Cao Jinde,et al.WeightedTraffic-Network-Based Geographic Profiling for Serial Crime Location Prediction[J].EPL,2011,DOI:10.1209/0295-5075/93/68006
[5]Mohler G O,Short M B.Geographic Profiling from Kinetic Models of Criminal Behavior[J].SIAM JAppl Math,2012,72(1):163-180
[6]Martineau M,Beauregard E.Journey to Murder:Examining the Correlates of Criminal Mobility in Sexual Homicide[J].Police Practice and Research,2016,17(1):68-83
[7]Song C,Koren T,Wang P,et al.Modelling the Scaling Properties of Human Mobility[J].Nat Phys,2010,6(10):818-823
[8]Sadilek A,Kautz H,Bigham J P.Finding Your Friends and Following Them to Where You Are[C].The 5th ACM International Conference on Web Search and Data Mining,Seattle,USA,2012
[9]Gogate V,Dechter R,Bidyuk B.Modeling Transportation Routines Using Hybrid Dynamic Mixed Networks[C].The 21st Conference on Uncertainty in Artificial Intelligence,Edinburgh,Scotland,2005
[10]Wang Y,Yuan N J,Lian D,et al.Regularity and Conformity:Location Prediction Using Heterogeneous Mobility Data[C].The 21th ACM SIGKDDInternational Conference on Knowledge Discovery and Data Mining,Sydney,Australia,2015
[11]Xiao Xiangye,Zheng Yu,Luo Qiong,et al.Finding Similar Users Using Category-Based Location History[C].The 18th ACM SIGSPATIAL Conference on Advances in Geographical Information Systems,San Jose,USA,2010
[12]Levine N,Block R.Bayesian Journey-to-Crime Estimation:An Improvement in Geographic Profiling Methodology[J].Professional Geographer,2011,63(2):213-229
[13]Tayebi M A,Glasser U,Ester M,et al.Personalized Crime Location Prediction[J].Eur J Appl Math,2016,27(3):422-450
[14]Gao Huiji,Tang Jiliang,Liu Huan.Mobile Location Prediction in Spatio-Temporal Context[C].Mobile Data Challenge 2012(by Nokia)Workshop,Newcastle,Britain,2012
[15]Yuan Quan,Cong Gao,Ma Zongyang,et al.Who,Where,When and What:Discover Spatio-Temporal Topics for Twitter Users[C].The 19th ACM SIG-KDD International Conference on Knowledge Discovery and Data Mining,Sydney,Australia,2013
[2]Hammond L.Geographical Profiling in a Novel Context:Prioritising the Search for New Zealand Sex Offenders[J].Psychology,Crime and Law,2014,20(4):358-371
[3]Chen Naichang,Shi Wei,Song Dawei.Prediction of Series Criminals:An Approach Based on Modeling[C].International Conference on Computational and Information Sciences,Chengdu,China,2010
[4]Qian Cheng,WangYubo,Cao Jinde,et al.WeightedTraffic-Network-Based Geographic Profiling for Serial Crime Location Prediction[J].EPL,2011,DOI:10.1209/0295-5075/93/68006
[5]Mohler G O,Short M B.Geographic Profiling from Kinetic Models of Criminal Behavior[J].SIAM JAppl Math,2012,72(1):163-180
[6]Martineau M,Beauregard E.Journey to Murder:Examining the Correlates of Criminal Mobility in Sexual Homicide[J].Police Practice and Research,2016,17(1):68-83
[7]Song C,Koren T,Wang P,et al.Modelling the Scaling Properties of Human Mobility[J].Nat Phys,2010,6(10):818-823
[8]Sadilek A,Kautz H,Bigham J P.Finding Your Friends and Following Them to Where You Are[C].The 5th ACM International Conference on Web Search and Data Mining,Seattle,USA,2012
[9]Gogate V,Dechter R,Bidyuk B.Modeling Transportation Routines Using Hybrid Dynamic Mixed Networks[C].The 21st Conference on Uncertainty in Artificial Intelligence,Edinburgh,Scotland,2005
[10]Wang Y,Yuan N J,Lian D,et al.Regularity and Conformity:Location Prediction Using Heterogeneous Mobility Data[C].The 21th ACM SIGKDDInternational Conference on Knowledge Discovery and Data Mining,Sydney,Australia,2015
[11]Xiao Xiangye,Zheng Yu,Luo Qiong,et al.Finding Similar Users Using Category-Based Location History[C].The 18th ACM SIGSPATIAL Conference on Advances in Geographical Information Systems,San Jose,USA,2010
[12]Levine N,Block R.Bayesian Journey-to-Crime Estimation:An Improvement in Geographic Profiling Methodology[J].Professional Geographer,2011,63(2):213-229
[13]Tayebi M A,Glasser U,Ester M,et al.Personalized Crime Location Prediction[J].Eur J Appl Math,2016,27(3):422-450
[14]Gao Huiji,Tang Jiliang,Liu Huan.Mobile Location Prediction in Spatio-Temporal Context[C].Mobile Data Challenge 2012(by Nokia)Workshop,Newcastle,Britain,2012
[15]Yuan Quan,Cong Gao,Ma Zongyang,et al.Who,Where,When and What:Discover Spatio-Temporal Topics for Twitter Users[C].The 19th ACM SIG-KDD International Conference on Knowledge Discovery and Data Mining,Sydney,Australia,2013