CN112508295A - Incoming call service prediction method and system based on service map and Markov model - Google Patents
Incoming call service prediction method and system based on service map and Markov model Download PDFInfo
- Publication number
- CN112508295A CN112508295A CN202011490382.2A CN202011490382A CN112508295A CN 112508295 A CN112508295 A CN 112508295A CN 202011490382 A CN202011490382 A CN 202011490382A CN 112508295 A CN112508295 A CN 112508295A
- Authority
- CN
- China
- Prior art keywords
- service
- incoming call
- probability
- matrix
- client
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 239000011159 matrix material Substances 0.000 claims abstract description 105
- 230000007704 transition Effects 0.000 claims abstract description 77
- 238000006243 chemical reaction Methods 0.000 claims abstract description 48
- 239000013598 vector Substances 0.000 claims description 13
- 230000008859 change Effects 0.000 claims description 11
- 230000009466 transformation Effects 0.000 claims description 4
- 238000009826 distribution Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 16
- 238000010586 diagram Methods 0.000 description 11
- 238000004590 computer program Methods 0.000 description 7
- 238000011160 research Methods 0.000 description 6
- 238000012546 transfer Methods 0.000 description 6
- 238000011161 development Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000005457 optimization Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000005653 Brownian motion process Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005537 brownian motion Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 230000003137 locomotive effect Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 238000005309 stochastic process Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/60—Business processes related to postal services
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Economics (AREA)
- Computational Mathematics (AREA)
- Marketing (AREA)
- Algebra (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Game Theory and Decision Science (AREA)
- Artificial Intelligence (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Telephonic Communication Services (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
Abstract
The invention relates to an incoming call service prediction method and system based on a service map and a Markov model, wherein the method comprises the following steps: establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type; setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service conversion matrix; and predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix. According to the method, the relation of the incoming call of the client aiming at the service type is analyzed, and the state transition probability matrix is constructed on the basis of the service map and the Markov model and is used for predicting the service type aiming at the incoming call of the client next time, so that the customer service staff are assisted to intervene in advance, and the service efficiency is improved. Through the state transition probability matrix, the traffic of the future day is predicted, the customer center manager is supported to accurately develop a personnel scheduling plan, and the operation efficiency is improved.
Description
Technical Field
The invention relates to the technical field of business prediction, in particular to an incoming call business prediction method and system based on a business map and a Markov model.
Background
In the prior art, the service industry rises rapidly in China, wherein a customer service center is more and more concerned by enterprises as a novel business competition tool in the 21 st century. With the continuous change of the customer requirements in the large market, the customer requirements are more and more strict, the balance between enterprise management and the customer requirements is more and more important, a large number of enterprises gradually establish a customer service department or establish professional calls to help the enterprises manage and reserve new and old customers, and the competitiveness of enterprise cores is improved. Particularly, with the implementation of an enterprise strategy of 'three-type two-network and top-class world' of electric power, the client needs to grasp the client requirements and expectations, grasp the client preferences and behavior rules, and finally achieve the improvement and improvement of the client experience through personalized differentiated services from the comprehensive insight of the client.
At present, electric power enterprises conduct diversified research aiming at customer service (95598) business, but mainly surround the application aspect of customer labels, and lack research on the self management aspect of customer service business. However, the electric power 95598 has huge traffic, the cost of human resources accounts for a large share, and how to improve the operation efficiency becomes a problem to be urgently solved by managers.
On the other hand, efficient management of the customer service center requires that resources of the telephone customer service center are called according to workload, and accurate customer service incoming call prediction is important. At present, the incoming call amount prediction of a customer service center is mainly carried out by management workers according to experience and simple model prediction, the problems of large workload and randomness from person to person exist, the prediction accuracy is often insufficient, and a plurality of problems are brought to the following detailed management of telephone operators.
In the existing service prediction methods of many telephone call centers, simple regression prediction is mainly carried out by taking historical service volume as a reference, the randomness of the telephone of a client is not considered, and the prediction accuracy is reduced. In order to improve the customer service capability, customer service personnel need to predict the type of consultation service of the next call of the same customer in advance so as to intervene in advance and improve the service efficiency and the customer satisfaction, but at present, electric power enterprises do not make intensive research on the type of consultation service.
Disclosure of Invention
The invention provides an incoming call service prediction method and system based on a service map and a Markov model, which solve the problem of service prediction of customer incoming calls in customer service in the power industry in the prior art.
According to one aspect of the invention, an incoming call service prediction method based on a service graph and a Markov model is provided, and comprises the following steps:
establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type;
setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service conversion matrix;
and predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix.
The setting of the Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type comprises the following steps:
and acquiring the occurrence probability of the incoming call of the client aiming at the service type according to the historical data of the incoming call of the client.
The probability of occurrence of the client incoming call for the service type comprises the following steps:
setting the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, wherein each service jointly forms a state probability vector in the set time length; the set period of time is preferably one day.
The setting of the Markov state transition probability matrix includes:
according to the service conversion matrix, the state transition probability matrix is obtained as follows:
wherein, P (n) is a probability value when the time quantum n is used, and describes the probability distribution of m states of mutual transition; bmmThe number of times of occurrence of the condition that the state m is transferred to the state m in one step; pmmProbability value for service m to transition to service m.
The method comprises the following steps of establishing a service conversion matrix according to the conversion condition of the incoming call of a customer for the service type, wherein the service conversion matrix comprises the following steps:
forming a service map according to the change relation of the client incoming call aiming at the service type;
and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
The method further comprises the following steps:
acquiring the total traffic of the incoming call of the client within a set time length;
and predicting the predicted traffic in the next set time length according to the Markov state transition probability matrix and the total traffic.
According to another aspect of the present invention, there is provided an incoming call traffic prediction system based on a traffic pattern and a markov model, the system comprising:
the service conversion matrix unit is used for establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type;
a state probability transition matrix unit, which is used for setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service transition matrix;
and the service prediction unit is used for predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix.
The state probability transition matrix unit is specifically configured to:
acquiring the occurrence probability of the incoming call of the client aiming at the service type according to the historical data of the incoming call of the client; setting the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, wherein each service jointly forms a state probability vector in the set time length; the set period of time is preferably one day.
The service conversion matrix unit is specifically configured to:
forming a service map according to the change relation of the client incoming call aiming at the service type; and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
The system further comprises:
the service volume prediction unit is used for acquiring the total service volume of the incoming call of the client in a set time length; predicting the predicted traffic in the next set duration according to the Markov state transition probability matrix set by the Markov state transition probability matrix unit and the total traffic; the set period of time is preferably one day.
The beneficial effect who adopts above-mentioned scheme is:
in the scheme of the invention, the state transition probability matrix is constructed by analyzing the relation of the incoming call of the client to the service type and based on the service map and the Markov model, and is used for predicting the service type of the incoming call of the client next time, assisting customer service personnel to intervene in advance and improving the service efficiency. Through the state transition probability matrix, the traffic of one day in the future or the traffic of one week in the future is predicted, a customer center manager is supported to accurately develop a personnel scheduling plan, and the operation efficiency is improved.
Drawings
Fig. 1 is a schematic flow chart of an incoming call service prediction method based on a service graph and a markov model according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of a service transformation matrix according to an embodiment of the present invention.
Fig. 3 is a service map of the current day of the incoming call of the client according to the embodiment of the present invention.
Fig. 4 is a traffic map of traffic for a future day provided by an embodiment of the present invention.
Fig. 5 is a traffic map of a future week of traffic provided by an embodiment of the present invention.
Fig. 6 is a schematic structural diagram of an incoming call service prediction system based on a service graph and a markov model according to an embodiment of the present invention.
Detailed Description
The principles and features of this invention are described below in conjunction with the following drawings, which are set forth by way of illustration only and are not intended to limit the scope of the invention. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
The current telephone customer service center traffic prediction method mainly comprises a time series prediction method and a mathematical statistical analysis method. The regression analysis is a mathematical analysis method with wide application, and is mainly used for determining the interdependent quantitative relation between two or more variables.
The linear regression prediction model has important significance for the prediction of the economic and social systems, but because the influence factors of the customer service volume are complex and the change of the customer service volume presents certain random fluctuation characteristics, the prediction result can generate errors.
The Markov chain describes the future development state of a random dynamic system according to a transition matrix between system states, and the transition matrix reflects the regularity between the states. The main feature of the markov chain process is the inefficiency, the states after time are only related to the states at time t, and not to the previous states. Therefore, in the prediction of customer service business, the Markov chain model can be adopted to improve the prediction accuracy.
Markov chain is a general model of law that can mathematically explain natural changes, proposed by the well-known russian mathematician markov in about 1910. The Markov process is already an important aspect of the random process theory in probability theory. After a hundred years of development, the markov process has penetrated into various fields and played an important role, for example, in the fields of economy and communication, and in addition, in the fields of natural sciences such as geological disasters, medical health services, biology and the like.
In the research on practical problems, it is found that a plurality of phenomena can be generated along with the continuous development and change of time. Still other phenomena or processes may be expressed as follows: in the case where "now" is known, the "future" of such a course of change is unrelated to the "past". That is to say that future occurrences of such a process do not depend on past developmental changes. A process with the above properties is called markov process. The markov process can describe many phenomena in real life. Such as brownian motion made by particles in a liquid, daily sales to be studied in commercial activities, voice signals in digital communications, video signals, etc. The Markov chain has a plurality of applications in other fields, such as the management of bad assets of banks, locomotive management, enterprise management, ecological environment evolution, urban water consumption simulation, information processing and other scientific research and production life.
The method considers that most customer service forecast of the customer service center aims at the forecast of the service volume and lacks of forecast of the service type at present, so that the pertinence accuracy of the service is reduced. The invention uses Markov chain to construct business type state transition probability matrix, to master the business type probability of next incoming call, and use the matrix to predict the business volume in future day.
In addition, at present, many customer service centers do not carry out research on service type prediction of the next incoming call of the same customer, and the method adopts Markov to carry out prediction innovatively based on a service map.
The method is based on a client service 95598 work order acceptance table of the power enterprise, the occurrence frequency of each business subtype is counted according to days, then the transition probability among all the business subtypes is obtained by applying a transition probability matrix calculation method based on the optimal thought in a Markov chain optimization prediction method, the transition probability is used for representing 95598 business relevance and a business map and is used for predicting the business type probability of the next call of a client, and meanwhile, the traffic of the next day can be predicted by utilizing the transition matrix, so that the client service is assisted in intervening in advance and personnel scheduling is carried out. The main contents comprise: firstly, constructing a transition probability matrix; secondly, predicting the service type of the next call of the client; thirdly, predicting the traffic of the future day; fourth, traffic is predicted for a future week or other period.
The invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, a schematic flow chart of an incoming call service prediction method based on a service graph and a markov model provided in embodiment 1 of the present invention is specifically as follows:
and step 11, establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type.
In the invention, a service map is formed according to the change relation of the client incoming call aiming at the service type; and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
And establishing a service conversion matrix according to the actual conversion situation of the incoming call of the client aiming at the service.
Combing the relation between different customer incoming calls and aiming at the service types to form a service map, which comprises the following steps:
customer 1: service 1-service 2- … … -service m.
And (3) a client 2: service 1-service 3- … … -service m.
And 3, a client: service 2-service 4- … … -service m.
、、、、、、、、、、。
And constructing a service conversion matrix according to the service map of the incoming call of the client aiming at the service type. See in particular fig. 2. Wherein, b11Representing the number of occurrences of a one-step transition from state 1 to state 1, and bmmRepresenting the number of occurrences of a one-step transition from state m to state m. The relationship between the various service types can be intuitively reflected according to the service conversion matrix.
And step 12, setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service conversion matrix.
In the embodiment of the invention, the occurrence probability of the incoming call of the client aiming at the service type is obtained according to the historical data of the incoming call of the client.
The method comprises the steps that the occurrence probability of a client incoming call aiming at a service type comprises the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, and all the services jointly form a state probability vector in the set time length; the set period of time is preferably one day.
Specifically, it is first required to determine the status of the customer service system, that is, the status of the customer call for the specific system of the business, which may be the call record of the power system customer service system such as 95598 and the record of the customer for the business type.
Generally, the occurrence frequency of each sub-service in a certain day is used to represent the state probability of the sub-service in the day, and the services together form the state probability vector of the day. Of course, the time duration may not be one day, but may be a self-set time duration, and may be half a day, two days, one week, or any other set time duration.
The Markov Model is a statistical Model, and is widely applied in the application fields of speech recognition, automatic part-of-speech tagging, phonetic-to-character conversion, probabilistic grammar and other natural language processing. Assume a random process with one value per time period, let XnRepresenting its value over a time period n, we want to apply to a series of successive values X0，X1…, a probabilistic model is built, assuming that the value at the n +1 state depends on the value of the nth state, which defines a Markov chain: let { XnN-0, 1, 2, … is a random process of finite or infinite possible values, and unless specifically noted, this set of possible values of the random process will be referred to as a set of non-negative integers {0, 1, 2, … }. if X is not positivenI, the process is said to be in state i at time t. We assume that as long as the process is in state i, there is a fixed probability P of having it in the next stateState j, i.e. we assume for all states i0，i1，…in-1If ij and all n are greater than or equal to 0, have
P{Xn+1＝j|Xn＝i，Xn-1＝in-1，…，X1＝i1，X0＝i0}＝Pij
Such a stochastic process is a markov chain. For a Markov chain, X is the state in the past given0，X1，…Xn-1And the present state XnTime, future state Xn+1Is independent of past conditions and depends only on present conditions.
The Markov chain describes the future development state of a random dynamic system according to a transition matrix between system states, and the transition matrix reflects the regularity between the states. The main feature of the markov chain process is the inefficiency, i.e., the states after time are only related to the states at time t, and not to the previous states. Therefore, in the prediction of customer service business, the Markov chain model can be adopted to improve the prediction accuracy.
In order to obtain a more accurate state transition probability matrix, an optimization model can be established by using the markov optimization idea, namely, the criterion that the sum of the squares of the errors of the probability vectors of the actual states and the state probability vectors calculated theoretically is to be minimized in m moments.
A transition probability matrix P (n) is constructed.
According to the traffic transition matrix, the state transition probability matrix is obtained as follows:
p (n) describes the probability distribution of m state transitions to each other, which can be represented by the following equation.
PijProbability of transition to service j for service i, e.g. Pl2Indicating the probability of transition from service l to service 2.
And step 13, predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix.
The conversion relation between the incoming call of the client and the service is recorded in the service conversion matrix, the state transition probability matrix is a probability matrix of the incoming call of the client and the service type, which is established according to a Markov model, the data of the incoming call of the client and the service type are fused, the service data in the service conversion matrix is input into the probability matrix of the service type, and the probability of the incoming call of the next client aiming at the service type can be obtained, namely the next incoming call of the client aiming at the service type can be predicted.
Noting that the service type of the current latest incoming call of the client is i, namely, the initial state of the client is i, according to the transition probability matrix obtained by the service map, the probability vector for the state i to be transferred to other states in one step is Pi(n)＝(pi1，pi2，…，pin) Record PiMaximum value in (n) is pijThen, it means that the maximum probability of the state i is transferred to the state j, that is, the service type j is the service type for the next incoming call.
The present embodiment may also utilize a state transition probability matrix for the traffic type to predict the change in total traffic. Acquiring the total traffic of the incoming call of the client within a set time length; and predicting the predicted traffic in the next set time length according to the Markov state transition probability matrix and the total traffic.
Taking the total traffic of one day as an example, the total amount processed by each service type of the day is input into the state transition probability matrix so as to predict the total amount of each service type of the next day.
Inputting: record the occurrence status of each business on the day as S (0)1×mI.e. vector S (0)1×m＝(s1，s2，…，sm) Indicates the frequency of each service occurrence on the day, wherein siRepresenting the total amount of traffic of a certain service i on the day. In particular toAs shown in fig. 3.
Recording the traffic state of one day in the future as S (1)1×m＝(s1，s2，…，sm) According to the transition probability matrix obtained by the service map, the predicted result of the service volume obtained in the future day can be obtained as follows:
S(0)1×m·P(n)m×m。
and (3) outputting: s (1)1×m＝(s1，s2，…，sm). As shown in particular in fig. 4.
Take the example of predicting traffic volume in the next week. The client service map is obtained by describing the service transfer track of each client, the service transfer track of a single client reflects the transfer relation between the most real services, and the service type of incoming call consultation of the client center has stability, so that the one-step service transfer probability matrix obtained by calculation based on the service map also has stability; in addition, the service matrix table is obtained by comprehensively considering the service transfer trajectories of all the clients in the client center, so that the obtained one-step service transfer probability matrix has the characteristic of ergodicity.
To sum up, the one-step traffic transition probability matrix has stability and ergodicity, so the state vector after K-step transition can be recorded as: s (k)1×m＝S(0)1×m·Pk(n)m×m。
Inputting: record the occurrence status of each business on the day as S (0)1×mI.e. vector S (0)1×m＝(s1，s2，…，sm) Indicates the frequency of each service occurrence on the day, wherein siRepresenting the total amount of traffic of a certain service i on the day. As shown in fig. 3.
Recording the traffic state of i days in the future as S (i)1×m＝(s1，s2，…，sm) According to the transition probability matrix obtained by the service map, the service volume prediction result obtained in the future i days can be obtained as follows:
S(0)1×m·Pi(n)m×m
in summary, the traffic volume in the future one week is:
S(7)1×m＝S(0)1×m·P(n)m×m+S(0)1×m·P2(n)m×m+…+S(0)1×m·P7(n)m×m
and (3) outputting: s (7)1×m＝(s1，s2，…，sm). As shown in fig. 5.
In the embodiment, the relation of the incoming call of the client to the service type is analyzed, and the state transition probability matrix is constructed based on the service map and the Markov model and is used for predicting the service type of the incoming call of the client next time, so that the customer service staff is assisted to intervene in advance, and the service efficiency is improved. Through the state transition probability matrix, the traffic of the future day is predicted, the customer center manager is supported to accurately develop a personnel scheduling plan, and the operation efficiency is improved.
As shown in fig. 6, a schematic structural diagram of an incoming call service prediction system based on a service graph and a markov model provided by the present invention includes:
a service conversion matrix unit 21, configured to establish a service conversion matrix according to a conversion condition of a client incoming call for a service type;
a state probability transition matrix unit 22, configured to set a markov state transition probability matrix according to the occurrence probability of the customer incoming call for the service type and the service transition matrix;
and the service prediction unit 23 is configured to predict a service type of the next incoming call of the client for the service according to the state transition probability matrix.
The state probability transition matrix unit 22 is specifically configured to:
acquiring the occurrence probability of the incoming call of the client aiming at the service type according to the historical data of the incoming call of the client; setting the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, wherein each service jointly forms a state probability vector in the set time length; the set period of time is preferably one day.
The service conversion matrix unit 21 is specifically configured to:
forming a service map according to the change relation of the client incoming call aiming at the service type; and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
The system further comprises:
a traffic prediction unit 24, configured to obtain a total traffic of the incoming call of the client within a set time length; predicting the predicted traffic in the next set duration according to the Markov state transition probability matrix set by the state transition probability matrix unit 22 and the total traffic; the set period of time is preferably one day.
In summary, in the scheme of the present invention, a state transition probability matrix is constructed by analyzing the relationship of the client incoming call to the service type and based on the service map and the markov model, and is used for predicting the service type targeted by the next client incoming call, so as to assist the customer service staff to intervene in advance and improve the service efficiency. Through the state transition probability matrix, the traffic of one day in the future or the traffic of one week in the future is predicted, a customer center manager is supported to accurately develop a personnel scheduling plan, and the operation efficiency is improved.
The present invention has been described in detail with reference to specific embodiments, but the above embodiments are merely illustrative, and the present invention is not limited to the above embodiments.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (10)
1. An incoming call service prediction method based on a service graph and a Markov model, the method comprising:
establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type;
setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service conversion matrix;
and predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix.
2. The method of claim 1, wherein setting a markov state transition probability matrix based on a probability of occurrence of a customer incoming call for a traffic type comprises:
and acquiring the occurrence probability of the incoming call of the client aiming at the service type according to the historical data of the incoming call of the client.
3. The method of claim 2, wherein the probability of occurrence of the client call for a type of service comprises:
setting the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, wherein each service jointly forms a state probability vector in the set time length; the set period of time is preferably one day.
4. The method of claim 1, wherein setting the markov state transition probability matrix comprises:
according to the service conversion matrix, the state transition probability matrix is obtained as follows:
wherein, P (n) is a probability value when the time quantum n is used, and describes the probability distribution of m states of mutual transition; bmmThe number of times of occurrence of the condition that the state m is transferred to the state m in one step; pmmProbability value for service m to transition to service m.
5. The method of claim 1, wherein the establishing a traffic transformation matrix based on the transformation of the customer's incoming call for the traffic type comprises:
forming a service map according to the change relation of the client incoming call aiming at the service type;
and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
6. The method of any of claims 1 to 5, further comprising:
acquiring the total traffic of the incoming call of the client within a set time length;
and predicting the predicted traffic in the next set time length according to the Markov state transition probability matrix and the total traffic.
7. An incoming call traffic prediction system based on traffic maps and markov models, the system comprising:
the service conversion matrix unit is used for establishing a service conversion matrix according to the conversion condition of the incoming call of the client aiming at the service type;
a state probability transition matrix unit, which is used for setting a Markov state transition probability matrix according to the occurrence probability of the customer incoming call aiming at the service type and the service transition matrix;
and the service prediction unit is used for predicting the service type of the next incoming call of the client aiming at the service according to the state transition probability matrix.
8. The system of claim 7, wherein the state probability transition matrix unit is specifically configured to:
acquiring the occurrence probability of the incoming call of the client aiming at the service type according to the historical data of the incoming call of the client; setting the occurrence frequency of different services in a set time length as the state probability of the services in the current time length, wherein each service jointly forms a state probability vector in the set time length; the set period of time is preferably one day.
9. The system of claim 7, wherein the traffic transformation matrix unit is specifically configured to:
forming a service map according to the change relation of the client incoming call aiming at the service type; and according to the service map, establishing a service conversion matrix for the service type according to the conversion probability.
10. The system of any of claims 7 to 9, further comprising:
the service volume prediction unit is used for acquiring the total service volume of the incoming call of the client in a set time length; predicting the predicted traffic in the next set duration according to the Markov state transition probability matrix set by the Markov state transition probability matrix unit and the total traffic; the set period of time is preferably one day.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011490382.2A CN112508295A (en) | 2020-12-16 | 2020-12-16 | Incoming call service prediction method and system based on service map and Markov model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011490382.2A CN112508295A (en) | 2020-12-16 | 2020-12-16 | Incoming call service prediction method and system based on service map and Markov model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112508295A true CN112508295A (en) | 2021-03-16 |
Family
ID=74972935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011490382.2A Pending CN112508295A (en) | 2020-12-16 | 2020-12-16 | Incoming call service prediction method and system based on service map and Markov model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112508295A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117035176A (en) * | 2023-08-09 | 2023-11-10 | 上海智租物联科技有限公司 | Markov chain-based user power-change behavior prediction and directional recall method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106792517A (en) * | 2016-12-05 | 2017-05-31 | 武汉大学 | Base station service number time sequence forecasting method based on mobile phone location Time-spatial diversion probability |
| CN107506870A (en) * | 2017-09-06 | 2017-12-22 | 国家电网公司 | A kind of electric service hotspot prediction method based on hot word |
| CN107862413A (en) * | 2017-11-16 | 2018-03-30 | 四维口袋科技（北京）有限公司 | Based on markovian Forecasting Methodology and device |
| CN110110881A (en) * | 2019-03-21 | 2019-08-09 | 贵州电网有限责任公司 | Power customer requirement forecasting analysis method and system |
| CN110475032A (en) * | 2019-06-28 | 2019-11-19 | 平安科技（深圳）有限公司 | Multi-service interface switching method, device, computer installation and storage medium |
| CN111211830A (en) * | 2020-01-13 | 2020-05-29 | 电子科技大学 | Satellite uplink bandwidth resource allocation method based on Markov prediction |
| CN111444302A (en) * | 2020-04-17 | 2020-07-24 | 中国传媒大学 | Mobility prediction method, system and device based on user classification |
-
2020
- 2020-12-16 CN CN202011490382.2A patent/CN112508295A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106792517A (en) * | 2016-12-05 | 2017-05-31 | 武汉大学 | Base station service number time sequence forecasting method based on mobile phone location Time-spatial diversion probability |
| CN107506870A (en) * | 2017-09-06 | 2017-12-22 | 国家电网公司 | A kind of electric service hotspot prediction method based on hot word |
| CN107862413A (en) * | 2017-11-16 | 2018-03-30 | 四维口袋科技（北京）有限公司 | Based on markovian Forecasting Methodology and device |
| CN110110881A (en) * | 2019-03-21 | 2019-08-09 | 贵州电网有限责任公司 | Power customer requirement forecasting analysis method and system |
| CN110475032A (en) * | 2019-06-28 | 2019-11-19 | 平安科技（深圳）有限公司 | Multi-service interface switching method, device, computer installation and storage medium |
| CN111211830A (en) * | 2020-01-13 | 2020-05-29 | 电子科技大学 | Satellite uplink bandwidth resource allocation method based on Markov prediction |
| CN111444302A (en) * | 2020-04-17 | 2020-07-24 | 中国传媒大学 | Mobility prediction method, system and device based on user classification |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117035176A (en) * | 2023-08-09 | 2023-11-10 | 上海智租物联科技有限公司 | Markov chain-based user power-change behavior prediction and directional recall method |
| CN117035176B (en) * | 2023-08-09 | 2024-03-15 | 上海智租物联科技有限公司 | Markov chain-based user power-change behavior prediction and directional recall method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103999049B (en) | Method and apparatus for predicting virtual machine demand | |
| CN109191312A (en) | A kind of anti-fraud air control method and device of Claims Resolution | |
| JP2023537278A (en) | Methods and Systems for Scalable Contact Center Agent Scheduling Using Automated AI Modeling and Multi-Objective Optimization | |
| CN110533382A (en) | Processing method, device, server and the readable storage medium storing program for executing of meeting summary | |
| Rahardja et al. | Model Scheduling Optimization Workforce Management Marketing | |
| CN110445939B (en) | Capacity resource prediction method and device | |
| CN114118691A (en) | Scheduling scheme processing method, device, equipment and medium based on traffic | |
| CN110618867A (en) | Method and device for predicting resource usage amount | |
| CN111476460A (en) | Method, equipment and medium for intelligent operation scheduling of bank self-service equipment | |
| CN113657634A (en) | Bank branch queuing waiting time prediction method and device | |
| CN115423450A (en) | Processing method and device for scheduling data and computer equipment | |
| CN112508295A (en) | Incoming call service prediction method and system based on service map and Markov model | |
| CN113191533A (en) | Warehouse employment prediction method, device, equipment and storage medium | |
| CN109766541B (en) | Marketing strategy identification method, server and computer storage medium | |
| CN111353642A (en) | Telephone traffic scheduling prediction method, device, electronic equipment and storage medium | |
| Liu et al. | A multi-agent-based mould due date setting approach in stochastic production | |
| CN115330219A (en) | Resource scheduling method and device | |
| CN115760329A (en) | Bank outlet customer distribution method and system based on deep learning | |
| Wang et al. | A simulation study of workforce management for a two-stage multi-skill customer service center | |
| CN115700627A (en) | Scheduling method and device for personnel at bank outlets | |
| CN114581130A (en) | Bank website number assigning method and device based on customer portrait and storage medium | |
| CN103514503A (en) | IT base installation alteration assessment system | |
| CN113723663A (en) | Power work order data processing method and device, electronic equipment and storage medium | |
| US20230297909A1 (en) | System and method for predicting service metrics using historical data | |
| US20230325736A1 (en) | System and method for allocating multi-functional resources |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210316 |