WO2016127507A1 - 一种分析用户轨迹的方法及装置 - Google Patents

一种分析用户轨迹的方法及装置 Download PDF

Info

Publication number
WO2016127507A1
WO2016127507A1 PCT/CN2015/078220 CN2015078220W WO2016127507A1 WO 2016127507 A1 WO2016127507 A1 WO 2016127507A1 CN 2015078220 W CN2015078220 W CN 2015078220W WO 2016127507 A1 WO2016127507 A1 WO 2016127507A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
cell
time
signaling data
given user
Prior art date
Application number
PCT/CN2015/078220
Other languages
English (en)
French (fr)
Inventor
杨魁
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2016127507A1 publication Critical patent/WO2016127507A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04WWIRELESS COMMUNICATION NETWORKS
    • H04W24/00Supervisory, monitoring or testing arrangements
    • H04W24/02Arrangements for optimising operational condition
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16ZINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
    • G16Z99/00Subject matter not provided for in other main groups of this subclass

Definitions

  • the present invention relates to the field of mobile communication and data mining technologies, and in particular, to a method and apparatus for analyzing user trajectories.
  • a user location analysis system based on signaling data is a method for mining user trajectory features and predicting user location through signaling data: first, collecting user tools by installing terminal tools on user mobile devices A certain amount of trajectory data is abstracted, and the PrefixSpan mining algorithm is called to obtain and model the motion pattern based on the user position information, and the pattern tree is constructed (the pattern tree contains all motion modes and their probability of adopting different starting points and end points) At the same time, the user's online motion situation is analyzed to obtain a set of motion patterns organized according to the starting point and the end point position data, and the motion mode set result and the mined motion pattern are matched and searched to predict the user position.
  • the user's real-time location data such as time field information and location field information is collected through the wireless communication information of the mobile communication network user, and the acquired user mobile data is cleaned and processed, and a processed user mobile data information is output and combined.
  • the transition probability matrix obtained by the user's historical mobile behavior analysis the Markov model is constructed for analysis and calculation, thereby predicting the possibility of accessing each location, and making the location prediction of the user's maximum possible access.
  • the two methods have the following disadvantages: the analysis method used for the signaling data is complicated and the calculation amount is large, and the optimization processing of the historical data is lacking.
  • An object of the embodiments of the present invention is to provide a method and apparatus for analyzing a user trajectory, which can analyze signaling data simply and efficiently.
  • an embodiment of the present invention provides a method for analyzing a user trajectory, the method comprising:
  • the motion trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, an earliest time and a latest time when the user enters each cell, and the user is in each cell.
  • the trajectory of a given user is determined based on the unique identification of the given user and the motion trajectory parameters of the user.
  • the step of cleaning the collected signaling data of the user includes:
  • the demodulated processing of the corrected user's signaling data is performed.
  • the step of obtaining the motion trajectory parameter of the user according to the cleaned signaling data includes:
  • the daily trajectory parameters of the user in the preset number of days are obtained, and the daily trajectory parameters include the unique identifier of the user, the unique identifier of each cell, the earliest time and the latest time when the user enters each cell every day, and The minimum dwell time and maximum dwell time of the user in each cell per day;
  • the user's daily trajectory parameters are weighted in the preset number of days to obtain the user's motion trajectory parameters, wherein the farther from the current time, the preset weight of the trajectory parameter is smaller.
  • the steps of determining the trajectory of the given user include:
  • the average time that the given user enters the cell is within the second preset time period, further determining whether the dwell time of the given user in the cell covers the third preset time period;
  • the staying time of the given user in the cell covers the third preset time period, it is determined that the cell is the residence of the given user.
  • the method further includes: after the step of acquiring the motion trajectory data of the user, the shortest dwell time of the given user in the cell is the motion track data corresponding to the first preset time, according to the unique identifier of the given user, the method further includes:
  • the average time that the given user enters the cell is within the fourth preset time period, it is further determined whether the dwell time of the given user in the cell covers the fifth preset time period and the sixth preset time period;
  • the staying time of the given user in the cell covers the fifth preset time period and the sixth preset time period, it is determined that the cell is the working location of the given user.
  • the steps of determining the trajectory of the given user include:
  • the time when the given user leaves the current cell is obtained by scanning the motion track parameter of the user according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell.
  • the steps of determining the trajectory of the given user include:
  • the unique identifier of a given user the unique identifier of the cell in which the given user is currently located, and the time of entering the cell, by scanning the motion trajectory parameter of the user, the unique identifier of the next most likely cell of the given user is obtained and The time spent in the cell.
  • Embodiments of the present invention also provide an apparatus for analyzing a user trajectory, the apparatus comprising:
  • the cleaning module is configured to clean the signaling data of the collected user
  • Obtaining a module configured to obtain a motion trajectory parameter of the user according to the cleaned signaling data, where the motion trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, an earliest time and a latest time when the user enters each cell, and a user Minimum dwell time and longest dwell time in each cell;
  • a determination module is configured to determine a trajectory of a given user based on a unique identification of the given user and a motion trajectory parameter of the user.
  • the cleaning module includes:
  • a completion unit configured to complete the missing signaling data in the collected signaling data of the user
  • the determining unit is configured to determine, according to the signaling data rule, whether the signaling data of the completed user is error signaling data, and trigger the correcting unit when the completed signaling data of the user is error signaling data;
  • a correcting unit configured to correct the error signaling data according to a trigger of the determining unit
  • the deduplication unit is configured to perform deduplication processing on the corrected user's signaling data.
  • the acquisition module includes:
  • the first unit is configured to obtain the daily signaling data of the user from the cleaned signaling data, and sort the acquired daily signaling data according to the time sequence of the user entering different cells;
  • the second unit is configured to obtain, according to the sorted signaling data, a daily trajectory parameter of the user within a preset number of days, and the daily trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, and the earliest user enters each cell every day. Time and latest time and the minimum dwell time and maximum dwell time of the user in each cell per day;
  • the third unit is configured to perform weighting processing on the trajectory parameters of the user in the preset number of days according to the preset weight, to obtain a motion trajectory parameter of the user, wherein the farther from the current time, the preset weight of the trajectory parameter is smaller.
  • the determining module comprises:
  • the fourth unit is configured to: according to the unique identifier of the given user, scan the motion trajectory parameter of the user, and obtain the motion trajectory data corresponding to the shortest stay time of the given user in the cell as the first preset time;
  • the fifth unit is configured to determine whether the average time that the given user enters the cell is within the second preset time period, and trigger when the average time that the given user enters the cell is within the second preset time period Unit 6;
  • the sixth unit is configured to further determine, according to the triggering of the fifth unit, whether the staying time of the given user in the cell covers a third preset time period, and cover the third pre-preservation time when the given user stays in the cell When the time period is set, it is determined that the cell is the residence of the given user.
  • the determining module further includes:
  • the seventh unit is configured to determine whether the average time that the given user enters the cell is within a fourth preset time period, and trigger when the average time that the given user enters the cell is within the fourth preset time period.
  • the eighth unit is configured to further determine, according to the triggering of the seventh unit, whether the staying time of the given user in the cell covers the fifth preset time period and the sixth preset time period, and when the given user is in the cell When the stay time covers the fifth preset time period and the sixth preset time period, it is determined that the cell is the work place of the given user.
  • the determining module further includes:
  • the ninth unit is configured to: according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell, the time of the given user leaving the current cell is obtained by scanning the motion track parameter of the user. .
  • the determining module further includes:
  • the tenth unit is set to obtain the next most likely occurrence of the given user by scanning the trajectory parameter of the user according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell. The unique identity of the cell and the time spent in the cell.
  • the collected user's signaling data is cleaned, and according to the cleaned signaling data, the user's motion trajectory parameters are obtained, and then according to the unique identifier of the given user.
  • the trajectory of the given user is determined from the user's motion trajectory parameters, thereby simply and efficiently analyzing the signaling data to determine the trajectory of the given user.
  • FIG. 1 is a flowchart of a method for analyzing a user track in an embodiment of the present invention
  • FIG. 2 is a flowchart of specific steps of step 11 in FIG. 1 according to an embodiment of the present invention
  • step 12 in FIG. 1 is a flowchart of specific steps of step 12 in FIG. 1 according to an embodiment of the present invention
  • FIG. 4 is a schematic structural diagram of an apparatus for analyzing a user trajectory according to an embodiment of the present invention.
  • the present invention is directed to the problem of analyzing signaling data in the prior art, and provides a method and apparatus for analyzing user trajectories, which can analyze signaling data simply and efficiently.
  • an embodiment of the present invention provides a method for analyzing a user trajectory, the method comprising:
  • step 11 the collected signaling data of the user is cleaned.
  • cleaning the collected signaling data of the user may improve the correctness of subsequent data analysis.
  • Step 12 Obtain a motion trajectory parameter of the user according to the cleaned signaling data, where the motion trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, an earliest time and a latest time when the user enters each cell, and the user is in each The shortest dwell time and the longest dwell time in each cell.
  • user trajectory analysis is performed based on the cleaned signaling data.
  • the PrefixSpan algorithm can be used to analyze the user's signaling data, and the user's place of residence, work place, and user habitual motion trajectory are mined.
  • step 13 the trajectory of the given user is determined according to the unique identifier of the given user and the motion trajectory parameters of the user.
  • the trajectory of the given user may be determined from the motion trajectory parameters of the user obtained in step 12 according to the unique identifier of the given user.
  • the user behavior trajectory is analyzed by using the PrefixSpan algorithm, which is simple and efficient, so as to implement simple and efficient analysis of signaling data.
  • the use of the PrefixSpan algorithm to analyze the user behavior trajectory is common knowledge to those skilled in the art, and details are not described herein.
  • step 11 are:
  • Step 21 Completing the missing signaling data in the collected signaling data of the user.
  • the missing signaling data in the collected signaling data of the user may be complemented by a common manner such as linearity and mean value. In the process of completion, if there is any signaling data that cannot be completed, the signaling data is discarded.
  • Step 22 Determine, according to the signaling data rule, whether the signaling data of the completed user is error signaling data.
  • step 23 if it is error signaling data, the error signaling data is corrected.
  • the signaling data rule it is necessary to determine the error signaling data in the completed signaling data according to the signaling data rule, and correct the erroneous signaling data. In the process of correction, if there is uncorrectable signaling data, the signaling data is discarded.
  • Step 24 Perform deduplication processing on the corrected signaling data of the user.
  • Step 31 Obtain the daily signaling data of the user from the cleaned signaling data, and sort the acquired daily signaling data according to the chronological order in which the user enters different cells.
  • Step 32 According to the sorted signaling data, obtain a daily trajectory parameter of the user within a preset number of days, and the daily trajectory parameters include a unique identifier of the user, a unique identifier of each cell, and an earliest time and maximum time each user enters each cell. Late time and the minimum dwell time and maximum dwell time of the user in each cell per day.
  • all signaling data of the user is obtained from the cleaned signaling data, and the format of the obtained signaling data may be ⁇ userID, [cellID1, entertime, residenceTime], [cellID2, Entertime, residenceTime], ... ⁇ , where: userID is the unique identifier of a user; cellID1 is the unique identifier of the user entering the cell; entertime is the entry time of the user userID entering the cellcellID1; and the retentionTime is the dwell time of the user userID entering the cellcellID1.
  • the user signaling data is sorted according to the order in which the user enters different cells in time, thereby obtaining the trajectory of the user every day.
  • the signaling data of each of the above days may be processed to obtain other trajectory parameters (for example, an entry time of the user entering the cell ID11 on average in a day, etc.).
  • the data in the following format can be obtained through analysis: ⁇ userID, [cellID1, firstEntertime, endEntertime, avgEntertime, minResidenceTime, maxResidenceTime], [cellID2, firstEntertime, endEntertime, minResidenceTime, maxResidenceTime, avgResidenceTime], Probability ⁇ , where: userID The unique ID of a user; cellID1 is the unique identifier of the user entering the cell; firstEntertime is the entry time of the user's userID into the cellID1 of the cell in the first day; endEntertime is the entry time of the user's userID to enter the cellID1 of the cell at the latest; avgEntertime is the user userID The average entry time of the cell ID1 entering the cell userID is the shortest time; the
  • Step 33 Perform weighting processing on the trajectory parameters of the user within the preset number of days according to the preset weight, to obtain a motion trajectory parameter of the user, wherein the preset weight of the trajectory parameter that is further away from the current time is smaller.
  • the data extracted by the user trajectory analysis may be updated in real time according to the current location of the user, and the latest (ie, closest to the current time) data is given a higher weight, and the historical data information is weakened. This provides up-to-date and reliable information for the analysis of user trajectories.
  • the specific preset number of days can be adjusted according to the needs of the business.
  • take 35 days as an example for explanation.
  • the 35-day data is divided into 5 weeks, and the earliest entry time, early entry time, minimum entry time, longest stay time, etc. of the user UserID entering the cell ID1 per week are calculated, and the obtained weekly data is ⁇ userID, [cellID1, firstEntertime, endEntertime, avgEntertime, minResidenceTime, maxResidenceTime], weights the five-week data.
  • the calculation weighting formula can be: 5* recent first week data +4* recent second week data +3*last third week data+2*last fourth week data+last fifth week data/(5+4+3 +2+1), according to this formula, the historical data is weakened, and the latest data is given a higher weight.
  • the earliest entry time, the latest entry time, the shortest dwell time, the longest dwell time, etc. of the user userID in the cell_cell1 can be obtained. It can be understood that, in the embodiment of the present invention, the specific formula of the weighting process is not limited, as long as the latest data is given a higher weight and the historical data is weakened.
  • the trajectory parameter of the user is dynamically adjusted according to the real-time location information of the user, and the accuracy of determining the trajectory of the user is improved.
  • the specific step of the foregoing step 13 may be: according to the unique identifier of the given user, by scanning the motion trajectory parameter of the user, obtaining the minimum stay time of the given user in the cell is the first a motion trajectory data corresponding to a preset time; determining whether an average time for the given user to enter the cell is within a second preset time period; if an average time for the given user to enter the cell is within a second preset time period And determining whether the dwell time of the given user in the cell covers a third preset time period; if the dwell time of the given user covers the third preset time period, determining that the cell is the given time The user's place of residence.
  • the first preset time may be set to 5 hours
  • the second preset time period is set to 18 points to 24 points
  • the third preset time period is set to 1 am to 6 am. It can be understood that, in the embodiment of the present invention, the specific values of the first preset time, the second preset time period, and the third preset time period are not limited.
  • the above analysis data is scanned to obtain data of the user's shortest stay time of about 5 hours in a certain cell, and then the average entry time into the cell is obtained, if the average entry time is entered.
  • the time is from 18 o'clock to 24 o'clock, and the dwell time can cover the cell from 1 am to 6 o'clock in the morning, and the cell can be judged as the residence of the user.
  • the method further includes: determining whether the average time that the given user enters the cell is within a fourth preset time period; if the average time that the given user enters the cell is within the fourth preset time period, Further Determining whether the dwell time of the given user in the cell covers the fifth preset time period and the sixth preset time period; if the dwell time of the given user in the cell covers the fifth preset time period and the sixth preset The time period determines that the cell is the working place of the given user.
  • the fourth preset time period may be set to 7:00 to 9:00, the fifth preset time period is set to 9:30 to 11:30, and the sixth preset time period is set to 14 Half past half to 17:30. It can be understood that the specific values of the fourth preset time period, the fifth preset time period, and the sixth preset time period are not limited in the embodiment of the present invention.
  • the above analysis data is scanned to obtain data of the user's shortest stay time of about 5 hours in a certain cell, and then the average entry time into the cell is obtained, and the average entry time is obtained. It is 7 to 9 o'clock, and the dwell time can cover the cell from 9:30 to 11:30 and 14:30 to 17:30, and the cell can be judged as the user's work place.
  • the specific step of the foregoing step 13 may further be: scanning the user according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell.
  • the motion trajectory parameter gives the time when the given user leaves the current cell.
  • the analysis data may be scanned according to the user ID of the given user, the ID of the current cell of the user, and the time of entering the cell, and the average stay time of the user in the cell is obtained, and then the user enters according to the user.
  • the time of the cell can get the time when the user may leave.
  • the specific step of the foregoing step 13 may further be: scanning the user according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell.
  • the motion trajectory parameter obtains the unique identifier of the next most likely cell of the given user and the time spent in the cell.
  • the analysis data may be scanned according to the user ID of the given user, the ID of the cell where the user is currently located, and the time of entering the cell, to obtain which cell the user is most likely to appear next.
  • the probability of occurrence in the next cell and the dwell time in the next cell can be obtained from the Probability field in the analysis data.
  • an embodiment of the present invention provides an apparatus for analyzing a user trajectory, and the apparatus includes:
  • the cleaning module 41 is configured to clean the collected signaling data of the user
  • the obtaining module 42 is configured to obtain a motion trajectory parameter of the user according to the cleaned signaling data, where the motion trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, an earliest time and a latest time when the user enters each cell, and The minimum dwell time and the longest dwell time of the user in each cell;
  • the determining module 43 is arranged to determine the trajectory of the given user based on the unique identification of the given user and the motion trajectory parameters of the user.
  • the cleaning module 41 includes:
  • a completion unit configured to complete the missing signaling data in the collected signaling data of the user
  • the determining unit is configured to determine, according to the signaling data rule, whether the signaling data of the completed user is error signaling data, and trigger the correcting unit when the completed signaling data of the user is error signaling data;
  • a correcting unit configured to correct the error signaling data according to a trigger of the determining unit
  • the deduplication unit is configured to perform deduplication processing on the corrected user's signaling data.
  • the obtaining module 42 includes:
  • the first unit is configured to obtain the daily signaling data of the user from the cleaned signaling data, and sort the acquired daily signaling data according to the time sequence of the user entering different cells;
  • the second unit is configured to obtain, according to the sorted signaling data, a daily trajectory parameter of the user within a preset number of days, and the daily trajectory parameter includes a unique identifier of the user, a unique identifier of each cell, and the earliest user enters each cell every day. Time and latest time and the minimum dwell time and maximum dwell time of the user in each cell per day;
  • the third unit is configured to perform weighting processing on the trajectory parameters of the user in the preset number of days according to the preset weight, to obtain a motion trajectory parameter of the user, wherein the farther from the current time, the preset weight of the trajectory parameter is smaller.
  • the determining module 43 includes:
  • the fourth unit is configured to: according to the unique identifier of the given user, scan the motion trajectory parameter of the user, and obtain the motion trajectory data corresponding to the shortest stay time of the given user in the cell as the first preset time;
  • the fifth unit is configured to determine whether the average time that the given user enters the cell is within the second preset time period, and trigger when the average time that the given user enters the cell is within the second preset time period Unit 6;
  • the sixth unit is configured to further determine, according to the triggering of the fifth unit, whether the staying time of the given user in the cell covers a third preset time period, and cover the third pre-preservation time when the given user stays in the cell When the time period is set, it is determined that the cell is the residence of the given user.
  • the determining module 43 further includes:
  • the seventh unit is configured to determine whether the average time that the given user enters the cell is within a fourth preset time period, and trigger when the average time that the given user enters the cell is within the fourth preset time period.
  • the eighth unit is configured to further determine, according to the triggering of the seventh unit, whether the staying time of the given user in the cell covers the fifth preset time period and the sixth preset time period, and when the given user is in the cell When the stay time covers the fifth preset time period and the sixth preset time period, it is determined that the cell is the work place of the given user.
  • the determining module 43 further includes:
  • the ninth unit is configured to: according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell, the time of the given user leaving the current cell is obtained by scanning the motion track parameter of the user. .
  • the determining module 43 further includes:
  • the tenth unit is set to obtain the next most likely occurrence of the given user by scanning the trajectory parameter of the user according to the unique identifier of the given user, the unique identifier of the cell in which the given user is currently located, and the time of entering the cell. The unique identity of the cell and the time spent in the cell.
  • the device for analyzing the user track provided by the embodiment of the present invention is the device applying the above method, that is, all the embodiments of the foregoing method are applicable to the device, and all of the same or similar beneficial effects can be achieved.
  • the collected signaling data of the user is cleaned, and the motion trajectory parameter of the user is obtained according to the cleaned signaling data, and then the user's unique identifier is obtained according to the unique identifier of the user.
  • the trajectory of the given user is determined in the motion trajectory parameter, so that the signaling data is analyzed simply and efficiently, and the trajectory of the given user is determined.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本发明提供了一种分析用户轨迹的方法及装置,该方法包括:对采集到的用户的信令数据进行清洗;根据清洗后的信令数据,得到用户的运动轨迹参数,运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹,本发明的方法能简单、高效地分析信令数据。

Description

一种分析用户轨迹的方法及装置 技术领域
本发明涉及移动通信及数据挖掘技术领域,特别涉及一种分析用户轨迹的方法及装置。
背景技术
分析用户轨迹可以预测用户位置,挖掘用户移动的行为特征。目前基于信令数据的用户位置分析系统,尤其是在移动通信领域中通过信令数据来挖掘用户轨迹特征、预测用户位置的方法有:第一种,通过用户移动设备上安装终端工具,收集用户一定量的轨迹数据,对其进行抽象化处理,调用PrefixSpan挖掘算法得到基于用户位置信息的运动模式并进行建模,构造模式树(模式树包含所有运动模式及其采用不同起点和终点的概率),同时分析用户在线运动情况得到按照起点和终点位置数据进行组织的运动模式集,将运动模式集结果和挖掘出的运动模式进行匹配和查找来预测用户位置。第二种,通过移动通信网用户无线上网信息采集用户实时位置数据如时间字段信息、地点字段信息,同时对获取到的用户移动数据进行清洗处理,输出一个已处理的用户移动数据信息,并结合根据用户的历史移动行为分析得到的转移概率矩阵,构造马尔科夫模型进行分析计算,从而预测其访问各个地点的可能性,做出用户最大可能访问的地点预测。但这两种方法存在如下缺点:对信令数据采用的分析方法复杂且计算量大,同时缺少对历史数据的优化处理。
发明内容
本发明实施例的目的在于提供一种分析用户轨迹的方法及装置,能简单、高效地分析信令数据。
为了达到上述目的,本发明的实施例提供了一种分析用户轨迹的方法,该方法包括:
对采集到的用户的信令数据进行清洗;
根据清洗后的信令数据,得到用户的运动轨迹参数,运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;
根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹。
其中,对采集到的用户的信令数据进行清洗的步骤包括:
对采集到的用户的信令数据中的残缺信令数据进行补全;
根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据;
若是错误信令数据,则纠正该错误信令数据;
对纠正后的用户的信令数据进行去重处理。
其中,根据清洗后的信令数据,得到用户的运动轨迹参数的步骤包括:
从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序;
根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间;
根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
其中,根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;
判断该给定用户进入该小区的平均时间是否在第二预设时间段内;
若该给定用户进入该小区的平均时间在第二预设时间段内,则进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段;
若该给定用户在该小区的停留时间覆盖第三预设时间段,则确定该小区为该给定用户的居住地。
其中,根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据的步骤之后,方法还包括:
判断该给定用户进入该小区的平均时间是否在第四预设时间段内;
若该给定用户进入该小区的平均时间在第四预设时间段内,则进一步判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段;
若该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段,则确定该小区为该给定用户的工作地。
其中,根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
其中,根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
本发明的实施例还提供了一种分析用户轨迹的装置,该装置包括:
清洗模块,设置为对采集到的用户的信令数据进行清洗;
获得模块,设置为根据清洗后的信令数据,得到用户的运动轨迹参数,运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;
确定模块,设置为根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹。
其中,清洗模块包括:
补全单元,设置为对采集到的用户的信令数据中的残缺信令数据进行补全;
判断单元,设置为根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据,并当补全后的用户的信令数据是错误信令数据时,触发纠正单元;
纠正单元,设置为根据判断单元的触发,纠正该错误信令数据;
去重单元,设置为对纠正后的用户的信令数据进行去重处理。
其中,获得模块包括:
第一单元,设置为从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序;
第二单元,设置为根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间;
第三单元,设置为根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
其中,确定模块包括:
第四单元,设置为根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;
第五单元,设置为判断该给定用户进入该小区的平均时间是否在第二预设时间段内,并当该给定用户进入该小区的平均时间在第二预设时间段内时,触发第六单元;
第六单元,设置为根据第五单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段,并当该给定用户在该小区的停留时间覆盖第三预设时间段时,确定该小区为该给定用户的居住地。
其中,确定模块还包括:
第七单元,设置为判断该给定用户进入该小区的平均时间是否在第四预设时间段内,并当该给定用户进入该小区的平均时间在第四预设时间段内时,触发第八单元;
第八单元,设置为根据第七单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段,并当该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段时,确定该小区为该给定用户的工作地。
其中,确定模块还包括:
第九单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
其中,确定模块还包括:
第十单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
本发明的上述方案至少包括以下有益效果:
在本发明的实施例的分析用户轨迹的方法中,对采集到的用户的信令数据进行清洗,并根据清洗后的信令数据,得到用户的运动轨迹参数,再根据给定用户的唯一标识从用户的运动轨迹参数中确定出该给定用户的轨迹,从而简单、高效地分析信令数据,确定出给定用户的轨迹。
附图说明
图1为本发明实施例中分析用户轨迹的方法的流程图;
图2为本发明实施例中图1中的步骤11的具体步骤流程图;
图3为本发明实施例中图1中的步骤12的具体步骤流程图;
图4为本发明实施例中分析用户轨迹的装置的结构示意图。
具体实施方式
为使本发明要解决的技术问题、技术方案和优点更加清楚,下面将结合附图及具体实施例进行详细描述。
本发明针对现有技术中分析信令数据较复杂的问题,提供了一种分析用户轨迹的方法及装置,能简单、高效地分析信令数据。
如图1所示,本发明的实施例提供了一种分析用户轨迹的方法,该方法包括:
步骤11,对采集到的用户的信令数据进行清洗。
在本发明的具体实施例中,对采集到的用户的信令数据进行清洗可以提高后续数据分析的正确性。
步骤12,根据清洗后的信令数据,得到用户的运动轨迹参数,运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间。
在本发明的具体实施例中,根据清洗后的信令数据,进行用户轨迹分析。具体地,可以使用PrefixSpan算法对用户的信令数据进行分析,挖掘出用户的居住地,工作地,及用户习惯性的运动轨迹等。
步骤13,根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹。
在本发明的具体实施例中,可根据给定用户的唯一标识,从步骤12中得到的用户的运动轨迹参数中确定出该给定用户的轨迹。
在本发明的具体实施例中,利用PrefixSpan算法分析用户行为轨迹,该分析方法简单高效,以便实现简单、高效地分析信令数据。其中利用PrefixSpan算法分析用户行为轨迹对于本领域的技术人员来说是公知常识,在此不再赘述。
其中,在本发明的上述实施例中,如图2所示,步骤11的具体步骤为:
步骤21,对采集到的用户的信令数据中的残缺信令数据进行补全。
在本发明的具体实施例中,可以通过线性、均值等常见的方式对采集到的用户的信令数据中的残缺信令数据进行补全。在补全的过程中,若出现无法补全的信令数据,则舍弃该信令数据。
步骤22,根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据。
步骤23,若是错误信令数据,则纠正该错误信令数据。
在本发明的具体实施例中,需要根据信令数据规则判断出补全后的信令数据中的错误信令数据,并纠正这些错误的信令数据。在纠正的过程中,若出现无法纠正的信令数据,则舍弃该信令数据。
步骤24,对纠正后的用户的信令数据进行去重处理。
在本发明的具体实施例中,需要去除纠正后的信令数据中的重复的信令数据,以便提高后续数据分析的正确性。
其中,在本发明的上述实施例中,如图3所示,上述步骤12的具体步骤为:
步骤31,从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序。
步骤32,根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间。
在本发明的具体实施例中,从清洗后的信令数据中获取用户每一天的所有信令数据,得到的信令数据的格式可以是{userID,[cellID1,entertime,residenceTime],[cellID2,entertime,residenceTime],…},其中:userID为某用户的唯一标识;cellID1为用户进入此小区的唯一标识;entertime为用户userID进入小区cellID1的进入时间;residenceTime为用户userID进入小区cellID1的停留时间。获得每一天的所有信令数据后,会按照用户进入不同小区在时间的先后顺序对用户信令数据排序,从而得到用户每一天的轨迹。进一步地,还可以对上述每一天的信令数据进行处理得到其它的轨迹参数(例如用户一天中平均进入小区cellID1的进入时间等)。具体地,可以经过分析处理得到如下格式的数据:{userID,[cellID1,firstEntertime,endEntertime,avgEntertime,minResidenceTime,maxResidenceTime],[cellID2,firstEntertime,endEntertime,minResidenceTime,maxResidenceTime,avgResidenceTime],Probability},其中:userID为某用户的唯一标识;cellID1为用户进入此小区的唯一标识;firstEntertime为用户userID一天中最早进入小区cellID1的进入时间;endEntertime为用户userID一天中最晚进入小区cellID1的进入时间;avgEntertime为用户userID一天中平均进入小区cellID1的进入时间;minResidence为用户userID进入小区cellID1的停留最短时间;maxResidence为用户userID进入小区cellID1的停留最长时间;avgResidenceTime为用户userID进入小区cellID1的平均停留时间;Probability为为用户userID已经小区cellID1后,可能会进入小区cellID2的概率。
步骤33,根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
在本发明的具体实施例中,可以根据用户当前的位置,实时更新用户轨迹分析挖掘出来的数据,且对最新(即离当前时刻最近)的数据赋予更高的权重,弱化历史数据的信息,从而为用户轨迹的分析提供最新的可靠信息。
在本发明的具体实施例中,具体的预设天数可根据业务的需求进行调整。接下来以35天为例进行说明。在获得用户UserID在小区cellID1的最近35天的数据后。将35天数据分为5周,计算出用户UserID每周进入小区cellID1的最早进入时间,早进入时间,最晚进入时间、最短停留时间,最长停留时间等,得到的每周的数据为{userID,[cellID1,firstEntertime,endEntertime,avgEntertime,minResidenceTime,maxResidenceTime],对五周数据进行加权处理。计算加权公式可以为:5*最近第一周数据+4*最近第二周数据+3*最近第三周数据+2*最近第四周数据+最近第五周数据/(5+4+3+2+1),根据此公式对历史数据进行了弱化,对最新数据赋予更高的权重。可得到用户userID在小区cellID1最终的最早进入时间,最晚进入时间、最短停留时间,最长停留时间等。可以理解的是,在本发明的实施例中,并不限定加权处理的具体公式,只要对最新的数据赋予更高的权重,弱化历史数据即可。
在本发明的具体实施例中,根据用户的实时位置信息动态调整用户的轨迹参数,提高了确定用户轨迹的精度。
其中,在本发明的具体实施例中,上述步骤13的具体步骤可以为:根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;判断该给定用户进入该小区的平均时间是否在第二预设时间段内;若该给定用户进入该小区的平均时间在第二预设时间段内,则进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段;若该给定用户在该小区的停留时间覆盖第三预设时间段,则确定该小区为该给定用户的居住地。
在本发明的具体实施例中,可以将第一预设时间设为5小时,第二预设时间段设为18点至24点,第三预设时间段设为凌晨1点至凌晨6点,可以理解的是,在本发明的实施例中,并不限定第一预设时间、第二预设时间段以及第三预设时间段的具体数值。
在本发明的具体实施例中,可根据给定用户放入userID,扫描上述分析数据获取此用户在某小区最短停留时间大约5小时的数据,再获取进入此小区的平均进入时间,若平均进入时间为18点至24点,且停留时间可覆盖凌晨1点凌至晨6点的小区,可判断此小区为用户的居住地。
其中,在本发明的具体实施例中,根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据的步骤之后,方法还包括:判断该给定用户进入该小区的平均时间是否在第四预设时间段内;若该给定用户进入该小区的平均时间在第四预设时间段内,则进一步 判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段;若该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段,则确定该小区为该给定用户的工作地。
在本发明的具体实施例中,可将第四预设时间段设为7点至9点,第五预设时间段设为9点半至11点半,第六预设时间段设为14点半至17点半。可以理解的是,在本发明的实施例中并不限定第四预设时间段、第五预设时间段以及第六预设时间段的具体数值。
在本发明的具体实施例中,可根据给定用户的userID,扫描上述分析数据获取此用户在某小区最短停留时间大约5小时的数据,再获取进入此小区的平均进入时间,若平均进入时间为7点至9点,且停留时间可覆盖9点半至11点半及14点半至17点半的小区,可判断此小区为用户的工作地。
其中,在本发明的上述实施例中,上述步骤13的具体步骤还可以为:根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
在本发明的具体实施例中,可根据给定用户的userID、用户当前所在小区的ID以及进入此小区的时间,扫描上述分析数据,获取此用户在此小区的平均停留时间,再根据用户进入该小区的时间,可得到用户可能离开的时间。
其中,在本发明的上述实施例中,上述步骤13的具体步骤还可以为:根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
在本发明的具体实施例中,可根据给定用户的userID、用户当前所在小区的ID以及进入此小区的时间,扫描上述分析数据,获取用户接下来最有可能出现在哪个小区。具体地,可从分析数据中的Probability字段中得到在下一小区出现的概率,及在下一小区的停留时间。
为了更好的实现上述目的,如图4所示,本发明的实施例提供了一种分析用户轨迹的装置,该装置包括:
清洗模块41,设置为对采集到的用户的信令数据进行清洗;
获得模块42,设置为根据清洗后的信令数据,得到用户的运动轨迹参数,运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;
确定模块43,设置为根据给定用户的唯一标识和用户的运动轨迹参数,确定给定用户的轨迹。
其中,清洗模块41包括:
补全单元,设置为对采集到的用户的信令数据中的残缺信令数据进行补全;
判断单元,设置为根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据,并当补全后的用户的信令数据是错误信令数据时,触发纠正单元;
纠正单元,设置为根据判断单元的触发,纠正该错误信令数据;
去重单元,设置为对纠正后的用户的信令数据进行去重处理。
其中,获得模块42包括:
第一单元,设置为从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序;
第二单元,设置为根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间;
第三单元,设置为根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
其中,确定模块43包括:
第四单元,设置为根据给定用户的唯一标识,通过扫描用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;
第五单元,设置为判断该给定用户进入该小区的平均时间是否在第二预设时间段内,并当该给定用户进入该小区的平均时间在第二预设时间段内时,触发第六单元;
第六单元,设置为根据第五单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段,并当该给定用户在该小区的停留时间覆盖第三预设时间段时,确定该小区为该给定用户的居住地。
其中,确定模块43还包括:
第七单元,设置为判断该给定用户进入该小区的平均时间是否在第四预设时间段内,并当该给定用户进入该小区的平均时间在第四预设时间段内时,触发第八单元;
第八单元,设置为根据第七单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段,并当该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段时,确定该小区为该给定用户的工作地。
其中,确定模块43还包括:
第九单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
其中,确定模块43还包括:
第十单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
需要说明的是,本发明实施例提供的分析用户轨迹的装置是应用上述方法的装置,即上述方法的所有实施例均适用于该装置,且均能达到相同或相似的有益效果。
以上所述是本发明的优选实施方式,应当指出,对于本技术领域的普通技术人员来说,在不脱离本发明所述原理的前提下,还可以作出若干改进和润饰,这些改进和润饰也应视为本发明的保护范围。
工业实用性
基于本发明实施例提供的上述技术方案,对采集到的用户的信令数据进行清洗,并根据清洗后的信令数据,得到用户的运动轨迹参数,再根据给定用户的唯一标识从用户的运动轨迹参数中确定出该给定用户的轨迹,从而简单、高效地分析信令数据,确定出给定用户的轨迹。

Claims (14)

  1. 一种分析用户轨迹的方法,包括:
    对采集到的用户的信令数据进行清洗;
    根据清洗后的信令数据,得到用户的运动轨迹参数,所述运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;
    根据给定用户的唯一标识和所述用户的运动轨迹参数,确定给定用户的轨迹。
  2. 如权利要求1所述的方法,其中,所述对采集到的用户的信令数据进行清洗的步骤包括:
    对采集到的用户的信令数据中的残缺信令数据进行补全;
    根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据;
    若是错误信令数据,则纠正该错误信令数据;
    对纠正后的用户的信令数据进行去重处理。
  3. 如权利要求1所述的方法,其中,所述根据清洗后的信令数据,得到用户的运动轨迹参数的步骤包括:
    从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序;
    根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,所述每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间;
    根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
  4. 如权利要求1所述的方法,其中,所述根据给定用户的唯一标识和所述用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
    根据给定用户的唯一标识,通过扫描所述用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;
    判断该给定用户进入该小区的平均时间是否在第二预设时间段内;
    若该给定用户进入该小区的平均时间在第二预设时间段内,则进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段;
    若该给定用户在该小区的停留时间覆盖第三预设时间段,则确定该小区为该给定用户的居住地。
  5. 如权利要求4所述的方法,其中,所述根据给定用户的唯一标识,通过扫描所述用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据的步骤之后,所述方法还包括:
    判断该给定用户进入该小区的平均时间是否在第四预设时间段内;
    若该给定用户进入该小区的平均时间在第四预设时间段内,则进一步判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段;
    若该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段,则确定该小区为该给定用户的工作地。
  6. 如权利要求1所述的方法,其中,所述根据给定用户的唯一标识和所述用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
    根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描所述用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
  7. 如权利要求1所述的方法,其中,所述根据给定用户的唯一标识和所述用户的运动轨迹参数,确定给定用户的轨迹的步骤包括:
    根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描所述用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
  8. 一种分析用户轨迹的装置,包括:
    清洗模块,设置为对采集到的用户的信令数据进行清洗;
    获得模块,设置为根据清洗后的信令数据,得到用户的运动轨迹参数,所述运动轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户进入每个小 区的最早时间和最晚时间以及用户在每个小区内的最短停留时间和最长停留时间;
    确定模块,设置为根据给定用户的唯一标识和所述用户的运动轨迹参数,确定给定用户的轨迹。
  9. 如权利要求8所述的装置,其中,所述清洗模块包括:
    补全单元,设置为对采集到的用户的信令数据中的残缺信令数据进行补全;
    判断单元,设置为根据信令数据规则,判断补全后的用户的信令数据是否为错误信令数据,并当补全后的用户的信令数据是错误信令数据时,触发纠正单元;
    纠正单元,设置为根据所述判断单元的触发,纠正该错误信令数据;
    去重单元,设置为对纠正后的用户的信令数据进行去重处理。
  10. 如权利要求8所述的装置,其中,所述获得模块包括:
    第一单元,设置为从清洗后的信令数据中获取用户每天的信令数据,并按照用户进入不同小区的时间先后顺序对获取到的每天的信令数据进行排序;
    第二单元,设置为根据排序后的信令数据,得到预设天数内用户每天的轨迹参数,所述每天的轨迹参数包括用户的唯一标识、每个小区的唯一标识、用户每天进入每个小区的最早时间和最晚时间以及用户每天在每个小区内的最短停留时间和最长停留时间;
    第三单元,设置为根据预设权值,对预设天数内用户每天的轨迹参数进行加权处理,得到用户的运动轨迹参数,其中,距离当前时刻越远的轨迹参数的预设权值越小。
  11. 如权利要求8所述的装置,其中,所述确定模块包括:
    第四单元,设置为根据给定用户的唯一标识,通过扫描所述用户的运动轨迹参数,获取该给定用户在小区内的最短停留时间为第一预设时间对应的运动轨迹数据;
    第五单元,设置为判断该给定用户进入该小区的平均时间是否在第二预设时间段内,并当该给定用户进入该小区的平均时间在第二预设时间段内时,触发第六单元;
    第六单元,设置为根据所述第五单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第三预设时间段,并当该给定用户在该小区的停留时间覆盖第三预设时间段时,确定该小区为该给定用户的居住地。
  12. 如权利要求11所述的装置,其中,所述确定模块还包括:
    第七单元,设置为判断该给定用户进入该小区的平均时间是否在第四预设时间段内,并当该给定用户进入该小区的平均时间在第四预设时间段内时,触发第八单元;
    第八单元,设置为根据所述第七单元的触发,进一步判断该给定用户在该小区的停留时间是否覆盖第五预设时间段和第六预设时间段,并当该给定用户在该小区的停留时间覆盖第五预设时间段和第六预设时间段时,确定该小区为该给定用户的工作地。
  13. 如权利要求8所述的装置,其中,所述确定模块还包括:
    第九单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描所述用户的运动轨迹参数,得到该给定用户离开当前所在小区的时间。
  14. 如权利要求8所述的装置,其中,所述确定模块还包括:
    第十单元,设置为根据给定用户的唯一标识、该给定用户当前所在小区的唯一标识以及进入该小区的时间,通过扫描所述用户的运动轨迹参数,得到该给定用户下一个最可能出现的小区的唯一标识以及在该小区中停留的时间。
PCT/CN2015/078220 2015-02-12 2015-05-04 一种分析用户轨迹的方法及装置 WO2016127507A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510075261.4A CN105989226A (zh) 2015-02-12 2015-02-12 一种分析用户轨迹的方法及装置
CN201510075261.4 2015-02-12

Publications (1)

Publication Number Publication Date
WO2016127507A1 true WO2016127507A1 (zh) 2016-08-18

Family

ID=56614245

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/078220 WO2016127507A1 (zh) 2015-02-12 2015-05-04 一种分析用户轨迹的方法及装置

Country Status (2)

Country Link
CN (1) CN105989226A (zh)
WO (1) WO2016127507A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111578933A (zh) * 2020-05-09 2020-08-25 北京上下文系统软件有限公司 一种快速识别用户进入特定地理区域内的方法
CN112561759A (zh) * 2020-01-02 2021-03-26 北京融信数联科技有限公司 一种基于移动信令大数据的毕业生去向动态监测方法
CN113537879A (zh) * 2021-06-28 2021-10-22 深圳市盈捷创想科技有限公司 基于大数据的物品派发方法、装置和计算机可读存储介质
CN116033354A (zh) * 2022-12-16 2023-04-28 中科世通亨奇(北京)科技有限公司 一种用户位置属性信息的分析方法及系统

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107948738B (zh) * 2016-10-13 2020-10-20 北京国双科技有限公司 一种网络电视去重用户数的计算方法及装置
CN106792514B (zh) * 2016-11-30 2020-10-30 南京华苏科技有限公司 基于信令数据的用户职住地分析方法
CN110572813A (zh) * 2018-05-19 2019-12-13 北京融信数联科技有限公司 一种基于移动大数据的手机用户行为相似度分析方法
CN110958558A (zh) * 2018-09-26 2020-04-03 北京融信数联科技有限公司 一种基于移动大数据的手机用户时空轨迹刻画方法
CN111417075B (zh) * 2018-12-18 2023-06-06 北京融信数联科技有限公司 一种基于移动通信大数据的用户工作地识别方法
CN111372234A (zh) * 2018-12-25 2020-07-03 北京融信数联科技有限公司 一种手机用户乒乓切换智能化判定方法
CN110837528A (zh) * 2019-11-15 2020-02-25 智慧足迹数据科技有限公司 人口统计方法、装置及服务器
CN111669710B (zh) * 2020-04-21 2021-07-06 上海因势智能科技有限公司 人口统计去重方法
CN111680059B (zh) * 2020-04-29 2023-08-15 国家计算机网络与信息安全管理中心 基于特定区域的活跃人员、常驻人员的分析方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2247126A2 (en) * 2009-04-28 2010-11-03 Telefonaktiebolaget L M Ericsson (Publ) Predicting presence of a mobile user equipment
CN102075850A (zh) * 2009-11-19 2011-05-25 中国移动通信集团吉林有限公司 确定移动用户职业信息的方法及其装置
CN103052022A (zh) * 2011-10-17 2013-04-17 中国移动通信集团公司 基于移动行为的用户稳定点发现方法和系统
CN103731916A (zh) * 2014-01-14 2014-04-16 上海河广信息科技有限公司 一种基于无线网络的用户位置预测系统和方法
CN103914563A (zh) * 2014-04-18 2014-07-09 中国科学院上海微系统与信息技术研究所 一种时空轨迹的模式挖掘方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6778834B2 (en) * 2001-02-27 2004-08-17 Nokia Corporation Push content filtering
CN101146262A (zh) * 2007-10-23 2008-03-19 中兴通讯股份有限公司 一种移动用户长期活动区域信息生成的方法及系统
CN102682041B (zh) * 2011-03-18 2014-06-04 日电(中国)有限公司 用户行为识别设备及方法
CN103096371B (zh) * 2013-01-16 2016-05-18 上海大唐移动通信设备有限公司 一种室内用户的识别方法和系统
CN103227821B (zh) * 2013-04-03 2015-07-01 腾讯科技(深圳)有限公司 一种目标用户位置数据的处理方法及装置
CN104348635A (zh) * 2013-07-24 2015-02-11 中国移动通信集团福建有限公司 一种移动用户行为分析方法和装置
CN103354633B (zh) * 2013-07-31 2016-09-21 上海欣方软件有限公司 基于信令数据挖掘分析的特定移动用户粗定位系统及方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2247126A2 (en) * 2009-04-28 2010-11-03 Telefonaktiebolaget L M Ericsson (Publ) Predicting presence of a mobile user equipment
CN102075850A (zh) * 2009-11-19 2011-05-25 中国移动通信集团吉林有限公司 确定移动用户职业信息的方法及其装置
CN103052022A (zh) * 2011-10-17 2013-04-17 中国移动通信集团公司 基于移动行为的用户稳定点发现方法和系统
CN103731916A (zh) * 2014-01-14 2014-04-16 上海河广信息科技有限公司 一种基于无线网络的用户位置预测系统和方法
CN103914563A (zh) * 2014-04-18 2014-07-09 中国科学院上海微系统与信息技术研究所 一种时空轨迹的模式挖掘方法

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112561759A (zh) * 2020-01-02 2021-03-26 北京融信数联科技有限公司 一种基于移动信令大数据的毕业生去向动态监测方法
CN112561759B (zh) * 2020-01-02 2023-08-04 北京融信数联科技有限公司 一种基于移动信令大数据的毕业生去向动态监测方法
CN111578933A (zh) * 2020-05-09 2020-08-25 北京上下文系统软件有限公司 一种快速识别用户进入特定地理区域内的方法
CN111578933B (zh) * 2020-05-09 2022-03-11 北京上下文系统软件有限公司 一种快速识别用户进入特定地理区域内的方法
CN113537879A (zh) * 2021-06-28 2021-10-22 深圳市盈捷创想科技有限公司 基于大数据的物品派发方法、装置和计算机可读存储介质
CN116033354A (zh) * 2022-12-16 2023-04-28 中科世通亨奇(北京)科技有限公司 一种用户位置属性信息的分析方法及系统

Also Published As

Publication number Publication date
CN105989226A (zh) 2016-10-05

Similar Documents

Publication Publication Date Title
WO2016127507A1 (zh) 一种分析用户轨迹的方法及装置
CN106992994B (zh) 一种云服务的自动化监控方法和系统
CN106548343B (zh) 一种非法交易检测方法及装置
CN109740648A (zh) 电力负荷异常数据识别方法、装置和计算机设备
CN109522221B (zh) 一种提高模糊测试效率的方法和系统
CN103150177A (zh) 一种更新公交线路数据的方法、装置和系统
CN111782460A (zh) 大规模日志数据的异常检测方法、装置和存储介质
CN113822366A (zh) 业务指标异常检测方法及装置、电子设备、存储介质
CN107862863B (zh) 一种交通数据质量提升的方法及装置
CN112770265B (zh) 一种行人身份信息获取方法、系统、服务器和存储介质
CN106294219A (zh) 一种设备识别、数据处理方法、装置及系统
US20210182384A1 (en) Detecting fraudulent facial recognition
CN110869942A (zh) 自馈深度学习方法和系统
CN102521534A (zh) 一种基于粗糙熵属性约简的入侵检测方法
CN112817814A (zh) 异常监控方法、系统、存储介质及电子装置
CN103096371B (zh) 一种室内用户的识别方法和系统
CN111160797A (zh) 风控模型的构建方法、装置、存储介质及终端
CN111444075A (zh) 一种自动发现关键影响力指标的方法
CN117421692B (zh) 垃圾投放站的垃圾违规投放识别方法、装置、设备
CN110728395A (zh) 主变短期电力负荷计算方法、装置、计算机和存储介质
CN112637888B (zh) 覆盖空洞区域识别方法、装置、设备及可读存储介质
WO2021262344A1 (en) Method and apparatus to detect scripted network traffic
CN114223139A (zh) 界面切换方法、装置、可穿戴电子设备及存储介质
CN115422263A (zh) 一种电力现场多功能通用型故障分析方法及系统
CN114139707A (zh) 设备数据采集平台

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15881668

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15881668

Country of ref document: EP

Kind code of ref document: A1