CN101441730B - Case detecting method and system - Google Patents
Case detecting method and system Download PDFInfo
- Publication number
- CN101441730B CN101441730B CN2007101667879A CN200710166787A CN101441730B CN 101441730 B CN101441730 B CN 101441730B CN 2007101667879 A CN2007101667879 A CN 2007101667879A CN 200710166787 A CN200710166787 A CN 200710166787A CN 101441730 B CN101441730 B CN 101441730B
- Authority
- CN
- China
- Prior art keywords
- event
- historical events
- incident
- data
- changes
- 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.)
- Active
Links
Images
Landscapes
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention provides an event detecting method, which generates at least one optimum lifecycle model according to the data of at least one historical event, at least one nutrition growth function and at least one event trigger point rule. The method comprises the following steps of receiving an event, describing the strength value of the event according to a lifecycle model corresponding to the event, judging whether an even trigger point is reached according to the strength of the event, sending out an event sending notice and then sending an event notice when the strength of the event changes to the even trigger point. The event detecting method has the advantages of improving the capability of tracing the subsequent development of the event, ensuring that the trigger of the event is more accurate and accords with the situation of the event actually taking place better, comprehending the evolution process of the event and filtering false alarms.
Description
Technical field
The invention relates to a kind of case detecting method, and be particularly to a kind of case detecting method that is evolved to the basis with incident intensity.
Background technology
Case detecting (Event Detection) technology has been widely used in many different systems, for example, and enterprise security, driving monitoring, sports ... Or the like.Traditional case detecting algorithm mainly is the characteristic variations that relies on after incident taken place, the foundation that takes place as decision event with taked in response to measure, when after detect similar event feature when changing, can take correspondence in response to measure.
Traditional case detecting algorithm is directly to detect from the original value of the incident factor, and is whether basis (Rule-Based) comes decision event to produce with the rule.The production process that its shortcoming is incident not only simply but also intuition but is too simplified for the reason of its formation, causes and ignores the potential meaning that incident takes place easily.
Therefore, the invention provides and a kind ofly be evolved to the case detecting method on basis, can improve the ability that the incident follow-up developments are followed the trail of, and make the detecting of incident more accurate with incident intensity.
Summary of the invention
Based on above-mentioned purpose, the embodiment of the invention has disclosed a kind of case detecting method.Obtain historical events data according to an incident factor from a historical events database; Select a nutrition growth function according to these historical events data from a nutrition growth function library, and select at least one Event triggered point rule from a trigger point rule database according to these historical events data.These historical events data and this nutrition growth function are applied mechanically to a life periodic model, to calculate the changes in intensity values of this historical events.The changes in intensity values of this historical events is compared with a historical events time of origin section of obtaining from this historical events database according to these historical events data, whether meet historical events with changes in intensity values and change according to this this historical events of Event triggered point rule judgment.Change if meet this historical events, then with this life cycle model as the righttest life cycle model.Reception is to a event data that should the incident factor, and utilizes this righttest life cycle model to depict the changes in intensity values of this event data.Changes in intensity values according to this event data judges whether to arrive a predeterminable event trigger point.When the changes in intensity values of this event data arrives this predeterminable event trigger point, then send a incident that should event data.
The embodiment of the invention has more disclosed a kind of case detecting system, comprises that a historical events database, a trigger point rule database, a nutrition growth function library, life cycle training module, an incident intensity tracing module and send event module.This historical events database should store many historical events data.This trigger point rule database stores many Event triggered point rules.This nutrition growth function library stores a plurality of nutrition growth functions.This life cycle training module is obtained historical events data according to an incident factor from a historical events database; Select a nutrition growth function according to these historical events data from a nutrition growth function library; Select at least one Event triggered point rule according to these historical events data from a trigger point rule database; These historical events data and this nutrition growth function are applied mechanically to a life periodic model; To calculate the changes in intensity values of this historical events; The changes in intensity values of this historical events is compared with a historical events time of origin section of obtaining from this historical events database according to these historical events data; Whether meet historical events with changes in intensity values and change, and if meeting this historical events changes according to this this historical events of Event triggered point rule judgment, then with this life cycle model as the righttest life cycle model.This incident intensity tracing module receives a event data that should the incident factor from a data receiver; And utilize this righttest life cycle model to depict the changes in intensity values of this event data; Changes in intensity values according to this event data judges whether to arrive a predeterminable event trigger point; And when the changes in intensity values of this event data arrives this Event triggered point, send a notice to make this transmission event module transmission to a incident that should event data.
Description of drawings
Fig. 1 is the configuration diagram that shows the case detecting system of the embodiment of the invention.
Fig. 2 is the generation synoptic diagram of right life cycle model that shows the embodiment of the invention.
Fig. 3 A～Fig. 3 D is the generation detailed process synoptic diagram of right life cycle model that show to implement Fig. 2.
Fig. 4 is the flow chart of steps that shows the case detecting method of the embodiment of the invention.
Symbol description:
110～historical events database
120～trigger point rule database
130～nutrition growth function library
140～life cycle training module
150～the righttest incident life cycle model
160～incident intensity tracing module
170～data receiver
180～transmission event module
Embodiment
For let the object of the invention, characteristic, and advantage can be more obviously understandable, hereinafter is special lifts preferred embodiment, and cooperates appended pictorial image 1 to Fig. 4, does detailed explanation.Instructions of the present invention provides various embodiment that the technical characterictic of the different embodiments of the present invention is described.Wherein, the usefulness that is configured to explanation of each assembly among the embodiment is not in order to restriction the present invention.And the part of reference numerals repeats among the embodiment, is for the purpose of simplifying the description, is not the relevance that means between the different embodiment.
The embodiment of the invention has disclosed a kind of case detecting method and system that is evolved to the basis with incident intensity.
In embodiments of the present invention, mainly be to be that example is explained it with the medical care, but do not as limit, also can be applicable to water quality detection, air quality monitoring real ... Or the like.
Fig. 1 is the configuration diagram that shows the case detecting system of the embodiment of the invention.
The case detecting system of the embodiment of the invention comprises that at least (Most AdaptableEvent Life Cycle Model, AEM) 150, one incident intensity tracing module (Event StrengthTracker) 160, one data receiver (Data Receiver) 170 and sends event module (FiringEvent Module) 180 to a historical events database (Historical Event Database) 110, one trigger point rule database (Firing Point Rule Database) 120, one nutrition growth function library (Nutrition Growing Library) 130, one life cycle training module (Life Cycle Trainer) 140, one the righttest incident life cycle model.
The righttest incident life cycle model 150 is the output of life cycle training module, and it comprises nutrition growth function and the Event triggered point rule that is applicable to the output incident.Incident intensity tracing module 160 receives data from data receiver 170, and according to the changes in intensity values of life cycle model that should data being depicted incident.When up-to-date changes in intensity values arrival event trigger point (Event Firing Point) time, then can a notice be sent to and send event module 180.Data receiver 170 is accepted the source data of coming of incident, and converts in order to the required data layout of the changes in intensity values of describing incident.Send the changes in intensity values of event module 180, convert the kind of incident to the form of notification alert and then transmission incident according to incident.
With reference to figure 1 and Fig. 2; When the training stage (Training); Life cycle training module 140 is obtained incident in the interval historical events data that take place of different time from historical events database 110; Rule database 120 (is for example obtained various Event triggered point rule from the trigger point; Slope (Slope), intensity of variation (Variation), critical value (Threshold), mean value (Average) ... Or the like); And form long function library 130 on one's own account and obtain various nutrition growth function (for example, regretional analysis function (Regression), variance analysis function (Difference), time-dependent function (Dependencyto Time), similarity comparison function (Similarity Comparison) ... Or the like).Life cycle training module 140 is compared historical events data and an event, and according to nutrition growth function of selecting and the regular the righttest life cycle model that produces this incident factor of Event triggered point.
For producing the implementing procedure of right life cycle model, with reference to figure 3A, at first to select the incident factor, for example, blood oxygen concentration, diastolic pressure, pulse, respiration rate, body temperature, pulse pressure ... Or the like.Then, select different nutrition growth functions, and apply mechanically nutrition growth function to the different life cycle model of selection, to depict the changes in intensity values of event data from nutrition growth function library 130.In the selection of nutrition growth function, originally be to select with the mode of " training ", tranmittance is found out only nutrition growth function to the mode of trigger point again.This training patterns is all computing comparisons together.
The selection of nutrition growth function also can be done selecting of some preset property according to the characteristic of the incident factor, like this can be than faster in the training comparison.The characteristic that is based on the following factor of selecting of the incident factor is implemented.With normal codomain, some incident factor has bound, the incident intensity level surpass the upper bound or lower bound just be regarded as undesired.With normal codomain and bounded above situation, it is exactly abnormal that the incident intensity level surpasses the upper bound.With normal codomain and bounded below situation, it is exactly abnormal that the incident intensity level surpasses lower bound.Some incident factor is considerable for the main body of detecting, for example, blood pressure, one has slightly, and ANOMALOUS VARIATIONS will cause very big influence to human body.Therefore, giving the Sensitivity Index (Sensitive Index) of this incident factor will be than higher.Just can stress the importance of the incident factor by adjustment Sensitivity Index height.
After obtaining the changes in intensity values of event data, this event data is overlapped in time section and this changes in intensity values that history takes place, the changes in intensity values when really taking place to find out this event data is shown in Fig. 3 B.Then, apply mechanically the Event triggered point rule of selection, the case point during with the comparison Event triggered and the historical time of origin of this event data are shown in Fig. 3 C.Last figure is the incident comparison of applying mechanically the critical value rule, and middle figure is the incident comparison of applying mechanically the slope rule, and figure below is the incident comparison of applying mechanically the intensity of variation rule, and wherein the point on the figure is for applying mechanically the triggering opportunity that Event triggered point rule causes.
When applying mechanically critical value when rule, then in the changes in intensity values of event data trigger event notice when a preset critical is above all.When applying mechanically slope when rule, then trigger event notice when changing (skyrocketing/fall suddenly) is significantly arranged at the slope of the intensity level of event data.When applying mechanically intensity of variation when rule, then trigger event notice when significantly making a variation is arranged at the intensity level of event data.With reference to figure 3C; Applying mechanically regular result and the historical situation that is produced of critical value meets most; Be that historical interval interior (shown in Fig. 3 D) all dropped in the trigger point, therefore, the life cycle model of applying mechanically the critical value rule promptly is chosen as the righttest incident life cycle model 150 of this event data.
The trigger point rule can be more rigorous on judging if add the dimension of a time.The below judgement of simple declaration trigger point rule again.
If apply mechanically the critical value rule, whether the intensity level of the decision event of indicating data is all maintaining on the preset critical within a period of time all the time.If the short time maintains on the preset critical, expression has the alarm trigger that immediacy takes place.If maintain for a long time on the preset critical, then expression confirms that the intensity level of event data must maintain the above alarm trigger of just doing of preset critical.If apply mechanically the intensity of variation rule, the expression meeting is from the incident intensity level of certain time point the recall the past period changes in intensity values of event data of (long-time or short time), to calculate its variation value of incident instantly.If apply mechanically the slope rule, be illustrated in rising or the variation of downtrending of intensity level of the event data of a certain period (long-time or short time) lining.For the attention degree of this event data whether the representative of the length of time.If the incident factor that susceptibility is high then gives the slope more interval than the short period and judges.In addition, can detect the change point (that is, will upwards change or change) of trend downwards from the intensity level of event data in online variation of time.
See through above method and find out the alarm trigger point of changes in intensity values, compare with the event history data again.Each number of point that utilizes above method to find out work as molecule, and and the time point that is overlapped between the event history generating region work as denominator, again result of calculation is represented with percent that it is accurate that this trigger point rule of the high more representative of percent heals.
When the execute phase (Running), event data is constantly detected and accepted to data receiver 170, and the incident that receives converts in order to describe the required data layout of incident intensity, sends this incident to incident intensity tracing module 160 then.Incident intensity tracing module 160 receives these event datas from data receiver 170, and according to life cycle model that should event data is depicted event data changes in intensity values.In the time of up-to-date changes in intensity values arrival event trigger point, then can an incident be sent notice and send to and send event module 180.Send event module 180 and receive this incident when sending notice,, convert the form of notification alert to, send an incident then according to the kind of incident with the result of variations of incident intensity.
Fig. 4 is the flow chart of steps that shows the case detecting method of the embodiment of the invention.
At first; Obtain historical events data (step S401) from a historical events database; Select nutrition growth function (step S402) from a nutrition growth function library; And the nutrition growth function of historical events data that obtain and selection applied mechanically to a life periodic model, with the changes in intensity values (step S403) of calculating this historical events.After obtaining the changes in intensity values of this historical events, this historical events is overlapped the incident changes in intensity values (step S404) when really taking place to find out this historical events at the time section of history generation and the changes in intensity values of this historical events.Select Event triggered point rule (step S405) from a trigger point rule database; And whether meet the righttest principle (promptly whether meeting the historical events variation) (step S406) of the trigger point rule judgment of preceding text according to the changes in intensity values of this historical events of Event triggered point rule judgment of selecting.
If do not meet the righttest principle, then get back to step S402 and select other nutrition growth function and Event triggered point rule, to carry out the matching operation of life cycle model again.If meet the righttest principle, then produce the righttest life cycle model (step S407).Track of events is come source data and receiving event data; And, judge whether arrival event trigger point (step S409) with changes in intensity values according to this event data according to the changes in intensity values (step S408) of the righttest life cycle model that should event data being depicted this event data.In the time of the changes in intensity values arrival event trigger point of this event data, then send an incident and send the part notice, and send an incident (step S410).If no show Event triggered point then continues the track of events data.
The case detecting System and method for of the embodiment of the invention mainly is the characteristic according to the event history digital data; The digital data of incident is converted to the numerical value of representative incident intensity; And through analyzing its intensity level in online variation of time; Determine whether this incident is enough to be triggered to the incident into real, and then form the notice and the alarm of incident.
The present invention more provides a kind of recording medium (for example discs, disk sheet and removable hard drive or the like), and it writes down the authority sign-off program of an embodied on computer readable, so that carry out above-mentioned case detecting method.At this; Be stored in the authority sign-off program on the recording medium; Basically (for example the setting up organization chart code segment, sign-off forms code segment, setting program code snippet and deployment program code snippet) formed by a plurality of code segment, and the function series of these code segment corresponds to the step of said method and the functional block diagram of said system.
Though the present invention discloses as above with preferred embodiment; Right its is not in order to limiting the present invention, anyly has the knack of this art, do not breaking away from the spirit and scope of the present invention; When can doing various changes and retouching, so protection scope of the present invention is as the criterion when looking the claim scope person of defining.
Claims (8)
1. a case detecting method comprises the following steps:
Obtain historical events data according to an incident factor from a historical events database, wherein this incident factor is in a blood oxygen concentration, a diastolic pressure, a pulse, a respiration rate, a body temperature or the pulse pressure;
Select a nutrition growth function according to these historical events data from a nutrition growth function library, wherein this nutrition growth function is with the computing of carrying out of the raw data of incident, is converted into this incident and puts the nutritive value that is obtained at a time;
Select at least one Event triggered point rule according to these historical events data from a trigger point rule database;
These historical events data and this nutrition growth function are applied mechanically to a life periodic model, to calculate the changes in intensity values of this historical events;
The changes in intensity values of this historical events is compared with a historical events time of origin section of obtaining from this historical events database according to these historical events data; Whether meeting historical events with the changes in intensity values according to this this historical events of Event triggered point rule judgment changes; Apply mechanically the Event triggered point rule of selection, the case point during with the comparison Event triggered and the historical time of origin of this incident;
Change if meet this historical events, then with this life cycle model as the righttest life cycle model;
Reception is to a event data that should the incident factor, and utilizes this righttest life cycle model to depict the changes in intensity values of this event data;
Changes in intensity values according to this event data judges whether to arrive a predeterminable event trigger point; And when the changes in intensity values of this event data arrives this predeterminable event trigger point, then send a incident that should event data;
Wherein, whether this Event triggered point rule will positively send to the data receiving end for whether be enough to trigger the rule of a real incident in order to the changes in intensity values of decision event with the notice that determines Event triggered.
2. case detecting method as claimed in claim 1; Wherein, Select a plurality of nutrition growth functions from this nutrition growth function library, and apply mechanically nutrition growth function to a plurality of life cycle models of selection, to describe the incident Strength Changes of corresponding each life cycle model.
3. case detecting method as claimed in claim 1 wherein, when this Event triggered point rule is critical value rule, is then notified at changes in intensity values trigger event when a preset critical is above of this event data.
4. case detecting system comprises:
One historical events database; Store many historical events data; Be used for obtaining historical events data according to an incident factor from this historical events database, wherein this incident factor is in a blood oxygen concentration, a diastolic pressure, a pulse, a respiration rate, a body temperature or the pulse pressure;
One trigger point rule database stores many Event triggered point rules, is used for selecting at least one Event triggered point rule according to these historical events data from this trigger point rule database;
One nutrition growth function library; Store a plurality of nutrition growth functions; Be used for selecting a nutrition growth function from a nutrition growth function library according to these historical events data; Wherein this nutrition growth function is with the computing of carrying out of the raw data of incident, is converted into this incident and puts the nutritive value that is obtained at a time;
One life cycle training module is used for these historical events data and this nutrition growth function are applied mechanically to a life periodic model, to calculate the changes in intensity values of this historical events; The changes in intensity values of this historical events is compared with a historical events time of origin section of obtaining from this historical events database according to these historical events data; Whether meeting historical events with the changes in intensity values according to this this historical events of Event triggered point rule judgment changes; Apply mechanically the Event triggered point rule of selection, the case point during with the comparison Event triggered and the historical time of origin of this incident; Change if meet this historical events, then with this life cycle model as the righttest life cycle model;
One incident intensity tracing module is used to receive to a event data that should the incident factor, and utilizes this righttest life cycle model to depict the changes in intensity values of this event data;
One judge module is used for judging whether to arrive a predeterminable event trigger point according to the changes in intensity values of this event data; And
One sends event module; When the changes in intensity values of this event data arrives this predeterminable event trigger point, then send a incident that should event data.
5. case detecting as claimed in claim 4 system; Wherein, This life cycle training module is selected a plurality of nutrition growth functions from this nutrition growth function library, and applies mechanically nutrition growth function to a plurality of life cycle models of selection, to describe the incident Strength Changes of corresponding each life cycle model.
6. case detecting as claimed in claim 4 system, wherein, when this Event triggered point rule is critical value rule, then in changes in intensity values trigger event notice when a preset critical is above of this event data.
7. case detecting as claimed in claim 4 system; Wherein, This life cycle training module utilizes the said nutrition growth function in this nutrition growth function library to calculate an event data and be converted into this event data at the nutritive value that is obtained sometime; And calculate the up-to-date changes in intensity values of this event data according to this nutritive value, and according to different historical events nests with different nutrition growth functions.
8. case detecting as claimed in claim 4 system, wherein, this sends the changes in intensity values of event module according to this event data, converts the kind of this event data the form of this notice to, and then sends this incident.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101667879A CN101441730B (en) | 2007-11-19 | 2007-11-19 | Case detecting method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101667879A CN101441730B (en) | 2007-11-19 | 2007-11-19 | Case detecting method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101441730A CN101441730A (en) | 2009-05-27 |
| CN101441730B true CN101441730B (en) | 2012-05-23 |
Family
ID=40726152
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2007101667879A Active CN101441730B (en) | 2007-11-19 | 2007-11-19 | Case detecting method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101441730B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103530796B (en) * | 2013-10-10 | 2016-06-01 | 北京智谷睿拓技术服务有限公司 | The active period detection method of application program and active period detection system |
| CN106375429A (en) * | 2016-08-31 | 2017-02-01 | 宁波金唐软件有限公司 | Task push system |
| TWI624815B (en) * | 2016-11-23 | 2018-05-21 | 財團法人資訊工業策進會 | Behavior detection system and method thereof |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070038587A1 (en) * | 2005-07-27 | 2007-02-15 | Fujitsu Limited | Predicting apparatus, predicting method, and computer product |
-
2007
- 2007-11-19 CN CN2007101667879A patent/CN101441730B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070038587A1 (en) * | 2005-07-27 | 2007-02-15 | Fujitsu Limited | Predicting apparatus, predicting method, and computer product |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101441730A (en) | 2009-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20140266690A1 (en) | Automated event severity determination in an emergency assistance system | |
| KR101851690B1 (en) | A Appartus and Method for Anomaly Detection of the Circadian Rhythm Monitored with Wearable Sensors | |
| CN109325692A (en) | The data real-time analysis method and device of pipe network | |
| US11922796B2 (en) | Predicting critical alarms | |
| CN101441730B (en) | Case detecting method and system | |
| EP3495905A2 (en) | Operational improvement effect calculation device, operational improvement effect calculation method and recording medium | |
| CN114601478B (en) | Method, device and equipment for improving alertness of driver and readable storage medium | |
| US20220183566A1 (en) | Systems and methods for diurnal curve tracking and analysis | |
| CN114680889A (en) | Method and device for identifying unsafe behaviors of drilling operation personnel | |
| US11033226B2 (en) | Detecting non-evident contributing values | |
| US20210321956A1 (en) | Determination of health status of systems equipped with sensors | |
| CN103810809B (en) | The method of fire is judged based on DS evidence theory | |
| CN107690687B (en) | Detection and early prediction system for patients with chronic obstructive bronchitis when they are close to attack | |
| CN110811583A (en) | Heart rate monitoring method and device, electronic equipment and computer readable storage medium | |
| CN110517731A (en) | Genetic test quality monitoring data processing method and system | |
| Zhang et al. | Statistical monitoring-based alarming systems in modeling the AIDS epidemic in the United States, 1985-2011 | |
| Dreesman et al. | Challenges for signal generation from medical social media data | |
| CN101458738A (en) | Disease serious degree ordering method, system thereof and recording medium | |
| Poh et al. | Anomaly detection for home activity based on sequence pattern | |
| CN207125724U (en) | Wearable custodial care facility system | |
| JP5109059B2 (en) | How to operate complex dynamic processes | |
| Woldaregay et al. | Electronic Disease Surveillance System Based on Input from People with Diabetes: An Early Outbreak Detection Mechanism | |
| Bersimis et al. | Public health monitoring using control charts based on convex hull | |
| US20220142521A1 (en) | Methods and apparatus for predicting whether and when a hypo/hyper analyte concentration event will occur | |
| EP3833240A1 (en) | Incorporating contextual data in a clinical assessment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |