FI125573B

FI125573B - Adaptive management of services that take into account the disruptive effect

Info

Publication number: FI125573B
Application number: FI20135865A
Authority: FI
Inventors: Jukka Rantala; Markus Takala
Original assignee: Elisa Oyj
Priority date: 2013-08-27
Filing date: 2013-08-27
Publication date: 2015-11-30
Also published as: WO2015028714A1; FI20135865A

Description

INTERFERENCE EFFECT OF HAVING REGARD TO ADAPTIVE MANAGEMENT SERVICE

The invention relates to a method according to the preamble of claim 1 for automatic detection and classification of the technical system disturbances in the computer system.

The invention is also a computer system according to the preamble of claim 15 for the corresponding purpose.

Background Art

The prior art includes. a variety of methods and systems for controlling communications.

US 7,539,752 B1 discloses a communications network for proactive and predictive system.

EP 2 408 143 AI describes the virtual networks, proactive resource management method.

WO 2010/112855 Al describes a method of analysis, in particular for telecommunications networks.

Description of the invention

The purpose of the invention is to create a new technical solution to control the technical system.

The invention is based on the fact that the technical system is received event information, which is classified according to different criteria and which to any of the deviation of the data, such as fault information, given the severity of values. The severity values are weighted class of data by the respective weighting coefficients and on the basis of these values calculated from a set of reference values, which are compared to the corresponding limit values. Based on the comparison can make a decision on possible further measures.

More specifically, the method according to the invention is characterized by what is disclosed in the characterizing part of claim 1.

According to the invention, the telephone and / or mobile network associated with or contained in a computer system is characterized by what is disclosed in the characterizing part of claim 15.

The invention offers significant benefits. The invention offers, among others. an efficient way to collect, process and filter a substantial interference from a large data set of individual transaction data. Thus the potential for further measures to be applied and the correct dimensions and to initiate corrective action or preventive action quickly.

In an embodiment in which the collection and processing of information will be the mobile network and / or telephone network elements by means of the method and system can be implemented very cost-effectively and a system it is possible to achieve a very comprehensive geographic coverage. In embodiments monitor the technical system can therefore be a system other than the telecommunications network itself. Thus the control of the telecommunications network outside the system can be utilized in telecommunications systems offered by efficient data processing capabilities and extensive geographical coverage.

In an embodiment where the parameters are set dynamically and automatically, the system can be implemented self-learning. In this case, the system parameters are initially the default values and system to adjust the parameters of the received event information, and possibly also on the basis of other information. Such a system is able to emphasize the events on the basis of information previously collected in such a way that individual random events is filtered off and the attention can be focused to the accumulated problems of interest.

THE FIGURES presentation

The invention is examined below by means of examples and with reference to the accompanying drawings.

Figure 1 shows a general system environment according to one embodiment.

Figure 2 illustrates a further data processing system shown in Figure 1 in accordance with one embodiment of the two.

Figure 3 shows processes according to the embodiments.

Figure 4 shows the enrichment of a single event in accordance with one embodiment.

Figure 5 illustrates the calculation of class-vertical thresholds in accordance with one embodiment.

Sovellusmuotoia invention,

The system of Figure 1 comprises a data processing system 2, which includes. the parameters given in 3 and 4. In addition, the data warehouse system comprises an input interface 1 to be treated through Whose information is received informants 0, and the output interface 5 through which the refined processing of information taken forward eg. operational control and monitoring systems or directly toimenpidekäskyinä employees.

The data processing system 2 is the preferred embodiment of the telephone system and data contained in or associated with a communication network, such as mediation solutions. Telephone and telecommunication network means that, for example, to a mobile network and a fixed telephone network. Telephone and data processing system included in or associated with a communication network, in turn, means that the data processing system involved in carrying out the measures relating to the functioning of the telephone and data transmission network. Such measures may include the collection of event data and telephone communications network, from network elements, event data processing, such as aggregation, correlation, enrichment and pricing. Telephone and data processing system included in or associated with a communication network not, therefore, in this embodiment, the mean of a system which is merely a communication connection of telephone and telecommunications network but which has no functional link the telephone and carried out measures and provided by a telecommunication network services. In such an embodiment the service management can be implemented very efficiently by utilizing efficient servers are used for controlling and monitoring the telephone and the communication network. These servers can thus be in accordance with embodiments of the program to provide basic duties, value-added services in accordance with the application forms.

In accordance with the embodiments may be 0 sources such as telephone and data network elements and telephone and the telecommunications network connection devices.

Figure 2 illustrates a further data processing system shown in Figure 1 in accordance with one embodiment of the two. The data processing system 2 of Figure 2 comprises one element of the online analysis module 21, which in turn contains an event (event), the modeling engine 211 and event analysis engine 212. In Figure 2, the data processing system 2 also comprises one element of the decision engine 23, which already contains the above-mentioned parameters 3 and the data store 4. In addition, decision engine 23 comprises for each classification of information used in the application form their own analysis engine. The embodiment of Figure 2, the analysis engines are the location analysis engine 223, the service analysis engine 224 and customer analysis engine 225. After processing the work of the Analysis The motto of the femur system can produce a variety of descriptions of the event, of which the horizontal, informative, proactive and reactive event description is shown in FIG. The event descriptions can then be disseminated through 5 output interface.

The system of Figure 2 can produce different types of event descriptions. Along with the description of the event of the original event (event) data in the weights and additional information. The second event descriptors type is an entirely new event or event description, which is created on the basis of information received from the event. New events or event descriptions can be created for each classification of knowledge, which may include, for example, customer, location, and service (Customer, Location, and Service). Each new event or event description may also be a disorder of the grade on the basis of threshold values, which may include, for example, Informative, Proactive and Reactive. Error Category Informative is intended for informative one system administrator. Error Category Proactive is intended for proactive measures to prevent the actual service prejudicial interference. Error Category reactive, in turn, is intended to illustrate that hinder service disruptions which require an immediate response.

Figure 3 illustrates processes that can be implemented in the system of Figure 2 through one input interface and an output interface 5 between. These measures are described below with reference to FIG 301 Collect and store event - received in the first stage, event descriptions are collected and stored in the system. Event descriptions are received records that contain information about the events.

302 Known event? - In the second stage, it is checked whether the description of the event described by event known event.

303 Parse all values (customer, location, services) - If the event was known to the system event, the event related parameters are picked up.

304 Values correct or missing? - At the time the record is checked from any incorrect or in case of missing data. This identified the records that contain the classification of the information provided.

305 Create Alert - Information reasonably sufficient to contain the records exported for further processing.

306 Read event profile - further processing initially includes information related to the event. Further process are exported to at least the identified records that include at least one offset information, which is interpreted as a sign disorder.

307 Check Prestored Location parameter - Check if the event justify any change in weight based on the pre-set fixed This refers bottom value.

308 Check Location Experience of value - Calculating site-specific experience value, ie, more specifically, is formed by a weighted grade-specific severity of the value of the record describing the event in terms of class Location. If someone shot on location begin to accumulate unusually large number of events as its weight increases and the weight factor is raised to the automatic form. Weight value can also be manually set higher for example. Due to the trade fair event. The method, therefore, be reviewed, whether or not to change the event based on the weight allocation. The new value can be determined at step 310, at least, in the case of automatic adjustment. Manual setting again can be implemented by changing the preset This refers solid base in the desired manner.

309 Check new information for event - any additional information audit. At this stage, it is checked whether Location weighting also set some new information this shot on location context, which should add event field. Additional information can be e.g.

Information from the running area public event. This step may according to the embodiment also combine to step 307.

310 Add value to Location Experience counter - Updating the counter. This stage of the event data contained in the record is extracted from the classification of the information provided, which contain category information associated with each of the specified event class, which in this embodiment are the Customer, Location and Service. The record contained in deviation of the knowledge is also prescribed for the severity of the value, and this is the weighted class of the corresponding weighting factor Location and thus formed a weighted grade-specific severity of the value is from the record by describing an event in terms of class Location. The resulting class specific gravity of adding value to the class location counter, the value of which can later be compared to the class location so-called vertical threshold value.

311 Check Prestored Service parameter - Check if the event justify any change in weight, based on pre-set to a fixed base Service value. Step corresponds to the phase be 307.

312 Check Service Experience value - Carry out step 308 of the same measures in respect of class Service. If any of the service begins to accumulate unusually large number of events, its importance is growing. Increase the weight coefficient can be automatically or manually, as described above in connection with step 308, for example. The actual value of the determination can be carried out e.g. in step 314, at least, in the case of automatic adjustment.

313 Check new information for event - any additional information audit. Step corresponds to step 309 described above and similar variations are also valid with respect to step 313. In terms of services, more information can be eg. Version of substitutions and the like relevant information related to the service changes.

314 Add value to counter Service Experience - This stage is a stage of 310 similar events in terms of class Service.

Check Prestored 315 Customer parameter - step described above corresponds to steps 307 and 311, respectively, and may be implemented with respect to the class Customer. The above-mentioned comments and modifications apply accordingly.

316 Check Customer Experience value - Step corresponds to the steps described above, 308 and 312 and can implement these accordingly with regard to the class Customer. The above-mentioned comments and modifications apply accordingly. As an example, the specific situation is a campaign carried out by the client eg.. The potential value of the automatic adjustment performed at step 318 the same way as described above.

317 Check new information for event - step corresponds to steps described above, 309 and 313, respectively, and can be implemented with respect to the class Customer. The above-mentioned comments and modifications apply accordingly. An example of additional information can be eg. Commercial business markets.

318 Add value to Customer Experience counter - phase corresponds to the steps described above, 310 and 314, respectively, and can be implemented with respect to the class Customer. The above-mentioned comments and modifications apply accordingly.

319 Read the summary of all values Experience: Customer, Location, Service - At this point, read the class values with grades counters. Step 319, therefore, to give as starting classes Customer, Location and Service class-counter values. Each class will therefore be considered separately. This is done after each event, or event when the original has passed through the entire chain weighting. According to another embodiment, step 319 is automatically performed in batch mode at specified time intervals or whenever a prescribed number of new eventtejä has arrived in the buffer.

At this point must pay attention to what time window looking at things. Events can be monitored eg. At intervals and by seconds, minutes, hours, or days, in which case, of course, for consideration accumulate different amounts of values. If, however, collected a certain number of values, look at the time interval varies, which must also be taken into account.

In one embodiment, for example, provides 4 different value every time each of the dimension, and yet each individual instance.:

Location Experience The amount Over 1 second

LocationExperiencesummaoverlminute

Location Experience Sum Over lhour LocationExperiencesummaoverlday Service Experience The amount Over 1 second S ervi ceExp erience_summa_over_ 1 minute Service experience the sum over l hour S ervi ceExperi ence_summa_over_ 1 day Customer Experience amount Over 1 second CustomerExperiencesummaoverlminute Customer experience the sum over l hour CustomerExperiencesummaoverlday

If the application is even 100 customers, 1,000 and 100 to the location of the service so the new values generated a ferocious amount of time per day:

CustomerExperiencesummaxx: - 100 pieces _over_day - 2400 pcs over_hour - 144,000 shares over hour - 8.64 million pieces over_second Location_Experience_summa_xx: - 1000 _over_day - 24000 pieces over_hour - 1440000 pcs over hour - 86.4 million pieces over_second ServiceExperiencesummaxx: - 100 pieces _over_day - 2400 pcs over_hour - 144000 pc over hour - 8640000 pcs over_second 320 Threshold exceeded? - At this stage, compared to the class-counter values obtained in step 319 is compared to the corresponding class-specific vertical thresholds. If the class-specific vertical threshold value is exceeded, proceed to step 321. The comparison takes into account of course, consider the time window.

321 Create Location, Customer Service or event - at this stage is formed by a class-specific event (event) for each class, a class-specific vertical threshold value was found to have been exceeded in step 320. The class-specific event illustrates that relates to that category of disorder. Although the analysis time windows can be of varying lengths and a new event is created the same priorities.

In one embodiment, the weight for using the same weighting factor as the original event in each of the weighting of (0-5).

In one embodiment, the Severity value is always a fixed value, e.g. 5, wherein the possible values would be 0, 5, 10, 15, 20 and 25.

In another embodiment, the Prestored Severity is a parameter that can be adjusted between 0-5 and it can be set separately for each class.

Sure, these numerical values are merely examples. The values can be selected as desired and the application object needs. The application form can also easily eg. To scale values 0-75, thus avoiding the need of different thresholds stages 323, 326 and 329.

323 Bigger than reactive? - At this point it is checked whether higher reference value from step 321 or as large class of reactive than the interference threshold value. If so, proceed to step 324 and 325. If the comparison value is not greater than the interference threshold value of the class of reactive, proceed to step 326.

324 Store of value - reference value is entered into a database.

325 Reactive Create Event - Forging a new event, which is multiplied by the type of reactive. Event may include all along the way the weight values obtained. This can be, Location, Service and Customer-based all-new event, or then the original event that has come through all the accents and additional information Appendices to this. In a similar manner also checked describing the impact of the event a combination of the value obtained in step 332. Created event description (event) can then be sent eg. The operator staff for further action. Since this is a reactive class event, the disorder requires immediate action and the system is designed to trigger an appropriate degree of alarm.

326 Bigger than Proactive? - At this point it is checked whether higher reference value from step 321 or as large as the interference class Proactive threshold value. If so, proceed to step 327 and 328. If the comparison value is not greater than the interference threshold value of the Proactive class, going to step 329.

327 Store of value - reference value is entered into a database.

328 Create Proactive event - Forging a new event, which is multiplied by the type of Proactive. This corresponds to step 325, and the difference is only the disturbance class.

329 Bigger than Informative? - At this point it is checked whether the comparison value from step 321 greater than the interference threshold value of the class Informative. If so, proceed to step 330 and 331. If the comparison value is not greater than the interference threshold value of the class Informative, the performance of the method of the leg ends.

330 Store of value - reference value is entered into a database.

331 Create Informative event - Forging a new event, which is multiplied by the type of Informative. This corresponds to steps 325 and 328. The only difference is in fault class.

332 Enrich weighted event -Identifies describing the effect of a combination of the categories value, which is calculated by summing all the values Customer, Location, Service Experience. This ratio can be scaled to range the same as those described above, the class-weighted values. According to one embodiment, for example, is calculated by summing the values in each category:

Severity * Weight (value of 0-25) and dividing the result by the number of categories, in this embodiment, three. This is where the event has, therefore, gone through all the accents and has received the final set of weighted values. At this stage, therefore, what is the interaction classes for that particular event. At the same time can be set to any new additional lines 309, 313, 317.

The combination description taken for comparison and classification step 323. This can look to whether the effect of the combination informative, proactive or reactive, and, if necessary, on this basis to create a new event, which is included in the additional information and weighting.

Application forms makes it possible to handle a large number of events and event descriptions (eventtejä).

The embodiment of Figure 3, the classification took place in accordance with the following dimensions: 1. The principal dimensions: customer, place and service.

2. Response sensitivity divided into three Time dimension: Informative, Proactive and Reactive.

Service Manager can be also used in other dimensions. At other sizes used could be e.g. tekniikkaorientoituneessa environment server, database, switch or the like.

Event (event) may be, for example, a single alarm, the measured variable, issue tracking, CDR or call.

The application form offers significant advantages compared to the previously known solutions: 1. To avoid the error situation by emphasizing eventtejä proactively in different dimensions.

2. Focus on the right eventtejä, when the situation is reactive and able to limit the impact.

3. External parameters can be influenced by individual customer, place, or event processing service 4. History information can be exploited to influence the processing of individual customer, location or service event.

5. Create a new eventtejä based on the different dimensions of values.

The following describes with reference to Figure 4 of example, with enrichment of a single event (event). In this embodiment, the major dimensions of the customer, the place and the service and responsiveness is divided into three Time dimension as above, ie the dimensions are informative, proactive and reactive.

Impulse for carrying out the method can serve as a single event or event description (event or event record), alarm, measurement variable, issue tracking, CDR or call.

Record related to the incident is passed through the system and it is emphasized in all three dimensions (customer, location, and service) weighting coefficients. The system is also stored for each disorder severity value (Severity), which is defined as a stable solid and is provided on technical grounds. This fixed value takes into account the processing of the record. Severity can contain the desired number of categories and corresponding values of the gravity. According to the classification corresponding event record is then added to the severity of the value, eg. A critical failure, a serious defect, service level, etc. descending disorder.

The embodiment of Figure 4 Weight is dynamically changeable weight factor, which can get a value from 0-5. Dynamically alterable herein means that the customer or the service provider can change the weight coefficient also temporarily provide the desired sensitivity.

A record of the description of the event (the raw event) in accordance with the system of Figure 4 arrives and is passed through the system as illustrated in Figure 3. Relating to each class method stages of record is responsible for the relevant category weighted class-specific severity of the value of the record describing the event. Category-specific value of severity (Customer Experience, Location Experience, Service Experience) can be calculated, for example, by multiplying the value of the event severity (Severity), or class specific weighting factor (Customer Weight, Weight Location, Service Weight). In this case, therefore, the following calculation must be carried out:

Customer Weight * Severity = Customer Experience

Location Weight * Sev cou = Location Experience

Service Weight * Severity = Service Experience

Record filming the event can be calculated by the so-called. Horizontal threshold (Horizontal Trigger), for example, by adding up the above-mentioned class-specific weighting factors. In this case, therefore, the following calculation is performed:

Customer Experience + Location + Service Experience Experience = Horizontal Trigger

Horizontal threshold (Horizontal trigger) an image of an individual event, the magnitude of the adverse reaction, and in this example, it can get a value from 0-75.

Horizontal threshold value can now be classified as severity categories to select the necessary follow-up treatment and to facilitate the evaluation. In this example, the horizontal threshold value are classified according to the following limits: • If the horizontal threshold value is less than 20, the severity of the class is informative • If the horizontal threshold value is less than 35, the severity of the class is proactive • If the horizontal threshold value is less than 50, the severity of the class is reactive

Examples of operational conditions may be mentioned, for example, a major public event organized in a certain area, eg. Trade show, or other event, such as a car crash or the authorities of a situation that you want to follow more closely the normal level of service a particular area. In this case, Location Weight parameter can be increased for the duration of the event for the region and thus get the operator's system to react more readily available to the event area Disorder information. This will help if necessary to focus resources quickly to that area to safeguard the level of service.

Other similar use cases can be eg. A customer-specific focus. An example would be to think about even though trade opening ceremonies, when the events related to the client (event) can be weighted and thus more likely to respond to them in a given time.

Similarly, can also be service-specific temporary responsiveness lift. This could be due to e.g. version upgrade terminals or the like.

Other reasons to increase system sensitivity can be eg. Severe weather. It was whatever reason, the system allows the system to adjust the sensitivity by setting the weighting coefficients to the values as desired.

The measures described above in connection with Figure 4 can be carried out, for example, the system of Figure 2 and performing the method shown in Figure 3. Described by Figure 4, the logical structure shown in Figure 2 is carried out decision engine through 23. The records will be handled online 4 structure analysis module of Figure 21. Location Analysis Engine 223 calculates the location of the seriousness of the value of each category of by (Location Experience), service analysis engine 224 calculates for each category of seriousness of the value of the service (Service Experience) and customer analysis engine 225 calculates the value of each category of severity according to the customer's (Customer Experience). After processing the work of the Analysis The motto of the femur system can produce a variety of event descriptions and forward it through the 5 output interface.

The following describes with reference to Figure 5 by way of example, category of vertical thresholds based on the received accounting events (event). In other respects the embodiment of Figure 5 corresponds to the embodiment of Figure 4 above. Vertical threshold values (Vertical Triggers) is formed by adding together the events given the class-gravity of the values obtained during a certain reference period. The calculation is done, for example, by the following formulas: • through Customer Experience Trigger = Sum of all during the period under review on the system of the past records in Customer Experience worth it. Customer Experience Trigger is calculated from customer to customer, ie for each customer separately for targeting this client events.

• Location Experience Trigger = Sum of all during the period under review on the system of the past records Location Experience worth it. Location Experience Trigger is calculated for each region, ie the area to be inspected separately on the basis of targeted events in this region.

• Service Experience Trigger = Sum of all during the period under review on the system of the past records Service Experience worth it. Service Experience Trigger is calculated for each service separately considered, by means of targeted this service events.

Vertical thresholds can be dimensioida responsiveness to achieve the desired sensitivity and further processing. Thresholds and during the time the application is selected in accordance with the requirements. The example of Figure 5, some of the possible thresholds and the follow-up measures are described. For example, considering hour time window counter value exceeds 100 (Informative) sent to the release and the counter value exceeds 200 (proactive) measures are taken to avoid any possible damage caused by the interference of the situation. Such measures may include the introduction of back-up system for customers, or emergency power sources supplying certain areas.

Vertical thresholds can also be used eg. Events just cause interference detection and identification of the impact and influence of interference interference (customer, location, service).

Vertical thresholds may also be used e.g. controlling the communications of information. Eg. A certain sum of value can not start eg. Customer information and the amount grows, started the construction of back-up facilities to the customer within a given area are exported to reserve power. A certain amount at value can also create an event error issue tracking, and on the other, higher value eg. To send directly to the technician.

One embodiment of the use according to the example it is found that the vertical threshold (Vertical Trigger) Location field shows $ Experience begins to rise because the same area has several base stations that is the bottom off. The system it may be inferred that the region has an electrical fault. On the basis of this observation may send an error report to the electricity distribution company. Such an embodiment, therefore, provides the functionality of a mobile phone network at the time of using the automatic grid failures and describing the extent of the defect region report. This allows for an efficient alert system, which is fast-acting and helps electricity distribution companies to locate the fault and to devise remedies effectively.

According to one embodiment be implemented as. Triggering a dynamic (Dynamic Triggering). In this case, the system can be parameterized dynamically to the situation in all the different principal dimensions. The parameter adjustment, or layout may be taken by the person operation or system parameters can be set automatically according to the program.

One embodiment, the parameters are set automatically and dynamically. In this case, the system parameters are initially the default values and system to adjust the parameters of the received event information, and possibly also on the basis of other information. Such a system is, therefore, a learning system.

One embodiment of the procedure, so that will be monitored to each number associated with the category of events and the corresponding per category based on the number of detected weighting factor is changed. This can be done, for example, that the number of data collected over the review periods provided for information and compared to the number of previous or earlier periods observed numbers. Observation period, in turn, may be e.g. 5 minutes, 15 minutes or an hour or 4 hours. Evaluation period can of course be chosen freely into something other. The review period may be, for example, a week or a month, or a minute.

Changing the weighting factor can be done, for example, that the class-specific weighting factor is increased by one step when it is detected that the number of the class received during the period of events, for example, 20% higher than the corresponding number in the previous period. Similarly, the class-specific weighting factor calculated by one notch when it is detected that the number of the class received during the period of events, for example, 20% less than the number corresponding to the previous period. The number of the previous period can also be replaced comparison, eg. The numbers from February to October of the previous period moving average.

Such an arrangement, the system sensitivity can be weighted dynamically in such a way that the emerging and growing problem situations can be detected more easily. The system thus raise the self-respect of sensitivity to cumulative problems and thus a way to filter out individual error messages caused by incidental factors. For example, if a particular area begins to emerge more disruption events, this may be a sign of poor condition of the system, improper sizing or other problem, which is good to quickly intervene in downtime and prevent damage.

Dynamic and automatic layout weighting factors obviously does not preclude the possibility that the weighting coefficient has given new values manually. The weighting factor can therefore also raises a person working in, for example, because of some public event. After this process, the weighting factor will start again to be directed to dynamically in accordance with the automated system.

The system can also be programmed, for example, so that the periods in the event the numbers stored in a database and retrieved correlations and repeating patterns. For example, it can be seen that in one location area is as a rule a greater number of events in a specific time, for example at 8-9. The system can then increase the weighting coefficients on the basis of observed regularities, ie an example of the case of, for example, increase the weight by one notch at that location in the area before 8am and again reduced after 9am.

It is therefore possible to implement a method of keeping track of events per class received numbers and automatically modifying at least one weighting factor based on the change observed per class received the number of events.

In one embodiment, the classroom weighting factor is increased automatically in response to the number of received events on the class detected increased compared to earlier. Classroom weighting factor can also be reduced only opened automatically in response to the number of received events on the class detected in the reduced compared to the past.

Each of the current flowing through the system event may therefore also change the weighting coefficients applied to future events.

Regardless of whether a contract change in the weighting coefficient automatically, or the person served basis or in both ways, change the weighting coefficients helps position the new situation, the sensitivity of the system to better vastaavaksi.Esimerkiksi case, the fair horizontal threshold (Horizontal trigger) can be made more sensitive to the weighting temporarily customer and location categories associated with participation in trade fairs (Customer and Location dimensions) by raising these weighting coefficients.

One working example that shows that mobile network datansiirtopalvelue provided in one area of fault messages. It may be inferred, for example, the fact that in that area the payment terminal service will slow down and this can be informed, for example, moving taxis in the area and, if desired, with other terminals operating in the territory holders.

According to one embodiment therefore provided a method for automatically detecting and classifying the technical system disturbances in the computer system. Observation of the technical system can be, for example, the electricity distribution system, the payment terminal system, mobile communication system, or a dial. In general observation system can be any system that can be connected to a computer system so that the computer system may be a system event data.

To the embodiment, the method is received records (e.g. event record) containing information about events (event), and identifies the records containing information provided for the classification and at least one offset information, which is interpreted as a sign disorder. Other information is also stored, analyzed, and used to establish priorities and limits. Further, the method of picking up the classification of information as specified in the identified records that contain category information associated with each of the specified event class. Thereafter, for each identified record contained in the deviation data is assigned by the severity of the value and emphasis on the seriousness of the value extracted from the class of this information by the respective weighting coefficients. Here's how to form the class-severity values weighted.

As already stated above, the weighting coefficients can be changed dynamically, and even in such a way that successive same class of data / events you may have a different emphasis.

Thereafter, the process is calculated based on the weighted class-severity values for a set of reference values and compared with reference values of the respective limit values. Based on comparison deciding on further measures in accordance with the classification procedure.

In one embodiment, the records are received telephone calls and / or data communications network, network elements.

In one embodiment, the classification information includes at least the following categories: customer, place and service. Directory information respectively identifying at least the customer associated with the event, location of the event or the location area and the service in question.

In one embodiment, the reference values include each calculated to the identified record in the so-called horizontal threshold, which is formed on the basis of the corresponding record of the weighted class-gravity of the values.

In one embodiment, the reference values include a first class data set for each class knowledge lowered so-called vertical thresholds, which are formed on the basis of all of the identified records which have been received and / or processed during a certain period of time and that include the class information.

In one further embodiment of the preceding embodiment, the first class data set comprises at least two first class to the class information and at least two of the first class information differing from another class.

In one further embodiment of the preceding embodiment, the first class data set comprises at least two class information in each of the following categories: a customer, location, and service.

In one further embodiment of the preceding embodiment, the first class data set is the same as the set of all category information.

In one embodiment, the vertical threshold value is formed of the weighted class-class information corresponding values on the basis of the severity.

In one further embodiment of the preceding embodiment, corresponding to the vertical threshold limit values provide vertical threshold values häiriöluokaksi one of three categories. Of these, the lowest fault class is intended for informative system administrators for the measures to the middle class of disorder within the meaning of proactive prevention of actual service prejudicial interference and highest fault class is intended to illustrate that hinder service disruptions which require an immediate response.

Along with the preceding embodiment further embodiment, the at least one weighting factor is changed for a certain period to change the method of responsiveness during this period of the relevant category of information concerned.

In accordance with one embodiment is provided a system and a computer software for implementing the methods described above.

According to one embodiment, a computer implemented system for automatic detection and classification of the technical system malfunctions.

According to one embodiment of the system is part of the telephone and / or data communications network and is adapted to receive the records of telephone and / or telecommunications network and network elements to implement the server system relating to the detection and classification of telephone and / or telecommunications network.

According to one embodiment the system is adapted to provide a user interface through which the weighting factors are set at desired at each time point. The user interface can be provided at the choice of the most desirable and functionalities in addition to the system administrator for the client so that the client can, for example, to change the weighting coefficients related to its own services.

According to one embodiment the system is adapted to provide a user interface, which in turn is arranged to provide class-perspective presentation of the technical system for each category of operating condition.

On the basis of the above examples it is clear that within the framework of the invention, numerous solutions differing from the embodiments described above can be implemented.

The invention is therefore not intended to be limited to the examples described above, but instead the patent protection should be examined within the full scope of the appended claims.

Claims

Method for automatically detecting and classifying noise states in a telephone and / or mobile communication network by means of a computer system (2) connected to the telephone and / or mobile communication network or which contains the telephone and / or mobile communication network, characterized in that : - at least from the telephone and / or mobile communications network's network elements, items containing event data are received; - items containing defined classification information and at least one deviation information, which can be interpreted as a sign of a disturbance, are identified; - from the identified entries, the defined classification information is selected, which contains class information for each defined class related to the event; - a critical value is assigned to the deviation information that each identified item contains and this critical value is weighed with weighting coefficients corresponding to the selected class information to form class-specific critical values; - on the basis of the weighted class-specific critical values, a group of comparison values is calculated; - the comparison values are compared with the corresponding limit values; and - on the basis of the comparisons, decisions on continuation measures are taken in accordance with a classification of measures.

Method according to claim 1, characterized in that the entries are received from the network elements of the telephone and / or mobile communication network.

Method according to claim 1 or 2, characterized in that - the classification information comprises at least the following classes: customer, position and service; and - the class information correspondingly specifies at least the customer related to the event, the event's position or location and the service in question.

Method according to claims 1-3, characterized in that the comparison values contain a so-called horizontal threshold value for each identified item, which threshold value is formed on the basis of the weighted class-specific critical values.

Method according to any one of claims 1-4, characterized in that the comparison values contain so-called vertical threshold values calculated for each class information of a first class information group, which threshold values are formed on the basis of all identified items received and / or processed during a specific time period and containing the class information in question.

Method according to claim 5, characterized in that the first class information group comprises at least two class information in the first class and at least two class information in the second class which differs from the first class.

Method according to claim 5 or 6, characterized in that the first class information group comprises at least two class information from each of the following groups: customer, position and service.

Method according to claims 5-7, characterized in that the first class information group is the same as the group of all class information.

Method according to any of claims 5-8, characterized in that the vertical threshold value is formed on the basis of the weighted class-specific critical values corresponding to the class information.

Method according to any one of claims 5-9, characterized in that the limit values corresponding to the vertical thresholds give the vertical threshold values as a disturbance class from one to three classes, of which the lowest disturbance class is intended to be informative to the system maintainer, the middle disturbance is intended for predictive measures to prevent a disturbance that damages the actual service, and the upper noise class is intended to describe disturbances, which damage the service and which require an immediate response.

Method according to any one of claims 1-10, characterized in that at least one weighting coefficient is changed for a specific period of time in order to change the reaction sensitivity of the method during that time period for the part of the class information in question.

Method according to any one of claims 1-11, characterized in that the class-specific number of events received is observed and at least one weighting coefficient is automatically changed class-specific on the basis of a change observed in the number of events received.

Method according to any one of claims 1-12, characterized in that a class' weighting coefficient is automatically raised in response to the fact that the number of received events relating to a class is observed to have grown in comparison with previous numbers.

Method according to any one of claims 1-13, characterized in that a class' weighting coefficient is automatically lowered in response to the fact that the number of received events relating to a class is observed to have decreased compared to previous numbers.

15. Computer system (2) for automatic detection and classification of disturbance states in a telephone and / or mobile communication network, characterized in that the system is arranged to: - receive at least from the telephone and / or mobile communication network's network elements, containing event data; - identify items containing defined classification information and at least one deviation information, which can be interpreted as a sign of a disturbance; - select from the identified records defined classification information, which contains class information for each defined class related to the event; - assigning a critical value to the deviation information contained in each identified record and weighing that critical value with weighting coefficients corresponding to the selected class information to form class-specific critical values; - on the basis of the weighted class-specific critical values, calculate a group of comparison values; - compare the comparison values with the corresponding limit values; and - on the basis of the comparisons, decide on continuation measures in accordance with a classification of measures.

System according to claim 15, characterized in that it forms part of a telephone and / or mobile communication network and is arranged to receive records from the telephone element and / or mobile communication network's network elements and realize detection and classification in a server system in connection with the telephone. - and / or the mobile communications network.

System according to claim 15 or 16, characterized in that it is arranged to implement a method as defined in at least one of claims 1-14.

System according to any one of claims 15-17, characterized in that it is arranged to provide a user interface via which the weighting coefficients can be set to desired values at each time point.

System according to any one of claims 15-18, characterized in that it is arranged to provide a user interface, which in turn can be arranged to provide class-specific windows for presenting the class-specific operating state of the technical system.