CN105704335B - Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process - Google Patents
Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process Download PDFInfo
- Publication number
- CN105704335B CN105704335B CN201610118087.1A CN201610118087A CN105704335B CN 105704335 B CN105704335 B CN 105704335B CN 201610118087 A CN201610118087 A CN 201610118087A CN 105704335 B CN105704335 B CN 105704335B
- Authority
- CN
- China
- Prior art keywords
- data
- module
- call
- time
- statistics
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/22—Arrangements for supervision, monitoring or testing
- H04M3/36—Statistical metering, e.g. recording occasions when traffic exceeds capacity of trunks
- H04M3/367—Traffic or load control
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
- H04M3/5175—Call or contact centers supervision arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
- H04M3/523—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing with call distribution or queueing
- H04M3/5232—Call distribution algorithms
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
- H04M3/523—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing with call distribution or queueing
- H04M3/5238—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing with call distribution or queueing with waiting time or load prediction arrangements
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Business, Economics & Management (AREA)
- Marketing (AREA)
- Telephonic Communication Services (AREA)
Abstract
The present invention relates to a kind of predictive outbound algorithms based on dynamic statistics process, interchanger dialing method and device, the inventive system comprises data input modules, predictive outbound algoritic module, interchanger, dynamic statistics module, data memory module and data export display module, data input module is to pretest outgoing call algoritic module input initialization supplemental characteristic, pretest outgoing call algoritic module is connected with interchanger, carry out information exchange, dynamic statistics module is connected with data memory module, receive the data of data memory module, the statistical data of dynamic statistics module statistics different frequency is simultaneously transferred to pretest outgoing call algoritic module, it is displayed data simultaneously by data output interface.The present invention is counted and is predicted according to the historical data made a phone call, and dynamically determines the telephone number that lower a moment should dial, and utilization rate of attending a banquet is improved while being no more than the target phone percent of call lost as far as possible, reaches optimization balance and completely substitution manual dial.
Description
Technical field
The present invention relates to telephone outbound call technical fields, and in particular to a kind of predictive outbound calculation based on dynamic statistics process
Method, interchanger dialing method and device.Statistics, data mining, queueing theory, Markov chain, raw go out opinion and calculation is utilized
The technology of the technologies such as method design and analysis, especially real time resources allocation schedule and feedback control optimization.
Background technique
Call center can be divided into incoming call type and two kinds of call-out type.Incoming call type call center provides the service of passive type,
And call-out type call center provide actively be customer service a kind of mode, carry out the marketing, questionnaire survey etc. for enterprise
Business provides a kind of efficient mode.With the development of computer and the communication technology, the call center system based on traditional technology
It has been difficult to meet turn of the market demand, automatic outer call is given birth to therefrom, and automatic outer call system actively initiates calling, will be unable to connect
Clients filter, the client being successfully connected is distributed to idle customer service, to improve the utilization rate of contact staff.
Existing outgoing call algorithm mainly has three classes: experience outgoing call prediction algorithm, integral model algorithm and queuing theory are pre-
Application in method of determining and calculating.
Experience outgoing call prediction algorithm is to adjust outgoing call by carrying out dynamic regulation to feedback rates value to outer group of recent result
Adjusting parameter (hypervelocity is outer when dialling>1, slow down outer dial when<1), so that the adaptability of algorithm is become strong.The disadvantage is that averagely idle duration point
Cloth is unstable, and call loss cannot be controlled preferably.
Integral model algorithm examines or check the connecting timing distribution situation of calling first: percent of call completed is distributed V (t), and rate of taking out stitches is distributed F
(t), then haveExamine or check the time series of the duration of call again: the duration of call is distributed h (t), meetsThen the duration of call is greater than the probability of time t, H (O)=l, H (+∞)=0.Assuming that system exhalation sequence C (t), then
In the access calling number of t moment, the current talking calls being somebody's turn to do in t indicates the calls being currently talking, including oneself passes through
With operator call calling and user oneself through off-hook it is unallocated to operator waiting calling.If it is desired to breathe out user's off-hook
Instant percent of call completed afterwards is g, i.e. the available free proportion of calls attended a banquet after exhalation user's off-hook, then just has N (t) * g < A, be simultaneously
Resource of attending a banquet is made full use of, and having N (t) >=A, A is the number of attending a banquet currently registered.It is i.e. available from above: A≤N (t) < A/
g.The exhalation time sequence C (t) of system can be calculated from this constraint condition.It being capable of basis according to this exhalation prediction algorithm
The call time sequence and current system resource situation of the result of history call, that is, previous carry out self feed back study, dynamically adjust
Whole subsequent exhalation.Although algorithm as above is really feasible, and plays a significant role in practical applications, found through analysis
Still having is worth improved place, for example it is optimal selection that n batches are divided into cycle T, and it is true that average handling time, which calculates L/2,
Do are situation or ideal state unalterable or have fluctuation, and fluctuation is how many? either the present situation extremely can or can not
Under the influence of a moment prediction, how is stability?
Application of the queuing theory in prediction algorithm: queuing theory (queuing theory) is existed by A.K. angstroms of erlang earliest
It solves to introduce when automatic telephone design problem.By prediction algorithm be summarized as solving the problems, such as two it is related: 1, sit maximizing
In the case where the busy factor of seat (utilization rate of attending a banquet), the percent of call lost is not allowed to be greater than an acceptable maximum value ARmax.2, to the greatest extent
In the case that amount reduces the percent of call lost, less than the one acceptable minimum level Umin of the busy factor that attends a banquet is not allowed.Algorithm is exactly
Meet the two above conditions and generates the number N for working as the cause calling.The algorithm can be collected or to receive system more relevant
Parameter: it the total number m that attends a banquet, throws phone arrival rate p into, average service time T, hit rate ρ.The algorithm assumes exhalation stream clothes
From the Poisson distribution of parameter lambda；Service time is the general distribution for obeying parameter μ within the system.In the prediction algorithm, it will exhale
The distribution flowed out is set to Poisson distribution, i.e., the time between the phone twice of adjacent dial-out just obeys exponential distribution.This and reality
In the case of the phenomenon that have very big discrepancy.In addition, service time obeys general distribution, find under study for action, the time meeting of service
Two stages are concentrated on, one is 3-10s, is not the general distribution an of standard between another 140-160s, this will make
Deviation at the calculated result of the model is larger.In addition, the state attended a banquet at present is not added in consideration by the algorithm, this is just
To will cause it is all suddenly attend a banquet that all the line is busy in the case where, appoint and so calculate number by former data and called, until mean time
Between T retrieved calculating (taking a long time), therefore will generate one section for a long time upheaval.A period of time is serviced
The case where time is reduced suddenly will also result in the upheaval that utilization rate of attending a banquet becomes smaller suddenly.
By being analyzed above as it can be seen that existing predictive outbound algorithm is mostly coupled with dynamic statistics part, even
There is no dynamic statistics part.And the most important determinant of predictive outbound algorithm is dynamic statistics part, existing outgoing call is calculated
Method some be to be obtained using theoretical formula method, it is desirable that establish a kind of mathematical model in advance, parameter value later is all logical
It crosses obtained by theoretical formula method；Although this method does not have to or seldom accesses database, establishes model and relatively fix, for
Some business, this model are no longer set up, and the adaptability of this model is poor in other words.Existing outgoing call algorithm also some
It is to be realized by dynamic statistics, but this method wants a large amount of access history data to carry out machine learning, training data
Either neural network etc., this method continually access database and need a large amount of data, waste many times, make
Decline at performance, and at the beginning without being difficult reasonably to be predicted when mass data.
Summary of the invention
In order to overcome above-mentioned defect existing in the prior art, the object of the present invention is to provide one kind to be based on dynamic statistics mistake
Predictive outbound algorithm, interchanger dialing method and the device of journey.
To achieve the goals above, according to the first aspect of the invention, the present invention provides one kind to be based on dynamic statistics
The predictive outbound algorithm of process comprising following steps:
S1 starts, and setting target is attended a banquet utilization rate AO, maximum percent of call lost ARmax, number of switches T, maximum waiting time
TQmax, time in predicting interval △ t, waiting list maximum length qL, user's patience 1/ η of time, user is reluctant that waiting list is added
Probability p0, and the initial value of quantity λ is made a phone call outward, enable the accumulated value al=0 of l.
S2 obtains current customer service sum N by interchanger in t moment, and idle customer service sum f is calling telephone number j,
Percent of call completed ρ, average 1/ β of jingle bell duration, the numerical value of 1/ μ of average call duration and waiting list length q, the t are positive number；
S3 judges whether to meet q < qL, goes to step S4 if meeting, and otherwise suspend mode to t+ time Δt goes to step S2 progress
Next round prediction, and current time is updated for t=t+ Δ t；
S4, minimum needed for calculating are attended a banquet several M, calculation formula are as follows:
S5 judges whether to meet N≤M, goes to step S17 if meeting, otherwise go to step S6；
S6, computing system predict maximum outgoing call value l, cumulative calculation al=al+l, wherein
S7, the percent of call lost AR at dynamic statistics last moment simultaneously judge whether to meet AR < ARmax, S8 is gone to step if meeting,
Otherwise S16 is gone to step；
S8 judges whether to meet current idle customer service sum f > al, goes to step S9 if meeting, otherwise go to step S13；
S9, prediction result s=(f-j)/ρ-q+al, al sets 0 if al > 1；
S10 calculates period of reservation of number Tq, Tq=q/ (N* μ)；
S11 judges whether to meet Tq < TQmax, S12 is gone to step if meeting, otherwise suspend mode to t+ time Δt is gone to step
S2 carries out next round prediction, and updates current time for t=t+ Δ t；
S12, updates prediction result s=s*ARrate, and ARrate is that the exhalation of system weakens ratio, ARrate=1-AR, AR
For the percent of call lost,
AR=p0B(θ,N)+(1-p0) λ β ρ η, go to step S18；
S13, prediction result s=f-j-q+al, al sets 0 if al > 1；
S14 calculates percent of call lost AR, AR=p0B (θ, N)+(1-p0)λβρη
B (θ, 0)=1,
S15 judges whether to meet AR < ARmax, S10 is gone to step if meeting, otherwise suspend mode to t+ time Δt is gone to step
S2 carries out next round prediction, and updates current time for t=t+ Δ t；
S16, prediction result s=f-j-q, goes to step S18；
S17, prediction result s=f-j, goes to step S18；
S18 is calculated and is finally made a phone call quantity λ, λ=(int) min (max (s, 0), T)；
S19 dials λ phone outward, and algorithm suspend mode to t+ time Δt goes to step S2 and carries out next round prediction, and updates
Current time is t=t+ Δ t；Above step is repeated, is run until application program terminates.
To achieve the goals above, according to the second aspect of the invention, the present invention provides one kind to be based on dynamic statistics
The interchanger dialing method of the predictive outbound algorithm of process, it is characterised in that include the following steps:
Step 1, predictive outbound algorithm according to claim 1, which predicts subsequent time, will make a phone call quantity
λ；
Step 2, interchanger dials λ phone according to this prediction result；
Step 3, after average 1/ β of ring time, the β is positive number, has the closing of the circuit of ρ ratio, is left 1- ρ ratio
The phone access failure of example is filtered, and the ρ is the positive number greater than 0 less than 1；
Step 4, whether the number metering-in control system of connection, control system available free can attend a banquet point according to current
Match；
Step 5, distribution is attended a banquet and is serviced for it if available free attend a banquet, service finish be dial it is successful；If not being free
Seat is sat back for a chat or enjoy one's leisure, the client for having p0 probability is ready that waiting list, which is added, to be waited, and the client of remaining 1-p0 probability is unwilling to add
Enter waiting list to abandon forming call loss；
Step 6, the client of queuing also some abandon forming call loss due to waiting more than its patience 1/ η of time, only
Have it is remaining can be idle according to what is serviced at present by control system not above its patience 1/ η of time
It attends a banquet and distributes the service of attending a banquet for it, finally service is completed.
To achieve the goals above, according to the third aspect of the present invention, the present invention provides a kind of operation bases of the present invention
In the device of the predictive outbound algorithm of dynamic statistics process comprising data input module, is handed over predictive outbound algoritic module
It changes planes, dynamic statistics module, data memory module and data output display module, the data input module is to predictive outbound
Algoritic module input initialization supplemental characteristic, the predictive outbound algoritic module are connected with interchanger, carry out information exchange, institute
It states dynamic statistics module to be connected with data memory module, receives the data of data memory module, the dynamic statistics module statistics
The statistical data of different frequency is simultaneously transferred to predictive outbound algoritic module, while carrying out data by data output interface and showing
Show；
The predictive outbound algoritic module includes configuration parameter unit, initiation parameter unit, theoretical calculation data sheet
Member, period statistical parameter unit, predictive outbound algorithm unit, prediction result unit；The configuration parameter unit is for receiving
Interchanger configuration parameter information, the initiation parameter unit various target component numbers that input module inputs for receiving data
According to the theoretical calculation data cell is used to receive theoretical calculation but wants the data of cyclomorphosis, the period statistical parameter list
Member is for receiving the various data that dynamic statistics module is come out with different statistic frequencies, the predictive outbound algorithm unit
For realizing predictive outbound algorithm described in claim 1, the prediction result unit is used to transmit prediction knot to interchanger
Fruit；
The dynamic statistics module include obtain historical data unit, selection statistics policy unit, statistics policy unit, to
Predictive outbound algoritic module transmits the unit of statistical data and the unit to output interface transmitting statistical data, and the acquisition is gone through
Be stored in history data cell including obtain written data library data and interchanger save service but also it is following must
And the data in write-in database, the selection statistics policy unit is for selecting using which kind of statistics strategy, the statistics plan
Slightly unit includes a variety of specific statistics strategies.
The present invention has following advantageous effects:
1, the present invention is counted and is predicted according to the historical data made a phone call, and dynamically determines that lower a moment should dial
Telephone number, improve and attend a banquet utilization rate while being no more than the target phone percent of call lost as far as possible, reach optimization balance and complete
Replace manual dial.
2, queuing theory model when outgoing call algorithm of the invention foundation attends a banquet not enough, and ring time, service time
With statistics strategy then using the statistical model of manual dial data, the model of foundation is similarly being exhaled closer to actual conditions
Utilization rate of attending a banquet under loss rate can greatly improve.Outgoing call algorithm algorithm of the invention is based on queuing theory, but initial parameter is (as serviced
Time, percent of call completed etc.) it is very big for different business difference, so these parameters by dynamic statistics can be well adapted for it is various
Business.Utilization rate of attending a banquet in the case where number of attending a banquet is 10 has 65-70%, percent of call lost 3-4%；In the case where number of attending a banquet is 20
Utilization rate of attending a banquet has 70-73%, the percent of call lost 2% or so；Utilization rate of attending a banquet in the case where number of attending a banquet is 30 has 72-75%, exhales
Loss rate 1-2%；Utilization rate of attending a banquet in the case where number of attending a banquet is 40 has 80-85%, the percent of call lost 1% or so；It is 50 in number of attending a banquet
In the case where utilization rate of attending a banquet have 82-85%, the percent of call lost is less than 1%；And with several increase of attending a banquet, more and more steadily, exhale
Loss rate has just started with regard to very little even without fluctuation also very little；Hardly there is call loss when number of attending a banquet is 60 or more and keeps
Utilization rate of attending a banquet 80%.And there is good effect for different business datums.
3, the present invention, which opens, a kind of is divided into two important modules: predictive outbound algorithm mould predictive outbound algorithm
Block and dynamic statistics module；So as to optimize from both direction, the bottleneck that will appear so as to avoid only optimization algorithm is asked
Topic.And the specific statistics of parameter need not be concerned about in optimization algorithm, when statistical parameter without consider algorithm realization and whether
It is useful to algorithm.
4, joined dynamic statistics module, improve the adaptability of algorithm, can adapt to a variety of different business and
The too big situation of data fluctuations；The thinking combined using statistical model with theoretical model avoids theoretical model in extreme feelings
Deviate biggish defect under condition, enables theoretical model preferably close to statistical model.
5, the thinking combined using feedback with prediction, enables those to predict accurately to become more acurrate, and prediction is less
Accurately it can become accurate by feedback, thus the advantages of combining prediction and feedback.
Additional aspect and advantage of the invention will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
Above-mentioned and/or additional aspect of the invention and advantage will become from the description of the embodiment in conjunction with the following figures
Obviously and it is readily appreciated that, in which:
Fig. 1 is predictive outbound algorithm flow chart in a kind of preferred embodiment of the present invention；
Fig. 2 is the flow diagram of interchanger dialing method in a kind of preferred embodiment of the present invention；
Fig. 3 is the composition knot for inventing the device of the predictive outbound algorithm based on dynamic statistics in a kind of preferred embodiment
Structure schematic diagram.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, and for explaining only the invention, and is not considered as limiting the invention.
In the description of the present invention, unless otherwise specified and limited, it should be noted that term " installation ", " connected ",
" connection " shall be understood in a broad sense, for example, it may be mechanical connection or electrical connection, the connection being also possible to inside two elements can
, can also indirectly connected through an intermediary, for the ordinary skill in the art to be to be connected directly, it can basis
Concrete condition understands the concrete meaning of above-mentioned term.
Meaning of parameters in the present invention is explained as follows:
Ring duration: dialing a phone, is not turned on institute from starting to dial into a phone successfully to connect or identify
It takes time.
Service duration: the service time of the phone after successfully connecting.
It arranges duration: (a part that can also regard service time as) is done the business record time to client.
Waiting list length: phone is connected in the case where that cannot service, is ready to wait online, this queue length is just
It is waiting list length.
User's patience time: user Zi waiting list is added is reluctant that active is waited to abandon the time hung up the telephone to it.
Period of reservation of number: the time that user actually waits in queuing process.
Maximum waiting time: all clients waited online are ready the maximum time upper limit waited.
Customer service utilization rate: the service duration summation of all customer services and the ratio of the online total duration of all customer services.
The telephone number that the phone percent of call lost: connecting but is not serviced accounts for the ratio for connecting total call.
After the present invention carries out data statistic analysis according to actual artificial outgoing call dial data: service time approximation takes
From quantum condition entropy, ring time approximation obeys generalized Poisson distribution.
The present invention provides a kind of predictive outbound algorithms based on dynamic statistics process, as shown in Figure 1 comprising as follows
Step:
S1 starts, and parameter value, configuration parameter and initiation parameter including interchanger is arranged, for example, specially target is sat
Seat utilization rate AO, maximum percent of call lost ARmax, number of switches T, maximum waiting time TQmax, time in predicting interval △ t, waiting team
Column maximum length qL, user's patience 1/ η of time, user are reluctant that the Probability p of waiting list is added0, and number is made a phone call outward
Measure the initial value of λ.
In the present embodiment, workload, maximum percent of call lost AR are setmax=0.02, number of switches T=10000, mesh
Mark attend a banquet utilization rate AO=0.8, maximum waiting time TQmax=15s, time in predicting interval △ t=5s, waiting list most greatly enhance
(capacity of queue) qL=2, user patience 1/ η of time=10s are spent, user is reluctant that the Probability p of waiting list is added0=0.01 equal mesh
Mark parameter.The numerical value of equal sign back is the default value of parameter, enables the accumulated value al=0 of l.
S2 obtains current customer service sum N by interchanger in t moment, and idle customer service sum f is calling telephone number j,
Percent of call completed ρ, average 1/ β of jingle bell duration, the numerical value of 1/ μ of average call duration and waiting list length q, the t are positive number；
S3 judges whether to meet q < qL, goes to step S4 if meeting, and otherwise suspend mode to t+ time Δt goes to step S2 progress
Next round prediction, and current time is updated for t=t+ Δ t；
S4, minimum needed for calculating are attended a banquet several M, calculation formula are as follows:
S5 judges whether to meet N≤M, goes to step S17 if meeting, otherwise go to step S6；
S6, computing system predict maximum outgoing call value l, cumulative calculation al=al+l, wherein
S7, statistics policy module utilize the percent of call lost AR at the statistical method dynamic statistics last moment prestored and judge whether
Meet AR < ARmax, S8 is gone to step if meeting, otherwise goes to step S16；
S8 judges whether to meet current idle customer service sum f > al, goes to step S9 if meeting, otherwise go to step S13；
S9, prediction result s=(f-j)/ρ-q+al, al sets 0 if al > 1；
S10 calculates period of reservation of number Tq, and Tq=q/ (N* μ), wherein q is the number being currently lined up, the table if q is 0
Show currently without waiting client；
S11 judges whether to meet Tq < TQmax, S12 is gone to step if meeting, otherwise suspend mode to t+ time Δt is gone to step
S2 carries out next round prediction, and updates current time for t=t+ Δ t；
S12, updates prediction result s=s*ARrate, and ARrate is that the exhalation of system weakens ratio, ARrate=1-AR, AR
For the percent of call lost, AR=p0B(θ,N)+(1-p0) λ β ρ η, go to step S18；
S13, prediction result s=f-j-q+al, al sets 0 if al > 1；
S14 calculates percent of call lost AR, AR=p0B(θ,N)+(1-p0) λ β ρ η, phone call loss, which can be divided into being reluctant to enter, to be waited
The sum of queue call loss is ready that two parts into waiting list call loss, first half only occur in the case where i=N, it is assumed that θ
=λ β ρ (1-p0)/μ is successful connection phone input load, according to Erlang-B formula, can calculate N number of customer service be not it is free
The probability π of idleN, i.e. the percent of call lost, this probability can obtain for crossing Erlang formula iteration:
B (θ, 0)=1,
S15 judges whether to meet AR < ARmax, S10 is gone to step if meeting, otherwise suspend mode to t+ time Δt is gone to step
S2 carries out next round prediction, and updates current time for t=t+ Δ t；
S16, prediction result s=f-j-q, goes to step S18；
S17, prediction result s=f-j turn rapid S18；
S18 is calculated and is finally made a phone call quantity λ, λ=(int) min (max (s, 0), T)；
S19 dials λ phone outward, and algorithm suspend mode to t+ time Δt goes to step S2 and carries out next round prediction, and updates
Current time is t=t+ Δ t；Above step is repeated, is run until application program terminates.
In the present embodiment, the percent of call lost AR at dynamic statistics last moment is from before current time when t1 in step S7
All phone numbers connected to the phone number of stage at current time call loss divided by this stage are carved, the t1 is positive number.
In more preferred embodiment of the invention, t1 be 15 minutes to 45 minutes between time, preferably 20 minutes.
If the interval chosen is too small, the data in the possible period are very little, are not enough to reflect true situation.If choosing
The interval taken is excessive, cannot the change of response parameter in real time be also not enough to reflect true and when data fluctuations are more apparent
Fluctuation situation.
The present invention has the advantage that and is established by raw go out by with queueing theory relative to traditional predictive outbound algorithm
Queuing theory model when attending a banquet not enough, and ring time, service time and statistics strategy are then using manual dial data
Statistical model, closer to actual conditions, utilization rate of attending a banquet under the same percent of call lost can greatly improve the model of foundation；This hair
Predictive outbound algorithm a kind of is divided into two important modules: predictive outbound algoritic module and dynamic statistics mould by bright open
Block.So as to optimize from both direction, both direction has many prioritization schemes, can go out so as to avoid only optimization algorithm
Existing bottleneck problem；And the specific statistics of parameter need not be concerned about in optimization algorithm, without considering algorithm when statistical parameter
It realizes and whether useful to algorithm；It joined dynamic statistics module, improve the adaptability of algorithm, a variety of differences can be adapted to
Business and the too big situation of data fluctuations；The thinking combined using statistical model with theoretical model avoids theoretical mould
Type deviates biggish defect in extreme circumstances, enables theoretical model preferably close to statistical model；Using feedback and prediction phase
In conjunction with thinking, enable those to predict accurately to become more acurrate, prediction less accurately can by feedback become accurate
A bit, thus the advantages of combining prediction and feedback.
Automatic outer call system is a very complicated calling system, is especially had to customer service utilization rate and the phone percent of call lost
It has in the case of asking.So a predictive outbound algorithm will not only consider that above-mentioned affecting parameters are also contemplated that system
Actual motion state parameter.Therefore the present invention proposes a kind of prediction of binding isotherm modeling and system real-time running state
Outgoing call algorithm is predicted, so that the customer service utilization rate and the phone percent of call lost of outer paging system reach certain requirement.
As shown in Fig. 2, using the predictive outbound algorithm of the invention based on dynamic statistics process, interchanger dialing method
Include the following steps:
Step 1, predictive outbound algorithm according to the present invention, which predicts subsequent time, will make a phone call quantity λ；
Step 2, interchanger dials λ phone according to this prediction result；
Step 3, after average 1/ β of ring time, only the closing of the circuit of ρ ratio, the phone of remaining 1- ρ ratio are not connect
It is logical to be filtered；
Step 4, whether the number metering-in control system of connection, control system available free can attend a banquet point according to current
Match；
Step 5, distribution is attended a banquet and is serviced for it if available free attend a banquet, service finish be dial it is successful；If not being free
Seat is sat back for a chat or enjoy one's leisure, the client for having p0 probability is ready that waiting list, which is added, to be waited, and the client of remaining 1-p0 probability is unwilling to add
Enter waiting list to abandon forming call loss；
Step 6, the client of queuing also some abandon forming call loss due to waiting more than its patience 1/ η of time, only
Have it is remaining can be idle according to what is serviced at present by control system not above its patience 1/ η of time
It attends a banquet and distributes the service of attending a banquet for it, finally service is completed.
As shown in figure 3, the present invention runs the device of the predictive outbound algorithm based on dynamic statistics process comprising data
Input module, predictive outbound algoritic module, interchanger, dynamic statistics module, data memory module and data output display mould
Block, data input module predictive outbound algoritic module and are handed over to predictive outbound algoritic module input initialization supplemental characteristic
It changes planes connected, carries out information exchange, dynamic statistics module is connected with data memory module, the data of data memory module are received,
The statistical data of dynamic statistics module statistics different frequency is simultaneously transferred to predictive outbound algoritic module, while being exported by data
Interface displays data.Predictive outbound algoritic module is for historical data, the dynamic of intermediate data according to last moment
Statistical result and preset parameter (including target component) and gross data are dialed result to lower a moment and are predicted, and according to
Prediction result decision makes a phone call or adds up parameter to lower a moment.Dynamic statistics module according to historical data for being calculated
The statistics of parameter or intermediate result needed for method, and the parameter of some measure algorithm indexs is carried out real-time Dynamic Display.
Predictive outbound algoritic module includes configuration parameter unit, initiation parameter unit, theoretical calculation data cell, week
Phase statistical parameter unit, predictive outbound algorithm unit, prediction result unit.Configuration parameter unit is configured for desampler
Parameter information (including number of switches, maximum waiting time, current queue number, currently available number etc. of attending a banquet), initialization
The parameters unit various object parameter datas that input module inputs for receiving data, theoretical calculation data cell is for receiving reason
By calculating but want the data of cyclomorphosis (including as theory AR etc, all formula that passes through are calculated but without using statistics
The data that method obtains are the data using theoretical calculation), period statistical parameter unit is for receiving dynamic statistics module use
The various data that different statistic frequencies come out, predictive outbound algorithm unit is for realizing predictive outbound algorithm, prediction
As a result unit is used to transmit prediction result to interchanger.
Dynamic statistics module includes obtaining historical data unit, selection statistics policy unit, statistics policy unit (statistics plan
Statistics strategy is slightly prestored in unit, such as according to week, month or festivals or holidays statistics, is classified according to morning and afternoon, according to
The sum of daily statistics of each period according to cumulative statistics strategy etc.), transmit and count to predictive outbound algoritic module
The unit of data and the unit that statistical data is transmitted to output interface are obtained to be stored in historical data unit and write including obtaining
What the data and interchanger for entering database saved is servicing but is also having not enough time to the data in write-in database, selection statistics plan
For slightly unit for selecting using which kind of statistics strategy, statistics policy unit includes a variety of specific statistics strategies.
The device of predictive outbound algorithm based on dynamic statistics of the invention, can be embedded in an existing interchanger
In, the interface that can also provide one and interchanger is individually disposed.Those of ordinary skill in the art will appreciate that realizing this
The process of the predictive outbound algorithm based on dynamic statistics of invention, can be completed, specific program can by programming
To be stored in a kind of readable storage medium storing program for executing of the device of predictive outbound algorithm based on dynamic statistics, the program is when being executed
Execute the correspondence step in the above method.The storage medium can be such as: ROM/RAM, magnetic disk, CD and hard disk.
The present invention can regard outer paging system as one queuing model, and telephone number in systems is regarded as and is become at any time
The function of change, then this system is exactly the Markov Chain of a Time Continuous, system setting maximum dials quantity (work
Amount) it is W, it is assumed that and call center, which has, N number of identical attend a banquet and allows to have capacity to be that c waits in line queue length and T
Interchanger.By to a large amount of manual dial's data of company data mining and theory analysis, telephone outbound call number is modeled as obeying
Parameter is the Poisson distribution of λ, and phone ringing duration obeys the Poisson distribution that mean value is 1/ β seconds, and phone is averaged percent of call completed as ρ, connects
The probability that logical client does not enter waiting list (directly hanging up) in the case where cannot get service scenario is p0, while the patient time of client
It is 1/ η seconds, the service time for connecting client obeys the quantum condition entropy that mean value is 1/ μ seconds.
Present outgoing call systematic procedure can be considered the Probability State conversion process under the Markov Chain of Time Continuous, X (t)
For total telephone number in t moment system, including the telephone number serviced and in waiting list.Then Time Continuous at this time
The state space E={ 0,1,2 ..., N+c } that is possessed of Markov Chain, for many queuing processes, client's
Reaching (state is caused to add 1) is represented by " life ", and leaving away and (state being caused to subtract 1) for client is represented by " dead ", then X (t)
When=k, when subsequent time state change, can only become k ± 1, therefore it is λ that whole process, which becomes " life " probability,k, " dead " is general
Rate is μkIndependent random process-birth and death process.
Wherein:
Probability distribution { the π of steady state can so be reached come computing system according to " life is gone out " process0,π1,…,
πN+c, wherein ∑i∈Eπi=1, according to entering the state in unit time when steady state and leave the average time phase of the state
Etc. formula can be listed:
In the system that obtains after the probability distribution under each state, so that it may calculate some fingers for measuring system performance
Mark:
1. the expectation TL of system waiting list length can be acquired according to the probability distribution of system in this state:
TL=∑I ∈ E, i > N(i-N)πi
2. client is not obtaining servicing average latency E [W] expectation that also do not abandon when waiting:
3. the utilization rate expectation U that attends a banquet of customer service:
The core of common phone predictive dialer system is predictive outbound algorithm, and this system opens a kind of prediction type
Outgoing call algorithm is divided into two important modules: predictive outbound algoritic module and dynamic statistics module.So as to from two sides
Always optimize, both direction there are many prioritization schemes, the bottleneck problem that will appear so as to avoid only optimization algorithm.And
The specific statistics of parameter need not be concerned about when optimization algorithm, without considering the realization of algorithm and whether having to algorithm when statistical parameter
With.
It is typically greater than idle number of seats at that time by the telephone number that predictive outbound algorithm is dialed, this is because phone
Percent of call completed is usually more much smaller than 1, so can greatly improve seat utilization compared to the dialing for dialling and being not previously predicted by hand function
Rate will result in that phone is got through but service for it of not attending a banquet yields, such case referred to as " is exhaled but if dialling more
Damage ", must inhibit the ratio of call loss, the i.e. percent of call lost in practice, otherwise can cause greatly to harass to user, so algorithm will
It is balanced between utilization rate and the phone percent of call lost attending a banquet.As long as still can be reached nevertheless, algorithm is reasonably predicted
To very high utilization rate of attending a banquet, good prediction effect is generated.But this requires algorithms can establish preferable model, to each seed ginseng
Several Changeements goes out a set of theoretical (including queueing theory), and then adapts to the situation of various complexity.Monochromatic light is preferable pre- by establishing
Survey model not enough, because prediction must be predicted according to historical data, need which and how many historical data, and how logarithm
Relatively reasonable value etc. can be obtained according to analysis is carried out, and here it is the effect of dynamic statistics module, the present invention pertains only to predict
Formula outgoing call algorithm.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
Centainly refer to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be any
One or more embodiment or examples in can be combined in any suitable manner.
Although an embodiment of the present invention has been shown and described, it will be understood by those skilled in the art that: not
A variety of change, modification, replacement and modification can be carried out to these embodiments in the case where being detached from the principle of the present invention and objective, this
The range of invention is defined by the claims and their equivalents.
Claims (5)
1. a kind of predictive outbound algorithm based on dynamic statistics process, which comprises the steps of:
S1 starts, and setting target is attended a banquet utilization rate AO, maximum percent of call lost ARmax, number of switches T, maximum waiting time TQmax,
Time in predicting interval △ t, waiting list maximum length qL, user's patience 1/ η of time, user are reluctant that the probability of waiting list is added
p0, and the initial value of quantity λ is made a phone call outward, enabling the accumulated value al=0, l of l is system prediction maximum outgoing call value；
S2 obtains current customer service sum N by interchanger in t moment, and idle customer service sum f is calling telephone number j, connects
Rate ρ, average 1/ β of jingle bell duration, the numerical value of 1/ μ of average call duration and waiting list length q, the t are positive number；
S3 judges whether to meet q < qL, goes to step S4 if meeting, otherwise suspend mode to t+ time Δt go to step S2 carry out it is next
Wheel prediction, and current time is updated for t=t+ Δ t；
S4, minimum needed for calculating are attended a banquet several M, calculation formula are as follows:
S5 judges whether to meet N≤M, goes to step S17 if meeting, otherwise go to step S6；
S6, computing system predict maximum outgoing call value l, cumulative calculation al=al+l, wherein
S7, the percent of call lost AR at dynamic statistics last moment simultaneously judge whether to meet AR < ARmax, S8 is gone to step if meeting, otherwise
Go to step S16；
S8 judges whether to meet current idle customer service sum f > al, goes to step S9 if meeting, otherwise go to step S13；
S9, prediction result s=(f-j)/ρ-q+al, al sets 0 if al > 1；
S10 calculates period of reservation of number Tq, Tq=q/ (N* μ)；
S11 judges whether to meet Tq < TQmax, if meet if go to step S12, otherwise suspend mode to t+ time Δt go to step S2 into
The prediction of row next round, and current time is updated for t=t+ Δ t；
S12 updates prediction result s=s*ARrate, and ARrate is that the exhalation of system weakens ratio, and ARrate=1-AR, AR are to exhale
Loss rate, AR=p0B(θ,N)+(1-p0) λ β ρ η, go to step S18；
S13, prediction result s=f-j-q+al, al sets 0 if al > 1；
S14 calculates percent of call lost AR, AR=p0B (θ, N)+(1-p0)λβρη
B (θ, N) be N number of customer service without it is idle when probability πN；
S15 judges whether to meet AR < ARmax, if meet if go to step S10, otherwise suspend mode to t+ time Δt go to step S2 into
The prediction of row next round, and current time is updated for t=t+ Δ t；
S16, prediction result s=f-j-q, goes to step S18；
S17, prediction result s=f-j, goes to step S18；
S18 is calculated and is finally made a phone call quantity λ, λ=(int) min (max (s, 0), T)；
S19 dials λ phone outward, and algorithm suspend mode to t+ time Δt goes to step S2 and carries out next round prediction, and updates current
Moment is t=t+ Δ t；Above step is repeated, is run until application program terminates.
2. the predictive outbound algorithm according to claim 1 based on dynamic statistics process, which is characterized in that the step
The percent of call lost AR at dynamic statistics last moment is the phone t1 moment to the current time stage call loss from before current time in S7
All phone numbers that number is connected divided by this stage, the t1 are positive number.
3. the predictive outbound algorithm according to claim 2 based on dynamic statistics process, which is characterized in that the t1 is
Time between 15 minutes to 45 minutes.
4. according to the interchanger side of dialing based on the predictive outbound algorithm described in claim 1 based on dynamic statistics process
Method, which comprises the steps of:
Step 1, predictive outbound algorithm according to claim 1, which predicts subsequent time, will make a phone call quantity λ；
Step 2, interchanger dials λ phone according to this prediction result；
Step 3, after average 1/ β of jingle bell duration, the β is positive number, has the closing of the circuit of ρ ratio, is left 1- ρ ratio
Phone access failure is filtered, and the ρ is the positive number greater than 0 less than 1；
Step 4, the number metering-in control system of connection, control system can be according to currently whether available free attend a banquet be allocated；
Step 5, distribution is attended a banquet and is serviced for it if available free attend a banquet, service finish be dial it is successful；If being sat without the free time
Seat, have p0 probability client be ready be added waiting list wait, the client of remaining 1-p0 probability be unwilling be added etc.
It abandons forming call loss to queue；
Step 6, the client of queuing also some abandon forming call loss due to waiting more than its patience 1/ η of time, it is only remaining
Under meeting idle being attended a banquet by control system according to what is serviced at present not above its patience 1/ η of time
The service of attending a banquet is distributed for it, and finally service is completed.
5. a kind of device for running the predictive outbound algorithm described in claim 1 based on dynamic statistics process, feature exist
In, including data input module, predictive outbound algoritic module, interchanger, dynamic statistics module, data memory module and data
Display module is exported, the data input module is to predictive outbound algoritic module input initialization supplemental characteristic, the prediction
Formula outgoing call algoritic module is connected with interchanger, carries out information exchange, and the dynamic statistics module is connected with data memory module, connects
The data of data memory module are received, the statistical data of the dynamic statistics module statistics different frequency is simultaneously transferred to predictive outbound
Algoritic module, while being displayed data by data output interface；
The predictive outbound algoritic module includes configuration parameter unit, initiation parameter unit, theoretical calculation data cell, week
Phase statistical parameter unit, predictive outbound algorithm unit, prediction result unit；The configuration parameter unit is used for desampler
Configuration parameter information, the initiation parameter unit various object parameter datas that input module inputs for receiving data, institute
It states theoretical calculation data cell to be used to receive theoretical calculation but want the data of cyclomorphosis, the period statistical parameter unit is used for
The various data that dynamic statistics module is come out with different statistic frequencies are received, the predictive outbound algorithm unit is for real
Existing predictive outbound algorithm described in claim 1, the prediction result unit are used to transmit prediction result to interchanger；
The dynamic statistics module include obtain historical data unit, selection statistics policy unit, statistics policy unit, to prediction
Formula outgoing call algoritic module transmits the unit of statistical data and the unit to output interface transmitting statistical data, the acquisition history number
According to be stored in unit including obtain written data library data and interchanger save service but also have not enough time to write
Enter the data in database, for the selection statistics policy unit for selecting using which kind of statistics strategy, the statistics strategy is single
Member includes a variety of specific statistics strategies.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610118087.1A CN105704335B (en) | 2016-03-02 | 2016-03-02 | Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610118087.1A CN105704335B (en) | 2016-03-02 | 2016-03-02 | Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105704335A CN105704335A (en) | 2016-06-22 |
| CN105704335B true CN105704335B (en) | 2018-12-18 |
Family
ID=56222835
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610118087.1A Active CN105704335B (en) | 2016-03-02 | 2016-03-02 | Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105704335B (en) |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106657693B (en) * | 2017-01-13 | 2019-11-08 | 深圳市彬讯科技有限公司 | A kind of calling system predictive outbound method and system |
| US9930180B1 (en) * | 2017-04-28 | 2018-03-27 | Afiniti, Ltd. | Techniques for behavioral pairing in a contact center system |
| CN108156331B (en) * | 2017-11-10 | 2021-04-30 | 大连金慧融智科技股份有限公司 | Intelligent automatic dialing outbound method |
| CN110581928B (en) * | 2018-06-07 | 2021-02-12 | 马上消费金融股份有限公司 | Telephone calling-out method, system, electronic equipment and storage medium |
| CN110581925B (en) * | 2018-06-07 | 2020-12-25 | 马上消费金融股份有限公司 | Method and system for adjusting parameters of predictive outbound algorithm, electronic equipment and storage medium |
| CN109246321A (en) * | 2018-07-27 | 2019-01-18 | 平安科技（深圳）有限公司 | Call management-control method, system, computer equipment and its storage medium |
| CN109040477B (en) * | 2018-07-30 | 2021-11-02 | 厦门快商通信息技术有限公司 | Optimization method for predicting outbound algorithm and outbound scheduling system |
| CN110858861A (en) * | 2018-08-24 | 2020-03-03 | 北京京东尚科信息技术有限公司 | Outbound system processing method and device |
| CN111147677B (en) * | 2018-11-05 | 2021-12-21 | 马上消费金融股份有限公司 | Call center system |
| CN110166641B (en) * | 2019-05-21 | 2020-11-27 | 深圳市中讯网联科技有限公司 | Dynamic predictive intelligent automatic outbound system algorithm |
| CN110266901B (en) * | 2019-07-08 | 2021-05-07 | 北京容联汇通信息技术有限公司 | Method for predicting outbound call by call center |
| CN111107229B (en) * | 2019-12-05 | 2021-05-04 | 北京高阳捷迅信息技术有限公司 | Outbound prediction method and device for intelligent customer service |
| CN111246029B (en) * | 2020-01-19 | 2021-08-27 | 中电智恒信息科技服务有限公司 | Outbound algorithm for dynamically adjusting calling frequency |
| CN111866290B (en) * | 2020-04-13 | 2021-07-27 | 马上消费金融股份有限公司 | Method for determining number of outgoing calls, automatic outgoing call method and equipment |
| CN111935362B (en) * | 2020-08-07 | 2021-07-16 | 北京新唐思创教育科技有限公司 | Call data processing method and device, electronic equipment and computer storage medium |
| CN113315876B (en) * | 2021-05-27 | 2023-01-20 | 中国银行股份有限公司 | Telephone bank service control method, device, server and storage medium |
| CN113179353B (en) * | 2021-05-28 | 2022-10-11 | 拉卡拉支付股份有限公司 | Call control method, device, electronic equipment, storage medium and program product |
| CN113225436B (en) * | 2021-05-28 | 2022-07-15 | 拉卡拉支付股份有限公司 | Call control method, device, electronic equipment, storage medium and program product |
| CN113382117B (en) * | 2021-06-04 | 2022-08-12 | 上海浦东发展银行股份有限公司 | Predictive outbound method based on machine learning |
| CN113452849A (en) * | 2021-06-28 | 2021-09-28 | 上海翰声信息技术有限公司 | Method and device for automatically predicting call volume of call center |
| CN113438376B (en) * | 2021-08-26 | 2022-01-07 | 阿里巴巴达摩院(杭州)科技有限公司 | Telephone calling-out method and device |
| CN113746990B (en) * | 2021-09-03 | 2023-03-21 | 科讯嘉联信息技术有限公司 | Outbound prediction system and method for changing human interaction after AI human-computer interaction |
| CN113923308B (en) * | 2021-10-15 | 2024-04-09 | 浙江工业大学 | Predictive outbound task allocation method and outbound system based on deep reinforcement learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103095937A (en) * | 2012-12-14 | 2013-05-08 | 广东电网公司佛山供电局 | Prediction method for number of seats in call center based on telephone traffic prediction |
| WO2013148452A1 (en) * | 2012-03-26 | 2013-10-03 | The Resource Group International, Ltd. | Call mapping systems and methods using bayesian mean regression (bmr) |
| CN104539814A (en) * | 2014-11-26 | 2015-04-22 | 小米科技有限责任公司 | Customer service call forwarding method and device |
| CN104778532A (en) * | 2015-03-12 | 2015-07-15 | 科大国创软件股份有限公司 | Intelligent calling center management system and intelligent scheduling control method |
-
2016
- 2016-03-02 CN CN201610118087.1A patent/CN105704335B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2013148452A1 (en) * | 2012-03-26 | 2013-10-03 | The Resource Group International, Ltd. | Call mapping systems and methods using bayesian mean regression (bmr) |
| CN103095937A (en) * | 2012-12-14 | 2013-05-08 | 广东电网公司佛山供电局 | Prediction method for number of seats in call center based on telephone traffic prediction |
| CN104539814A (en) * | 2014-11-26 | 2015-04-22 | 小米科技有限责任公司 | Customer service call forwarding method and device |
| CN104778532A (en) * | 2015-03-12 | 2015-07-15 | 科大国创软件股份有限公司 | Intelligent calling center management system and intelligent scheduling control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105704335A (en) | 2016-06-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105704335B (en) | Predictive outbound algorithm, interchanger dialing method and device based on dynamic statistics process | |
| US5436965A (en) | Method and system for optimization of telephone contact campaigns | |
| US7715546B2 (en) | System and method for updating contact records | |
| US5452350A (en) | Subscriber call routing processing system | |
| US5787163A (en) | Intelligent load balancing of special service calls based on availability of terminations | |
| CN106657693B (en) | A kind of calling system predictive outbound method and system | |
| US8386639B1 (en) | System and method for optimized and distributed resource management | |
| JP3759359B2 (en) | Adjusting call selection to achieve performance metrics targets based on call center segments | |
| US9232066B2 (en) | Method and apparatus for queue routing | |
| CA2441545C (en) | Self-tuning statistical resource allocation for multipoint network events | |
| CN102415068A (en) | Method and system for a multitenancy telephone network | |
| CN104125349A (en) | Voice interaction management method and system based on telephone traffic forecasting | |
| CN104883459A (en) | Call Routing Methods And Systems Based On Multiple Variable Standardized Scoring And Shadow Queue | |
| CN109343801A (en) | Date storage method, equipment and computer readable storage medium | |
| US9774740B2 (en) | Techniques for benchmarking pairing strategies in a contact center system | |
| CN109040477A (en) | A kind of optimization method and outgoing call scheduling system of prediction outgoing call algorithm | |
| CN110166641A (en) | Dynamic prediction type intelligence automatic outer call system algorithm | |
| EP3347863A1 (en) | Techniques for benchmarking pairing strategies in a contact center system | |
| WO2014014935A1 (en) | System and method for optimized and distributed resource management | |
| JPH02230352A (en) | Service control and communication network operating system therefor | |
| US11165908B2 (en) | Techniques for benchmarking pairing strategies in a contact center system | |
| CN103685788A (en) | Automatic outbound call and voice interaction system based on IP network | |
| CN100463468C (en) | System and method for measuring traffic volume distributed among areas | |
| CN113923308B (en) | Predictive outbound task allocation method and outbound system based on deep reinforcement learning | |
| US8023431B2 (en) | Digital subscriber line (DSL) tracking system |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20190426 Address after: Room 102, E District, 1st Floor, Building No. 201, Wumaobang Road, Qingpu District, Shanghai, 20179 Patentee after: Shanghai Hansheng Information Technology Co., Ltd. Address before: 400045 Sha Zheng street, Shapingba District, Chongqing City, No. 174 Patentee before: Chongqing University |