Summary of the invention
The technical problem to be solved in the present invention is, call bill data analytical method for prior art cannot obtain the defect of the analysis result for specific client, provides a kind of method of the user's of determination normality address points and a kind of method of carrying out call bill data analysis based on this user's normality address points.
The technical solution adopted for the present invention to solve the technical problems is: propose a kind of method determining user's normality address points, comprising:
A, obtain all call bill datas of same user within a period of time;
B, to determine in all call bill datas got four positions of most east, south, west, north according to the geographical location information of all call bill datas got, and with these four positions for according to delimiting a rectangle analyzed area;
C, this rectangle analyzed area is divided into multiple grids with predetermined length and width, and counts the number of the call bill data occurred in each grid;
D, with the grid with maximum call bill data number for according to delimiting out candidate region, place;
E, the periphery of the candidate region, place delimited out is all increased preset length, and the candidate region, place that the number that periphery is all increased call bill data in preset length rear region does not increase delimited as user's normality address points.
Above-mentionedly determine in the method for user's normality address points, in described step c, the length of grid and width are greater than 30 meters and are less than predetermined threshold value.
Above-mentionedly determine that, in the method for user's normality address points, the number counting the call bill data occurred in each grid in described step c specifically comprises:
From the some grids in this rectangle analyzed area, add up the call bill data occurred in each grid in this analyzed area one by one.
Above-mentionedly determine that, in the method for user's normality address points, described steps d specifically comprises:
The number of the call bill data of d1, more each grid, determines the grid with maximum call bill data numbers;
D2, the grid this with maximum call bill data number delimited as candidate region, place.
Above-mentionedly determine that, in the method for user's normality address points, described steps d specifically comprises:
The number of the call bill data of d1, more each grid, determines the grid with maximum call bill data numbers;
D2, be the center of circle by the grid center of a lattice that there is maximum call bill data number with this, there is with this border circular areas that length of grid of maximum call bill data number or the half of width are radius delimit as candidate region, place.
Above-mentionedly determine that, in the method for user's normality address points, described step e specifically comprises:
E1, the periphery of the candidate region, place delimited out is all increased preset length;
E2, judge peripheral increase preset length after whether there is call bill data in the region that increases;
If there is call bill data, then repeated execution of steps e 1 in the region that e3 increases;
If there is not call bill data in the region that e4 increases, then the candidate region, place of correspondence delimited as user's normality address points.
Above-mentionedly determine that, in the method for user's normality address points, described step a specifically comprises:
The all call bill datas of same user within a period of time are obtained in real time or regularly from each operator.
Above-mentionedly determine that, in the method for user's normality address points, described step a comprises further:
By the call bill data consolidation form obtained from each operator.
Above-mentionedly determine in the method for user's normality address points, described a period of time comprise one day, one week, one month or 1 year.
The present invention also proposes a kind of call bill data analytical method, comprising:
A, determine user's normality address points of user according to the method for the aforementioned user's of determination normality address points;
B, the call bill data analyzed in this user's normality address points, obtain the analysis data of user behavior.
Method of the present invention determines user's normality address points by carrying out the geographical location information of user's call bill data reasonably dividing, and user is often analyzed with the call bill data in address points, the analysis result obtained is pointed, according to these analysis results, better for specific client provides better communication environment and communication service, thus can carry out the division of high-end customer, the important in inhibiting such as definition, the guarantee of fine work network, accurately thoughtful marketing in VIP region for operator.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.
Fig. 1 shows the flow chart of the method 100 of the determination user normality address points that one embodiment of the invention provides.As shown in Figure 1, this determines that the method 100 of user's normality address points starts from step 102.
Subsequently, at next step 104, obtain all call bill datas of same user within a period of time.
Subsequently, at next step 106, to determine in all call bill datas got four positions of most east, south, west, north according to the geographical location information of all call bill datas got, and with these four positions for according to delimiting a rectangle analyzed area.Wherein, the method for delimiting rectangle analyzed area is specially: delimit 4 perpendicular to the straight line in plane with the most eastern position, the most southern position, the most western position, the most northern position respectively, and the region that 4 straight line intersection obtain is rectangle analyzed area.
Subsequently, at next step 108, this rectangle analyzed area is divided into multiple grids with predetermined length and width, and counts the number of the call bill data occurred in each grid.
Subsequently, at next step 110, with the grid with maximum call bill data number for according to delimiting out candidate region, place.
Subsequently, at next step 112, the periphery of the candidate region, place delimited out is all increased preset length, and the candidate region, place that the number that periphery is all increased call bill data in preset length rear region does not increase delimited as user's normality address points.
Finally, method 100 ends at step 114.
Fig. 2 shows the flow chart of the method 200 of the determination user normality address points that another specific embodiment of the present invention provides.As shown in Figure 2, this determines that the method 200 of user's normality address points starts from step 202.
Subsequently, at next step 204, obtain all call bill datas of same user within a period of time.In specific embodiment, step 204 can obtain all call bill datas of same user within a period of time from each operator in real time or regularly.This period of time can be one day, one week, one month or 1 year according to specific circumstances with needs.Such as, all call bill datas of party a subscriber on May 4th, 2010 are obtained in step 204.From the call bill data of the user of each operator acquisition in step 204, the problem that form is inconsistent may be there is.For the ease of subsequent analysis, after all call bill datas obtaining user, by these call bill data consolidation forms, such as, generate data format simple, extendible comma separated value (Comma-Separated Values, CSV) file.
Subsequently, at next step 206, determine four positions of most east, south, west, north in the call bill data obtained according to the geographical location information (such as latitude and longitude information) of call bill data, and with these four positions for according to delimiting a rectangle analyzed area.In this step 206, analyze the latitude and longitude information entrained by all call bill datas obtained, and the latitude and longitude information that more each call bill data is corresponding, judge which call bill data is in the most eastern position, which call bill data is in the most southern or the most western or the most northern position.After four positions determining most east, south, west, north, just can determine the zone of action of user within certain a period of time that this call bill data is corresponding.With determine four positions (see tetra-points of A, B, C, the D in Fig. 3 and Fig. 4) for foundation, delimit out a rectangle analyzed area (see rectangle frame in Fig. 3 and Fig. 4 301 and 401), as the scope of activities of user within certain a period of time.It will be appreciated by persons skilled in the art that and delimit analyzed area for rectangle, is for the ease of analyzing.
Subsequently, at next step 208, this analyzed area is divided into multiple grids with predetermined length and width.Such as, in a concrete example, this rectangle analyzed area be divided into multiple length and be widely all greater than 30 meters and be less than the grid of predetermined threshold value.There is error in the latitude and longitude information of carrying due to call bill data, error range is probably at about 30 meters, if therefore by the length of grid or widely divide too small, the analysis result obtained after analyzing the call bill data in grid is also not accurate enough, has little significance; If by the length of grid or wide division too much, the region obtained is easily excessive, and the analysis result obtained so is also not accurate enough.Therefore analyzed area can be divided into and long and widely all be greater than 30 meters and be less than the grid of predetermined threshold value, be such as divided into grow and wide be all the grid of 50 meters.
Subsequently, at next step 210, from the some grids in this analyzed area, add up the number of the call bill data occurred in each grid in this analyzed area one by one.In a specific embodiment, such as, can add up from the grid in the upper left corner of this analyzed area, certainly, also can add up from other grids, only the number of the call bill data in all grids need be come out.
Subsequently, at next step 212, the number of the call bill data in more each grid to determine the grid with maximum call bill data numbers, and with this grid for according to delimiting out candidate region, place.With the grid with maximum call bill data number for according to delimiting candidate region, place, can have various ways, such as Fig. 3 and Fig. 4 respectively illustrates two kinds of modes.In embodiment shown in Fig. 3, directly the grid 302 with maximum call bill data number delimited as candidate region, place.In embodiment shown in Fig. 4, being delimited by the border circular areas obtained as the center of circle, using the half of the length of this grid 402 as radius using the center of the grid 402 with maximum call bill data number is candidate region, place 403.Certainly, also the border circular areas obtained using the center of grid 402 as the center of circle, using the half of the width of grid 402 as radius can be delimited is candidate region, place 403.In different embodiment, the shape of candidate region, place can be rectangle, circle etc., specifically can set as required.
Subsequently, at next step 214, the periphery of the candidate region, place delimited out is all increased preset length.In this step, if as shown in Figure 3, candidate region, place is the grid 302 of rectangle, then make the length of grid 302 and widely all increase same preset length, obtains the region 303 after the increase that dotted line represents.If as shown in Figure 4, candidate region, place is border circular areas 403, then make the radius in region 403 increase the length preset, obtain the region 404 after the increase that dotted line represents.
Subsequently, at next step 216, judge whether there is call bill data in the region increased.The region memory increased after the periphery of candidate region, place being increased preset length if judge in the step 216 is at call bill data, then repeated execution of steps 214, the periphery of candidate region, place is continued increase the length preset, until no longer there is call bill data in the region increased.
There is not call bill data in the region increased if judge in the step 216, then perform step 218 subsequently.In step 218, the candidate region, place that after periphery is all increased preset length, the number of call bill data no longer increases delimited as user's normality address points.As shown in Figure 3, if there is not call bill data between candidate region, place 302 and the region 303 after increasing, then this candidate region, place 302 will as user's normality address points.Equally, as shown in Figure 4, if there is not call bill data between candidate region, place 403 and the region 404 after increasing, then this candidate region, place 403 will as user's normality address points.
Finally, method 200 ends at step 220.
Fig. 5 shows the flow chart of the call bill data analytical method 500 that one embodiment of the invention provides.As shown in Figure 5, this call bill data analytical method 500 starts from step 502.
Subsequently, at next step 504, determine user's normality address points of user according to the geographical location information of call bill data.Relevantly how to determine user's normality address points, be described in detail above, therefore do not repeat them here.
Subsequently, at next step 506, analyze the call bill data in this user's normality address points, obtain the analysis data of user behavior.In this step 506, the call bill data analyzed in user's normality address points comprises: judge the whether normal break-make of user's communication; Analyze quality in user's communication process as how.These analysis results, for the division of high-end customer, the important in inhibiting such as definition, the guarantee of fine work network, accurately thoughtful marketing in VIP region.
Finally, method 500 ends at step 508.
Utilize the above call bill data analytical method of the present invention introduced, the multiple user normality address points of user in one day, in one week, in one month or in 1 year can be determined, and by analyzing the call bill data in multiple user's normality address points, obtain the analysis data of user behavior, thus more targetedly for user provides better communication environment and communication service.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.