CN105574471A - Uploading method of user behavior data, and user behavior identification method and device - Google Patents
Uploading method of user behavior data, and user behavior identification method and device Download PDFInfo
- Publication number
- CN105574471A CN105574471A CN201410532912.3A CN201410532912A CN105574471A CN 105574471 A CN105574471 A CN 105574471A CN 201410532912 A CN201410532912 A CN 201410532912A CN 105574471 A CN105574471 A CN 105574471A
- Authority
- CN
- China
- Prior art keywords
- user
- behavior
- vector
- dynamic frequency
- frequency vector
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an uploading method of user behavior data, and a user behavior identification method and device. The uploading method of the user behavior data comprises the following steps: obtaining the acceleration speed of a user within a time period of a preset duration by an acquisition terminal, and determining and recording the behavior types of the user within the time period according to the acceleration speed; according to a preset period, carrying out statistics on the occurrence frequency of various recorded behavior types of the user within the preset period, and generating a dynamic frequency vector, wherein each component of the dynamic frequency vector is various behavior types sorted according to a preset sequence, and the value of each component corresponds to the occurrence frequencies of various behavior types; and uploading the dynamic frequency vector to a platform. The uploading method lowers requirements on the hardware handling capability and the requirements on the transmission capability of an acquisition terminal so as to finally realize a purpose that user behaviors are identified on the basis of increasing no cost.
Description
Technical field
The present invention relates to data service technical field, especially relate to a kind of method for uploading of user behavior data, the recognition methods of user behavior and device.
Background technology
Detailed User Activity identification can comprehensive monitoring User Activity, forms activity log, the main activities of recording user, and can be on this basis, according to the activity analysis result of user, exercise guidance and suggestion are carried out to user, improve the physical condition of user comprehensively.
Usually by acceleration raw data identification user behavior in prior art, existing Activity recognition algorithm carries out the identification of feature extraction and classifying device based on acceleration raw data.These algorithm realization behavioural analyses of direct employing have two kinds of strategies: (1) strengthens terminal transmission ability, raw data or feature are uploaded to rear end platform and calculate; (2) strengthen terminal computing power, realize Activity recognition in terminal; And in practical application, these two kinds of strategies are all difficult to realize.Wherein (1) volume of transmitted data of needing is large, can increase system energy consumption and hardware cost; (2) need stronger hardware processing capability, can hardware cost be increased.Both are all not suitable for low cost, the low-power consumption hardware platform that current Wearable extensively adopts.
Summary of the invention
The object of this invention is to provide a kind of method for uploading of user behavior data, the recognition methods of user behavior and device, overcome and strengthen terminal transmission ability in behavioural analysis, raw data or feature are uploaded to the defect that rear end platform calculates this strategy existence, overcome simultaneously and strengthen terminal computing power, realize the defect of this strategy existence of Activity recognition in terminal.
To achieve these goals, the invention provides a kind of method for uploading of user behavior data, comprise: the acceleration information of user in the time period that acquisition terminal obtains scheduled duration, determine according to this acceleration information and record the behavior kind of this user in this time period; According to predetermined period, add up the frequency of occurrences of this user's various actions kind of record in this predetermined period, and generate dynamic vector frequently, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding various actions kind; Described dynamic frequency vector is uploaded to platform.
Optionally, described acceleration information comprises 3-axis acceleration numerical value, describedly to determine according to this acceleration information and the behavior kind recording this user in this time period comprises: calculate described 3-axis acceleration numerical value square and value; Determine that described and belonging to value default value is interval, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, it is interval that different behavior kinds corresponds to different default values.
Optionally, the scope of described scheduled duration is 1 ~ 60S.
According to another aspect of the present invention, provide a kind of recognition methods of user behavior, comprise: receive the dynamic frequency vector that acquisition terminal periodically sends, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, the value of each component is the frequency of occurrences of corresponding various actions kind in predetermined period, and described behavior kind is that the acceleration information of user in the time period according to scheduled duration is determined; The behavior of this user is identified based on described dynamic frequency vector.
Optionally, after the described behavior based on described dynamic this user of frequency vector identification, described method also comprises: the display terminal behavior of this user being sent to this user.
Optionally, the described behavior based on described dynamic this user of frequency vector identification comprises: compare the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, as the behavior kind of this user in described predetermined period.
Optionally, the described proper vector to described dynamic frequency vector and various actions kind compares, using the behavior kind of the proper vector the most similar to described dynamic frequency vector as the behavior kind of this user in described predetermined period, comprising: the maximum comparability calculating described dynamic frequency vector and described proper vector; Using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in described predetermined period.
Optionally, before the proper vector of described dynamic frequency vector and various actions kind is compared, further according to default weight, each component of described dynamic frequency vector is weighted, wherein, each component has default importance index, and the component that importance index is higher has higher weight.
According to another aspect of the present invention, provide a kind of acquisition terminal, comprising: acquisition module, for obtain scheduled duration time period in the acceleration information of user, determine according to this acceleration information and record the behavior kind of this user in this time period; Processing module, for according to predetermined period, add up the frequency of occurrences of this user's various actions kind of record in this cycle, and generate dynamic vector frequently, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding various actions kind; First transport module, for being uploaded to platform by described dynamic frequency vector.
Optionally, described acquisition module comprises the first computing unit and the first comparing unit, described first computing unit for calculate described 3-axis acceleration numerical value square and value; Described first comparing unit is interval for determining the default value belonging to described and value, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, it is interval that different behavior kinds corresponds to different default values.
Optionally, described acquisition module obtain scheduled duration time period in user acceleration information in the scope of scheduled duration be 1 ~ 60S.
According to another aspect of the present invention, provide a kind of recognition system of user behavior, comprise: receiver module, for receiving the dynamic frequency vector that acquisition terminal periodically sends, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, the value of each component is the frequency of occurrences of corresponding various actions kind in predetermined period, and described behavior kind is that the acceleration information of user in the time period according to scheduled duration is determined; Identification module, for identifying the behavior of this user based on described dynamic frequency vector.
Optionally, described recognition system also comprises the second transport module, and described second transport module is used for the display terminal behavior of this user being sent to this user.
Optionally, described identification module comprises the second comparing unit, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, as the behavior kind of this user in described predetermined period.
Optionally, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, comprise in the behavior kind of described predetermined period as this user: the maximum comparability calculating described dynamic frequency vector and described proper vector; Using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in described predetermined period.
Optionally, described identification module also comprises the second computing unit, described second computing unit is used for, before the proper vector of described dynamic frequency vector and various actions kind is compared, further according to default weight, each component of described dynamic frequency vector is weighted, wherein, each component has default importance index, and the component that importance index is higher has higher weight.
The invention has the beneficial effects as follows: the present invention forms dynamic vector frequently by carrying out process at acquisition terminal to the acceleration information collected, then dynamic frequency vector is uploaded to platform, platform identifies the behavior of this user based on dynamic frequency vector, the present invention is by the combination of acquisition terminal and platform, avoid and carry out complex calculations at acquisition terminal, reduce the requirement of the processing power to hardware, thus decrease hardware cost, in addition, by dynamic frequency vector is uploaded to platform, reduce the requirement of the transmittability to acquisition terminal, thus decrease the energy ezpenditure of acquisition terminal, finally achieve and do not increasing on the basis of cost, the behavior of user is identified.
Accompanying drawing explanation
Fig. 1 represents the flow chart of steps of acquisition terminal in the method for uploading of user behavior data in embodiments of the invention;
Fig. 2 represents the flow chart of steps of platform in the recognition methods of user behavior in embodiments of the invention;
Fig. 3 represents that in embodiments of the invention, platform identifies the flow chart of steps of the behavior of this user based on dynamic frequency vector;
Fig. 4 represents the acceleration information of the user in the analysis result shown in RAD database in embodiments of the invention;
Fig. 5 represents the recognition result drawn according to the acceleration information in Fig. 4;
Fig. 6 represents the structured flowchart of acquisition terminal in embodiments of the invention; And
Fig. 7 represents the structured flowchart of the recognition system in embodiments of the invention.
Embodiment
Below with reference to accompanying drawings exemplary embodiment of the present disclosure is described in more detail.Although show exemplary embodiment of the present disclosure in accompanying drawing, however should be appreciated that can realize the disclosure in a variety of manners and not should limit by the embodiment set forth here.On the contrary, provide these embodiments to be in order to more thoroughly the disclosure can be understood, and complete for the scope of the present disclosure can be conveyed to those skilled in the art.
As shown in Figure 1, be the flow chart of steps of acquisition terminal in embodiments of the invention in the method for uploading of user behavior data, comprise the steps:
Step S101, the acceleration information of user in the time period obtaining scheduled duration, determines according to acceleration information and records the behavior kind of this user in this time period.
In the present embodiment, the acceleration information of user in the time period of acquisition terminal acquisition scheduled duration, preferably, at acquisition terminal, be reduce EMS memory occupation, the scope of described scheduled duration can be 1 ~ 60S.
Preferably, acceleration information can be 3-axis acceleration three numerical value axially, and acquisition terminal can by processing the behavior kind obtaining this user to the numerical value of 3-axis acceleration three axis.Concrete process can be, the acceleration value according to three axis carries out rough handling, obtains a result; Then, determine the behavior kind that this result is corresponding, and as the behavior kind of this user in this time period.
For example, rough handling can be, calculate 3-axis acceleration three axial numerical value square and value.Acquisition terminal first calculate 3-axis acceleration three axial numerical value square and value, suppose that scheduled duration is 2S, then calculate within the time period of each 2S a 3-axis acceleration three axial numerical value square and value.Above-mentioned rough handling can also be: the square root calculating 3-axis acceleration three axial numerical value, and then carries out read group total to the square root that three axis obtain, and obtains one and value; Or calculate 3-axis acceleration three axial numerical value square with value after, then the variance yields calculating and be worth; Or calculate 3-axis acceleration three axial numerical value absolute value and value, etc.That is, to the process of the numerical value of 3-axis acceleration, can select voluntarily according to specific environment and statistics experience, be not limited to above citing.
If the mode that rough handling carried out to the acceleration value of three axis be calculate 3-axis acceleration three axial numerical value square and value, then described in calculating and after value, can determine that described and belonging to value default value is interval, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, preset different behavior kinds and correspond to different default value intervals.Such as, the numerical intervals corresponding to different behavior kind is provided with in advance in described acquisition terminal, calculate 3-axis acceleration numerical value square with value after, this and value are contrasted from different numerical intervals, determine which numerical intervals this and value are positioned at, suppose that this and value are positioned at the first numerical intervals, then the behavior kind corresponding to the first numerical intervals is this and the behavior kind belonging to value.
If the mode that rough handling carried out to the acceleration value of three axis be calculate 3-axis acceleration three axial numerical value square and value, again the variance yields that variance calculates is carried out to this and value, so this variance yields can be contrasted from different numerical intervals in the embodiment of the present invention, determine which numerical intervals this variance yields is positioned at, suppose that this variance yields is positioned at second value interval, then the behavior kind corresponding to second value interval is the behavior kind belonging to this variance yields.The calculated amount of variance yields is less, and more can reflect the Changing Pattern of user's acceleration, and the embodiment of the present invention can preferably adopt variance yields to determine the behavior kind of user in acquisition terminal side.
Step S102, according to predetermined period, adds up the frequency of occurrences of this user's various actions kind of record in this predetermined period, and generates dynamic vector frequently.
In the present embodiment, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding behavior kind.Be described with specific embodiment at this, if various actions kind is respectively static, by bus, ride, activity and exercise 5 kinds of behavior kinds, then each component of described dynamic frequency vector is respectively static according to after predefined procedure sequence, by bus, ride, activity and exercise, the value of each component of described dynamic frequency vector is respectively static, by bus, ride, the movable frequency of occurrences with tempering 5 kinds of behavior kinds, suppose static, by bus, ride, the frequency of occurrences that is movable and exercise 5 kinds of behavior kinds is respectively D1, D2, D3, D4, D5, i.e. dynamic vectorial D=(D1 frequently, D2, D3, D4, D5).
Optionally, predetermined period can by hour in units of set, also can set in units of sky, namely predetermined period is not limited to concrete time restriction, suppose that predetermined period is 5Min, then acquisition terminal adds up the frequency of occurrences of this user's various actions kind of record in this 5Min, finally the frequency of occurrences of this user's various actions kind is formed dynamic vector frequently.
Step S103, described dynamic frequency vector is uploaded to platform, it should be noted that at this, acquisition terminal can be Wearable device, as passometer.
As shown in Figure 2, be the flow chart of steps of platform in embodiments of the invention in the recognition methods of user behavior, comprise the steps:
Step S201, receives the dynamic frequency vector that acquisition terminal periodically sends, it should be noted that, the dynamic frequency vector generated in the dynamic frequency vector of reception and step S102 is same dynamic frequency vector.
Step S202, identifies the behavior of this user based on described dynamic frequency vector, after obtaining the information of described dynamic frequency vector, by processing to described dynamic frequency vector the Activity recognition result finally obtaining this user.
Preferably, after the behavior identifying user, the behavior of this user can also be sent to the display terminal of this user, wherein, display terminal can be computer, mobile phone etc.
As shown in Figure 3, be the process flow diagram of the further refinement of step S202 in Fig. 2, comprise the steps:
Step S301, is weighted each component of described dynamic frequency vector.
In the present embodiment, be preset with the weight of dynamic each component of vector frequently in platform, each component of described dynamic frequency vector is weighted according to the weight of each component respectively; Preferably, each component has default importance index, and the component that importance index is higher then has higher weight, and weight corresponding to the behavior kind that namely user is different is different.Be described with specific embodiment at this, suppose that various actions kind is respectively static, rides, rides, movable and temper 5 kinds of behavior kinds, due to static, ride, ride, movable and temper the appearance that all can have static behavior in 5 kinds of behavior kinds, therefore the identification contribution of static behavior to user behavior is minimum, then the importance index of static behavior is just minimum, and weight is then minimum; On the contrary, exercise behavior only appears in exercise behavior kind usually, static, ride, ride, seldom there will be exercise behavior in the behavior kind such as movable, and it is maximum for therefore tempering the identification contribution of behavior to user behavior, the importance index of then tempering behavior is just the highest, and weight is then maximum; When static, ride, ride, after weight that is movable and that temper 5 kinds of behavior kinds determines, be then weighted respectively according to this weight each component to dynamic frequency vector.
Step S302, compares the proper vector of described dynamic frequency vector and various actions kind.
In the present embodiment, the proper vector of various actions kind is preset with in platform.Preferably, in platform, include sorter, in sorter, store the proper vector of various actions kind, if described proper vector did not carry out weighting, then described dynamic frequency vector and proper vector are compared; If described proper vector is the proper vector after weighting, then the dynamic frequency vector after weighting and the proper vector after weighting are compared.
Step S303, calculates the maximum comparability of described dynamic frequency vector and described proper vector, using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in predetermined period.
In the present embodiment, preferably, calculate the method that the computing method adopted in the similarity of described dynamic frequency vector and described proper vector are COS distance measured similarity, described COS distance is defined as
Wherein, X is described dynamic frequency vector, x
1for the one-component in described dynamic frequency vector, x
2for second component in described dynamic frequency vector, the like, x
nfor in described dynamic frequency vector n-th vector, Y be predetermined described various actions kind proper vector, y
1for the one-component of described proper vector, y
2for second component of described proper vector, the like, y
nfor the n-th component of described proper vector, θ be frequently vectorial X and various actions kind proper vector Y between angle.
When calculating the maximum comparability of described dynamic frequency vector and described proper vector, described dynamic frequency vector is compared with the proper vector of various actions kind, namely the cosine value of angle between described dynamic frequency vector and the proper vector of various actions kind is calculated respectively, cosine value is larger, then illustrate that similarity is larger, after obtaining maximum similarity, using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in predetermined period, namely kind is last recognition result the behavior.
Be described in this citing, suppose that various actions kind is respectively static, rides, rides, movable and temper 5 kinds of behavior kinds, the proper vector of static behavior is Y
1, the proper vector of behavior is by bus Y
2, the proper vector of driving behavior is Y
3, the proper vector of crawler behavior is Y
4, the proper vector of tempering behavior is Y
5, dynamic vector is frequently D, then calculate the proper vector Y of dynamic vectorial D and static behavior frequently respectively
1, by bus behavior proper vector Y
2, driving behavior proper vector Y
3, crawler behavior proper vector Y
4, temper the proper vector Y of behavior
5between similarity, namely calculate the proper vector Y of dynamic vectorial D and static behavior frequently respectively
1, by bus behavior proper vector Y
2, driving behavior proper vector Y
3, crawler behavior proper vector Y
4, temper the proper vector Y of behavior
5between the cosine value of angle, cosine value is larger, then illustrate that similarity is larger, gets the behavior kind of the corresponding proper vector of maximum similarity for this dynamic behavior kind frequently belonging to vector, namely the behavior kind of this user in predetermined period, is also last Activity recognition result.
It should be noted that at this, above-mentioned recognition result may comprise mistake accidental individually, can carry out aftertreatment according to the experience of life to recognition result.Such as, be mingled with the strenuous exercise of 5 minutes in multiple daily routines, normally judge by accident, can remove.Within 5 minutes that happen suddenly in long-time quiescing process, be also probably erroneous judgement by bus, can be removed.
The beneficial effect of this recognition methods is described below in conjunction with example.
As shown in Figure 4, Fig. 4 is the acceleration information of the user in the analysis result shown in RAD database, and these data can mark data as prompting assisted user.2 hours 5 minutes are about by when can find out data acquisition in Fig. 4.Actual conditions are 10 minutes sort articles and walk to parking lot (daily routines), within 35 minutes, drive, within 25 minutes, play basketball, walking in 5 minutes, within 15 minutes, drive, within 5 minutes, walk to family from parking lot, then always static.
Fig. 5 represents the recognition result drawn according to the acceleration information in Fig. 4, Fig. 5 user's various actions kind is divided into static, ride, ride, movable and temper 5 kinds of behavior kinds, almost conform to completely with the annotation results of step count information by finding out in Fig. 5 that this result is combined to remember with user, only have 1.5 hours places that once mistake occurs, namely last movable.This section of activity, for walk to family from parking lot, comprises parking, walking, takes elevator and sit down at home in 5 minutes, because 5 minutes movable complicated, finally ridden by being comprehensively identified as of mistake.
This mistake is understandable, in 5 minutes, comprise repeatedly State Transferring, and the excessive point (being namely switched to the period of another state from a state) of state can cause acquisition terminal identification to produce mistake, and then platform also can produce mistake.Test result on RAD data set shows, recognition accuracy reaches 94.21%.
As shown in Figure 6, be the structured flowchart of acquisition terminal in embodiments of the invention, acquisition terminal 400 comprises:
Acquisition module 401, for obtain scheduled duration time period in the acceleration information of user, determine according to this acceleration information and record the behavior kind of this user in this time period;
Processing module 402, for according to predetermined period, add up the frequency of occurrences of this user's various actions kind of record in this cycle, and generate dynamic vector frequently, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding various actions kind;
First transport module 403, for being uploaded to platform by described dynamic frequency vector.
Optionally, acquisition module 401 comprises the first computing unit and the first comparing unit, described first computing unit for calculate described 3-axis acceleration numerical value square and value; Described first comparing unit is interval for determining the default value belonging to described and value, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, it is interval that different behavior kinds corresponds to different default values.
Optionally, acquisition module 401 obtain scheduled duration time period in user acceleration information in the scope of scheduled duration be 1 ~ 60S.
As shown in Figure 7, be the structured flowchart of the recognition system of user behavior in embodiments of the invention, recognition system 500 comprises:
Receiver module 501, for receiving the dynamic frequency vector that acquisition terminal periodically sends, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, the value of each component is the frequency of occurrences of corresponding various actions kind in predetermined period, and described behavior kind is that the acceleration information of user in the time period according to scheduled duration is determined;
Identification module 502, for identifying the behavior of this user based on described dynamic frequency vector.
Optionally, recognition system 500 also comprises the second transport module, and described second transport module is used for the display terminal behavior of this user being sent to this user.
Optionally, identification module 502 comprises the second comparing unit, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, as the behavior kind of this user in described predetermined period.
Optionally, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, comprise in the behavior kind of described predetermined period as this user: the maximum comparability calculating described dynamic frequency vector and described proper vector; Using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in described predetermined period.
Optionally, identification module 502 also comprises the second computing unit, described second computing unit is used for, before the proper vector of described dynamic frequency vector and various actions kind is compared, further according to default weight, each component of described dynamic frequency vector is weighted, wherein, each component has default importance index, and the component that importance index is higher has higher weight.
Above-described is the preferred embodiment of the present invention; should be understood that the ordinary person for the art; can also make some improvements and modifications not departing under principle prerequisite of the present invention, these improvements and modifications are also in protection scope of the present invention.
Claims (16)
1. a method for uploading for user behavior data, is characterized in that, comprising:
The acceleration information of user in the time period that acquisition terminal obtains scheduled duration, determines according to this acceleration information and records the behavior kind of this user in this time period;
According to predetermined period, add up the frequency of occurrences of this user's various actions kind of record in this predetermined period, and generate dynamic vector frequently, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding various actions kind;
Described dynamic frequency vector is uploaded to platform.
2. method for uploading as claimed in claim 1, is characterized in that,
Described acceleration information comprises 3-axis acceleration numerical value, describedly to determine according to this acceleration information and the behavior kind recording this user in this time period comprises: calculate described 3-axis acceleration numerical value square and value; Determine that described and belonging to value default value is interval, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, it is interval that different behavior kinds corresponds to different default values.
3. method for uploading as claimed in claim 1, it is characterized in that, the scope of described scheduled duration is 1 ~ 60S.
4. a recognition methods for user behavior, is characterized in that, comprising:
Receive the dynamic frequency vector that acquisition terminal periodically sends, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, the value of each component is the frequency of occurrences of corresponding various actions kind in predetermined period, and described behavior kind is that the acceleration information of user in the time period according to scheduled duration is determined;
The behavior of this user is identified based on described dynamic frequency vector.
5. recognition methods as claimed in claim 4, is characterized in that, after the described behavior based on described dynamic this user of frequency vector identification, described method also comprises:
The behavior of this user is sent to the display terminal of this user.
6. recognition methods as claimed in claim 4, is characterized in that, the described behavior based on described dynamic this user of frequency vector identification comprises:
The proper vector of described dynamic frequency vector and various actions kind is compared, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, as the behavior kind of this user in described predetermined period.
7. recognition methods as claimed in claim 6, is characterized in that,
The described proper vector to described dynamic frequency vector and various actions kind compares, and using the behavior kind of the proper vector the most similar to described dynamic frequency vector as the behavior kind of this user in described predetermined period, comprising:
Calculate the maximum comparability of described dynamic frequency vector and described proper vector;
Using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in described predetermined period.
8. recognition methods as claimed in claim 7, is characterized in that,
Before the proper vector of described dynamic frequency vector and various actions kind is compared, further according to default weight, each component of described dynamic frequency vector is weighted, wherein, each component has default importance index, and the component that importance index is higher has higher weight.
9. an acquisition terminal, is characterized in that, comprising:
Acquisition module, for obtain scheduled duration time period in the acceleration information of user, determine according to this acceleration information and record the behavior kind of this user in this time period;
Processing module, for according to predetermined period, add up the frequency of occurrences of this user's various actions kind of record in this cycle, and generate dynamic vector frequently, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, and the value of each component is the frequency of occurrences of corresponding various actions kind;
First transport module, for being uploaded to platform by described dynamic frequency vector.
10. acquisition terminal as claimed in claim 9, it is characterized in that, described acquisition module comprises the first computing unit and the first comparing unit, described first computing unit for calculate described 3-axis acceleration numerical value square and value; Described first comparing unit is interval for determining the default value belonging to described and value, and by interval for described default value corresponding behavior kind, as the behavior kind of this user in this time period, wherein, it is interval that different behavior kinds corresponds to different default values.
11. acquisition terminals as claimed in claim 9, is characterized in that, in the time period that described acquisition module obtains scheduled duration user acceleration information in the scope of scheduled duration be 1 ~ 60S.
The recognition system of 12. 1 kinds of user behaviors, is characterized in that, comprising:
Receiver module, for receiving the dynamic frequency vector that acquisition terminal periodically sends, each component of described dynamic frequency vector is the various actions kind after sorting according to predefined procedure, the value of each component is the frequency of occurrences of corresponding various actions kind in predetermined period, and described behavior kind is that the acceleration information of user in the time period according to scheduled duration is determined;
Identification module, for identifying the behavior of this user based on described dynamic frequency vector.
13. recognition systems as claimed in claim 12, is characterized in that, also comprise the second transport module, and described second transport module is used for the display terminal behavior of this user being sent to this user.
14. recognition systems as claimed in claim 12, it is characterized in that, described identification module comprises the second comparing unit, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, as the behavior kind of this user in described predetermined period.
15. recognition systems as claimed in claim 14, it is characterized in that, described second comparing unit is used for comparing the proper vector of described dynamic frequency vector and various actions kind, by the behavior kind of the proper vector the most similar to described dynamic frequency vector, comprise in the behavior kind of described predetermined period as this user: the maximum comparability calculating described dynamic frequency vector and described proper vector; Using the behavior kind of maximum comparability characteristic of correspondence vector as the behavior kind of this user in described predetermined period.
16. recognition systems as claimed in claim 15, it is characterized in that, described identification module also comprises the second computing unit, described second computing unit is used for, before comparing the proper vector of described dynamic frequency vector and various actions kind, further according to default weight, each component of described dynamic frequency vector is weighted, wherein, each component has default importance index, and the component that importance index is higher has higher weight.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410532912.3A CN105574471B (en) | 2014-10-10 | 2014-10-10 | The recognition methods of the method for uploading, user behavior of user behavior data and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410532912.3A CN105574471B (en) | 2014-10-10 | 2014-10-10 | The recognition methods of the method for uploading, user behavior of user behavior data and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105574471A true CN105574471A (en) | 2016-05-11 |
| CN105574471B CN105574471B (en) | 2019-02-05 |
Family
ID=55884581
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410532912.3A Active CN105574471B (en) | 2014-10-10 | 2014-10-10 | The recognition methods of the method for uploading, user behavior of user behavior data and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105574471B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107463689A (en) * | 2017-08-10 | 2017-12-12 | 深圳市金立通信设备有限公司 | Generation method, moving state identification method and the terminal in motion characteristic data storehouse |
| CN109272432A (en) * | 2018-08-08 | 2019-01-25 | 广州杰赛科技股份有限公司 | User behavior monitoring method and system, computer equipment, computer storage medium |
| CN110520824A (en) * | 2017-04-14 | 2019-11-29 | 奇跃公司 | The tracking of multi-mode eyes |
| CN112115978A (en) * | 2020-08-24 | 2020-12-22 | 中国人民健康保险股份有限公司 | Motion recognition method and device and storage medium |
| CN114168948A (en) * | 2021-12-17 | 2022-03-11 | 北京华清信安科技有限公司 | Network security situation comprehensive analysis method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102087722A (en) * | 2009-12-04 | 2011-06-08 | 索尼公司 | Learning apparatus and method, prediction apparatus and method, and program |
| CA2758710A1 (en) * | 2011-11-07 | 2013-05-07 | Martin Bonac | Cadence detection |
| CN103886341A (en) * | 2014-03-19 | 2014-06-25 | 国家电网公司 | Gait behavior recognition method based on feature combination |
| CN103908259A (en) * | 2014-04-03 | 2014-07-09 | 深圳市德凯瑞科技有限公司 | Intelligent wearable device and human motion monitoring and recognition method |
-
2014
- 2014-10-10 CN CN201410532912.3A patent/CN105574471B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102087722A (en) * | 2009-12-04 | 2011-06-08 | 索尼公司 | Learning apparatus and method, prediction apparatus and method, and program |
| CA2758710A1 (en) * | 2011-11-07 | 2013-05-07 | Martin Bonac | Cadence detection |
| CN103886341A (en) * | 2014-03-19 | 2014-06-25 | 国家电网公司 | Gait behavior recognition method based on feature combination |
| CN103908259A (en) * | 2014-04-03 | 2014-07-09 | 深圳市德凯瑞科技有限公司 | Intelligent wearable device and human motion monitoring and recognition method |
Non-Patent Citations (2)
| Title |
|---|
| KAZUYA MURAO: "A Motion Recognition Method by Constancy-Decision", 《INTERNATIONAL SYMPOSIUM ON WEARABLE COMPUTERS (ISWC) 2010》 * |
| 徐川龙 等: "一种基于三维加速度传感器的人体行为识别方法", 《计算机系统应用》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110520824A (en) * | 2017-04-14 | 2019-11-29 | 奇跃公司 | The tracking of multi-mode eyes |
| CN110520824B (en) * | 2017-04-14 | 2023-11-24 | 奇跃公司 | Multimode eye tracking |
| CN107463689A (en) * | 2017-08-10 | 2017-12-12 | 深圳市金立通信设备有限公司 | Generation method, moving state identification method and the terminal in motion characteristic data storehouse |
| CN109272432A (en) * | 2018-08-08 | 2019-01-25 | 广州杰赛科技股份有限公司 | User behavior monitoring method and system, computer equipment, computer storage medium |
| CN109272432B (en) * | 2018-08-08 | 2020-11-13 | 广州杰赛科技股份有限公司 | User behavior monitoring method and system, computer device and computer storage medium |
| CN112115978A (en) * | 2020-08-24 | 2020-12-22 | 中国人民健康保险股份有限公司 | Motion recognition method and device and storage medium |
| CN114168948A (en) * | 2021-12-17 | 2022-03-11 | 北京华清信安科技有限公司 | Network security situation comprehensive analysis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105574471B (en) | 2019-02-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105574471A (en) | Uploading method of user behavior data, and user behavior identification method and device | |
| CN103801048B (en) | A kind of intelligence rope skipping and control method | |
| CN108345846A (en) | A kind of Human bodys' response method and identifying system based on convolutional neural networks | |
| CN106709235A (en) | Exercise training data processing method and device | |
| CN104022552B (en) | Intelligent detection method for electric vehicle charging control | |
| CN104090912A (en) | Information pushing method and device | |
| WO2005015476A3 (en) | System and method for determining equivalency factors for use in comparative performance analysis of industrial facilities | |
| CN103747523A (en) | User position predicating system and method based on wireless network | |
| CN112067015B (en) | Step counting method and device based on convolutional neural network and readable storage medium | |
| CN107341668A (en) | A kind of method and device for judging True Data | |
| CN103731916A (en) | Wireless-network-based user position predicting system and method | |
| CN107992595A (en) | Learning content recommendation method and device and intelligent equipment | |
| CN105335873A (en) | Method and device for acquiring quality of advertisement space and electronic equipment | |
| US11782496B2 (en) | Smart context subsampling on-device system | |
| WO2019052430A1 (en) | Method and apparatus for self-service of mobile terminal | |
| CN105589977A (en) | Movement frequency monitoring method and device | |
| CN112861019A (en) | Course recommendation method and system based on punishment course hot strength and user activity | |
| CN106598222A (en) | Scene mode switching method and system | |
| CN105303039A (en) | Acceleration sensor based behavior analysis system | |
| CN106994238B (en) | Data processing method and electronic device for executing the same | |
| CN115080840A (en) | Content pushing method and device and storage medium | |
| CN104077288A (en) | Web page content recommendation method and web page content recommendation equipment | |
| CN112363065A (en) | Electric quantity monitoring method and device applied to sensor of Internet of things | |
| CN107463689A (en) | Generation method, moving state identification method and the terminal in motion characteristic data storehouse | |
| CN102779120B (en) | Method, system and device for determining field information of station and judging correlation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |