• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The present invention relates to a contact center and method for servicing non-real time contact. The advancing time for the non-real time contact is determined by subtracting the service time for the preceding item from the service time for the subsequent item. For example, an item that is serviced immediately after some type of predetermined event, such as an event where a queue with non-real-time contact has no work agent, an event with an empty queue, an event with a changed system clock, and an event when the system is rebooted is a queued item. It is ignored when estimating the waiting time for. By using this method, accurate latency prediction is provided to manage the contact center more efficiently.
    公开号:KR20010098915A
    申请号:KR1020010022751
    申请日:2001-04-26
    公开日:2001-11-08
    发明作者:플록하트앤드류디；해리스마이클조셉
    申请人:빈 토마스 제이；아바야 테크놀러지 코퍼레이션；
    IPC主号:
    专利说明:

    WAIT TIME PREDICTION ARRANGEMENT FOR NON-REAL-TIME CUSTOMER CONTACTS}
    [13] The present invention relates generally to a method for estimating customer latency in customer contact systems, and more particularly to a method for estimating latency for non-real-time customer contact. will be.
    [14] In a customer contact system or use center, calls entered at the call center are answered and processed by multiple agents. The system automatically distributes and connects the input call to any agent that is suitable for handling this call and that is free, that is, not processing any other call at this moment. As used herein, "calling" refers to two entities including, for example, voice calls, VoIP, text-chat, email, fax, electronic documents, web forms, voice messages and video calls. Refers to any mode or type of contact between the two, but is not limited to voice calls and the like.
    [15] The system is often overloaded when no suitable agent is available to handle the call when the call is received. The calls are placed in different queues according to some predetermined criteria and also placed in each queue in the order of arrival and / or priority of the calls.
    [16] Information about how long individual or average callers have to wait for their call to be serviced is important to the service provider. This is often the most important factor in determining how to handle a call and serves as an important tool for providing customers with the right services. Estimated latency is also an important measure of service provider performance.
    [17] Numerous techniques have been developed several times to estimate the latency of calls in queues. One estimation technique uses average latency to answer calls in a queue. Another estimation technique uses the wait time of the first call in the queue. In addition, another estimation technique uses the number of calls in the queue divided by the number of agents available to handle the call and multiplied by the average time it takes for the agent to process the call as a wait time estimate. Another estimation technique is known as Weighted Advanced Time (WAT), which is an exponential moving average based on the average progress time (AAT) of each individual call in the queue. This method is disclosed in US Pat. No. 5,506,898, which is incorporated herein by reference.
    [18] This prediction method has been optimized for real time contacts, which may be inaccurate when non-real time contacts are received. As used herein, a real-time contact is a contact in which a contact entity or customer is present during a waiting (or queuing period) and may choose to abandon the contact or not connect if the call is not serviced within the appropriate period. Typical examples include voice calls, VoIP, text chat, video calls, etc. Non-real-time contacts are referred to as contacts that a contact entity or customer can't give up or can't connect to. Includes fax, electronic or written, web form submission, voice messages, etc. For example, non-real-time contacts can be queued at midnight, weekend or holiday when contact centers are not available. The advancing time from one queue to one location can be exceptionally large. This excessive propagation time can be deviated, thereby making the work distribution inefficient and / or ineffective.
    [19] Although some architectures have features that are considered for certain types of events that may deviate from latency predictions, the functionality may not be applicable or desirable for non-real time contacts. For example, one architecture marks or flags all calls in the queue when clock scaling occurs. The wait or progress time for all marked or flagged items is discarded when determining the predicted wait time. This method unnecessarily discards latency information that should be considered when determining the expected latency for non-real time contact.
    [20] The present invention solves these and other problems. The present invention generally relates to architectures for predicting latency in queues that eliminate undesirable deviations in latency prediction caused by predetermined events and / or real time contacts or calls served by the same or different systems in different ways. It relates to an architecture for estimating the waiting time in the queue for non-real time contact or call.
    [21] In one embodiment, the method comprises the following steps:
    [22] (a) providing a first item in a queue, for example at a first time for a work agent to service, at least one of the following events being true:
    [23] (I) the first time is at least one next event, i.e. (i) an event where the queue does not have a work agent (e.g., an unstaffed event in which a switching center (or call or contact center) is not configured), (ii) an event in which the queue is empty (e.g., an event where the contact center has a lazy work agent), (iii) an event in which the contact center (or call or switching center) does not operate (e.g. (Iv) an event that closes or does not provide a service), (iv) an event in which the contact agent's work agent decides not to service the item (e.g., And / or (v) changes to the clock accessed by the contact center with respect to timing information (such as an administrator or otherwise). Follow the occurrence of an event of the year).
    [24] (II) an event in which the time associated with the first item exceeds a predetermined time and
    [25] (b) an event that excludes a time interval, advancing time or other type of timing information associated with the first item from the estimation of the waiting time (eg, WAT, etc.) associated with the queue. The waiting time can be determined using any suitable method including the events described above.
    [26] The first item (or work item) may be a real-time or non-real time contact, which is typically typical.
    [27] This method can have many advantages. First, this method provides an efficient way to provide accurate latency predictions for non-real time contacts. For example, when a contact center is not configured during midnight, weekend or holiday, the first contact serviced will not cause the WAT value to be adjusted up to an illegal amount. Secondly, by immediately following a predetermined type of event or discarding only the affected item, the method avoids wasting unnecessary latency information. Third, by discarding items with advancing time exceeding a predetermined time interval, the method discards items that adversely affect latency prediction and may not be a predetermined type of event.
    [28] In one device, the event for the second item in the queue is false. In this event, a time interval, advancing time or other type of timing information associated with the second item is used when determining the waiting time prediction.
    [29] In another apparatus, the method determines the advancing time for the second item by measuring a time interval between successive queues served from the queue, ie, a service associated with a preceding item (which may be the first item) in the queue. The time is subtracted from the second time. Unlike conventional architectures, the advancing time is determined independently of the order of the second items in the queue.
    [30] In another apparatus, the method may be characterized by a marker, pointer or when the event is detected as being present (or one of the events ((i) to (v)) occurs) for an item in the queue and / or the queue itself. An additional step of setting (or deactivating a marker, pointer or other type of indicator) another indicator.
    [31] In another embodiment,
    [32] (a) providing a first item in a queue for servicing at a first time following the occurrence of a predetermined type of event;
    [33] (b) excluding a time interval, run time, or other type of timing information associated with the first item from the estimate of latency associated with the queue;
    [34] (c) providing a second item in the queue for service at a second time, wherein both the first and second are in the queue when a predetermined type of event occurs;
    [35] (d) using a time interval, run time, or other type of timing information associated with the second item in the estimate of the wait time associated with the queue. This method is particularly useful to account for latency drift that occurs when the system clock changes, the contact center does not operate and / or the work center of the contact center decides not to service the item.
    [36] In another embodiment,
    [37] (a) providing an item in the queue,
    [38] (b) filtering the items in the queue based on at least one predetermined criterion (eg, the event or condition described above) to form a filtered set of items, wherein the at least one item in the queue is filtered The step of excluding from the item of the set,
    [39] (c) determining the progress time for the selected item in the filtered set of items by subtracting the service time associated with the preceding item from the service time associated with the selected item;
    [40] (d) A wait time estimation method is provided that includes determining an estimated wait time in a queue for a particular item based on a travel time. Step (c) typically includes a partial step of determining a measured advancing time (WAT) based on the advancing time for the item in a particular queue, and step (d) is estimated in the queue for the particular item based on the WAT The wait time will be determined.
    [41] In another embodiment, the present invention includes a system architecture that performs the steps of the method described above.
    [42] In another embodiment,
    [43] (a) at least one item in the queue, wherein the first and second items are held in the first queue,
    [44] (b) a memory holding at least one item in the queue,
    [45] (c) a processor in communication with the memory for advancing the dequeued items in at least one item in the queue,
    [46] (d) a filter for filtering items serviced in the first queue to form a set of filtered items, the filtering being of the following type of event: (i) the first queue did not have a work agent (ii) an event in which the first queue was empty, (iii) an event in which the contact center associated with the first queue did not work, (iv) an event in which the contact agent of the contact center decided not to service the item, and / or (v Based on whether the item is queued in the first queue after at least one of the events of the change in the time setting of the clock that provides timing information to the contact center occurs,
    [47] (e) an apparatus for estimating latency of a particular item in an item in a queue comprising a latency estimator for predicting latency of a particular item in a first queue based on a time associated with the item in the filtered set of items Is provided.
    [48] The embodiments and apparatus described above are not complete or exhaustive. As can be appreciated, other embodiments of the invention are possible by using one or more features, alone or in combination, described above or in the following.
    [1] 1 is a block diagram of a contact center incorporating an exemplary embodiment of the invention,
    [2] 2 is a flow chart of a process used to determine average run time;
    [3] 3 is a flow diagram of a process used to mark items in a queue that are not considered when determining an expected latency.
    [4] Explanation of symbols for the main parts of the drawings
    [5] 10: central server 12: database
    [6] 13: another server 14: agent
    [7] 16: email server 20: web server
    [8] 22, 32: communication line 24: fax server
    [9] 28: PBX 30: Memory
    [10] 42, 46: call queue 44: Internet or intranet
    [11] 54: function 58: filter
    [12] 62: indicator
    [49] 1 illustrates an exemplary embodiment of the present invention. A contact center is a set of storage media or databases (including a central server 10) (such as a modified version of Lucent Technologies' CRM Central 2000 Server TM ), contact or customer-related information, and other information that can increase the efficiency and value of contact. 12), multiple servers, namely fax server 24, web server 20, email server 16 and other servers 13, private branch exchange (PBX 28) (or private automatic switchboard) (private automatic exchange (PAX)), a number of work agents 14 that operate computer workstations such as personal computers, and / or telephones or other types of voice communications devices, local area networks (LAN: 36) (or wide area networks (WAN). Fax server 24, web server 20 and email server 16 are connected to the Internet and / or intranet 44 via a communication link 40. Other servers 13 connects optional communication lines 22, 32 Can connect to the PBX 28 and / or the Internet or intranet 44. As can be seen, the other server 13 can be connected to a scanner (usually not connected to the PBX 28, the Internet or intranet 44), Interactive Voice Recognition (IVR) software, VoIP software, video calling software, voice messaging software, IP voice servers, etc. As will be appreciated, faxes may be configured by a suitably equipped personal computer (PSTN). ), Via the Internet or intranet 44. PBX 28, fax server 24, email server 16, web server 20 and database 12 are conventional devices.
    [50] In the architecture of FIG. 1, when the central server 10 sends a voice contact to the agent, the central server 10 may also retrieve information from the database 12 for the agent to provide better service to the customer. By pop-up display) to the agent's computer workstation.
    [51] As can be seen, the central server 10 is a communication component (e.g., PBX 28, fax server 24, email server 16, web server 20) that receives input contacts via LAN 36. And / or other servers 13) are informed of real-time or non-real-time input contacts. The input contact is held by the receiving communication component until it sends a command to the component that causes the central server 10 to send the contact to a particular department or agent 14. The server 10 distributes and connects these contacts to the departments of the available airagent 14 according to a predetermined set of criteria. The agent 14 handles this contact sent by the central server 10.
    [52] The memory 30 includes a plurality of sets 38 call queues 42 and 46. The call queues 42 and 46 of each set 38 are typically responsible for and hold contacts regarding different job types and / or real-time versus non-real-time contacts. In the illustrated embodiment, queue 42 is responsible for non-real-time contact, while queue 46 is responsible for real-time contact. This embodiment is particularly suitable under a Customer Relationship Management (CRM) environment where a customer is allowed to use any media to contact the enterprise. In a CRM environment, both real-time and non-real-time contacts must be handled and distributed with the same efficiency and effectiveness. In queues 42 and 46 of each set 38, each queue has a different priority and / or different type of contact (e.g., email, fax, electronic document or written, web form submission, voice message). , Voice calls, VoIP calls, text chat, video calls, etc.). Priority of contact is determined according to well-known pre-defined criteria. Each queue 42 and 46 typically operates on a first-in, first-out (FIFO) buffer memory and includes a number of entries or locations 50, each of which identifies a corresponding one queued contact. Position 50 at the head of the queue is considered position 1, the next subsequent position 50 is considered position number 2, and so on.
    [53] Memory 30 also includes an estimated latency (EWT) function 54 (or latency predictor). As the name suggests, this function determines an estimate of how long a contact located in queue 42 or 46 must wait before being delivered to agent 14 for service. This estimate is derived individually by the EWT function 54 for each cue 42 or 46 of each set 38.
    [54] For real-time contacts, this estimate is based on the average speed at which the call proceeds through the location 50 of the corresponding queue 46 of contacts. An example implementation of the EWT function 54 for real time contact is disclosed in US Pat. No. 5,506,898.
    [55] For non-real time contacts, this estimate is determined differently than for real time contacts. The system records the time each item is served from each queue. The advancing time is then calculated by measuring the time interval between the service time of the first item at the first location 50 at the head of the queue and the service time of the next second item at the second location. In other words, the running time is given by
    [56] Advance Time = (Service Time of Second Item)-(Service Time of First Item).
    [57] The measured average travel time (WAT) is then determined using the travel time, and the estimated waiting time (EWT) can be determined using the WAT.
    [58] In order to prevent substantial fluctuations in the duration of time for any type of event, a filter 58 is provided. Processor 34 sets indicator 62 such that filter 58 discards the duration of time associated with the marked item when a predetermined type of event occurs. The predetermined types of events are as follows:
    [59] (a) An event where each queue has no work agent available to service an item from the queue. For example, this event occurs when a contact center is not configured after regular work hours. Non-real-time contacts will still remain in the queue for a period that is not configured.
    [60] (b) Event where each queue is empty. For example, this event occurs during a quiet period when there are no items in the queue.
    [61] (c) Event when the system clock changes. For example, this event occurs when the system clock changes to or from daylight savings time. Non-real-time contacts may still remain in the queue during call change.
    [62] (d) The system is not working. For example, this event occurs when the system is temporarily down and then rebooted. Non-real-time contacts can still remain in the queue during the period when the system is off.
    [63] (e) An event that an agent delays the service of a queued item.
    [64] The memory 30 may also include a contact-selection (SEL) function 62. The function 62 is the current or initial wait time or CWT, measured mean progress time (WAT), expected wait time (EWT) and / or predicted wait time for each contact at the head of the queue during the real time. (PWT) (which is the sum of CWT and WAT), and for each available agent, function 62 connects to this agent and selects contacts from queues 42 and / or 46 to be processed. to be. This feature is also disclosed in US Pat. No. 5,905,793, which is incorporated herein by reference.
    [65] The operation of the EWT function 54 will now be described with reference to FIGS. Referring to FIG. 2, at block 100 the processor indicates an item from the queue to be serviced by the work agent. The processor records the time this item progresses from the queue to the work agent. The processor then adjusts the WAT and / or PWT if necessary. In block 104, the processor determines whether the ineligible indicator 62 has been set for the item being serviced from the queue. “Ineligible” means that the following run time calculation will not be used to adjust the wait time prediction (eg WAT, etc.). If not set, the processor determining the WAT new by using the block 108 determines the processing time, as described above in the equation WAT new = at block (112) (WAT old * 0.97) + (advance time * 0.03) Where WAT old is the WAT before the item has been serviced from the queue, and at block 116 the final service time (the service time of the serviced item before the final item is serviced) and the current service time (service time of the last item served Set the same as). If ineligible marker 62 has been set for this item, the processor bypasses blocks 108 and 112 without setting ineligible marker at block 120 and proceeds to block 116.
    [66] The operation of the marker or indicator 62 will now be described with reference to FIG. 3. When one of the aforementioned events occurs, at block 200 and 204 the processor sets the marker to position 1 in the associated queue 42 and / or 46. If an agent defers the service of an item in the queue (block 210), or if the queue does not have a work agent (block 208) or is empty (block 212), the processor may block these two events at block 200. Set the ineligible marker 62 only for queues that are directly contacted by one of the following: In some architectures, this block is performed for real-time queues as well as non-real-time queues, or the system clock is changed (block 216). Or, at an event where the system is rebooted (block 220), the processor sets a marker 62 for all non-real-time queues at block 204, which in some architectures not only for all non-real-time queues but also for real-time queues. 62. Although Figure 1 shows setting a marker to an item in position 1, the marker can also be set for the cue itself.
    [67] Various changes and modifications can be made to the invention. It will be possible to use only certain features of the invention without utilizing other features of the invention. For example, in one other embodiment each cue set 38 includes only one queue of items whose contacts with different priorities are queued in priority.
    [68] In other embodiments, the processor may include any conventional latency prediction method as described in the prior art instead of the method described above.
    [69] In another embodiment, the work agent has discretion when determining whether to handle real time and / or non real time contacts. In this embodiment, the processor 34 flags or marks any item in the queue with respect to the available agent exercising this discretion and determining not to service the item.
    [70] In another embodiment, the filter may be modified to discard timing information about an abnormally long item. For example, the filter may discard the time interval associated with the item being serviced if the time interval exceeds a predetermined time interval period. Alternatively, this survey can be used as a catch-all (after the processor checks for a predetermined event described above) to catch any pseudo time interval that is not due to a predetermined event.
    [71] In another embodiment, the indicator or marker is not set when an item in the queue is considered and inactive or is not set when an item in the queue is discarded by the filter 58. Therefore, the indicator is normally set in this embodiment.
    [72] In another embodiment, the formula for average run time disclosed in US Pat. No. 5,506,898 will be used by default, but the new formulation described above will be used for any contact during the agent free period. In this embodiment, all items in the queue will be marked when the queue has no agents. When servicing an item that is not marked in the queue, the original formula of "time at the queue / original cue location" will be used to calculate the average run time and adjust the WAT. When servicing the marked items in order, the serviced first mark item will be ignored, and subsequent serviced mark items will use the new formula "current service time-previous service time" to calculate progress and adjust the WAT. .
    [73] In another embodiment, queues are not serviced in strict FIFO order. The new formula can be used to accurately calculate the average run time. Since work items are not serviced in FIFO order, it is no longer possible to accurately predict the latency of any individual item in the queue. However, the contact center can determine how long it currently takes to service all items in the queue and evaluate whether the current staff level is inadequate to meet the service time target.
    [74] In another embodiment, the central server of FIG. 1 is replaced with the subscriber device disclosed in US Pat. Nos. 5,905,793, 5,206,903, and 5,206,903, all of which are incorporated herein by reference. In the contact center of FIG. 1, automatic call distribution (ACD) logic, including queues, EWT modules, and work spreaders, is located on an external server separate from the ACD switch.
    [75] In another embodiment, real-time and / or non-real-time contact enters the contact center through a computing component other than the computing component shown in FIG. 1. For example, this component could be another type of server, such as an IP voice server, or another type of communication device, such as an IP switch, voice messaging system, or the like.
    [76] The present invention includes various embodiments, components, methods, processes, systems, and / or apparatus including various embodiments, subcombinations, and sets of parts substantially illustrated and described herein. Those skilled in the art will know how to make and use the invention by understanding the concepts of the invention. The present invention is an event without items not shown and / or described herein, such as those that could be used in existing devices or processes to improve performance, easily implement and / or reduce implementation costs, for example. In various embodiments, including events without items, the present invention includes providing an apparatus and a process.
    [77] The foregoing description of the invention has been presented for purposes of illustration. The foregoing description is not intended to limit the invention to the type disclosed herein. Although the description of the present invention is limited to one or more embodiments and a description of certain changes and modifications, other changes and modifications that would be appreciated by those skilled in the art, for example, if one understood the concept of the invention, are within the scope of the invention. Whether alternatives, exchanges, and / or equivalent structures, functions, ranges, or steps are disclosed herein or otherwise, include other embodiments to the extent that the claims include these alternative exchanges and / or equivalent structures, functions, ranges, or steps. It is to obtain patent rights, not to share a patent subject matter with the public.
    [78] The present invention relates to a method and apparatus for estimating a waiting time, which effectively predicts a waiting time by not properly adjusting the WAT value, not discarding important waiting information, and discarding an item that adversely affects the waiting time prediction. To manage.
    权利要求:
    Claims (33)
    [1" claim-type="Currently amended] A method for estimating the waiting time of a particular item in a queue, the method comprising:
    Providing a first item in a queue for servicing at a first time, wherein the following conditions are met with respect to said first item:
    (a) the first time is followed by the occurrence of a predetermined type of event, wherein the predetermined type of event is (i) an event in which the queue has no working agents, (ii) an event in which the queue is empty, (iii) an event in which a contact center associated with the queue does not operate, and (iv) an event in which a work agent of the contact center decides not to service the item. At least one,
    (b) the providing step, wherein there is at least one of the conditions (a) and (b) that the time associated with the first item exceeds a predetermined time;
    Excluding a time interval associated with the first item from an estimate of a waiting time associated with the queue.
    How to estimate latency.
    [2" claim-type="Currently amended] The method of claim 1,
    And the first item is a non-real-time contact.
    [3" claim-type="Currently amended] The method of claim 1,
    Providing a second item in the queue for servicing at a second time;
    Determining whether any of the conditions (a) and (b) are present with respect to the second item;
    Calculating a time interval for the second item by subtracting a service time associated with a preceding item in the queue at the second time.
    How to estimate latency.
    [4" claim-type="Currently amended] The method of claim 3, wherein
    And the time interval is determined independently of the order of the second item in the queue.
    [5" claim-type="Currently amended] The method of claim 1,
    At least one of the following conditions includes only the condition identified in (a).
    [6" claim-type="Currently amended] The method of claim 1,
    And setting an indicator when at least one of the following conditions is found to be present for an item in the queue.
    [7" claim-type="Currently amended] The method of claim 3, wherein
    The first item is the preceding item,
    And recording the first time in response to the first item being dequeued at the first time and recording the second time in response to the second item being dequeued at the second time. Time estimation method.
    [8" claim-type="Currently amended] A method for estimating the waiting time of a particular item in a queue, the method comprising:
    Providing a first item in a queue for servicing at a first time, the first time following the occurrence of a predetermined type of event;
    Excluding timing information associated with the first item from an estimate of latency associated with the queue;
    Providing a second item in the queue to serve at a second time, wherein both the first and second items were in the queue when the predetermined type of event occurred; and
    Using the timing information associated with the second item in the estimation of a wait time associated with the queue.
    How to estimate latency.
    [9" claim-type="Currently amended] The method of claim 8,
    And the first and second items are non-real time contacts.
    [10" claim-type="Currently amended] The method of claim 8,
    The step of using
    Calculating timing information about the second item by subtracting a service time associated with a preceding item in the queue at the second time.
    [11" claim-type="Currently amended] The method of claim 10,
    The predetermined type of event may include (a) an event of a change to the system clock, (b) an event in which the contact center associated with the queue is not working, and (c) a work agent of the contact center serves the item. At least one of the events determined to not be present.
    [12" claim-type="Currently amended] The method of claim 1,
    And setting an indicator when the event of the predetermined type occurs.
    [13" claim-type="Currently amended] A method for estimating the waiting time of a particular item in a queue, the method comprising:
    Providing an item in the queue,
    Filtering items in the queue based on at least one predetermined criterion to form a filtered set of items, wherein at least one of the items in the queue is an item of the filtered set The filtering step excluded from;
    An advanced time for the selected item in the filtered set of items is obtained by subtracting a service time associated with an preceding item from a service time associated with a selected item. Determining,
    Determining an estimated waiting time in the queue for a particular item based on the advancing time,
    How to estimate latency.
    [14" claim-type="Currently amended] The method of claim 13,
    Wait time estimation method of all the items in the queue.
    [15" claim-type="Currently amended] The method of claim 13,
    And the advancing time is determined independently of the order of the second item in the queue.
    [16" claim-type="Currently amended] The method of claim 13,
    The predetermined criterion is
    (a) the service time of the item is an event of a predetermined type, wherein the event of the predetermined type is (i) an event in which the queue does not have a work agent, (ii) an event in which the queue is empty, (iii) An event following the occurrence of an event in which an associated contact center is inoperative, (iv) an event in which the work agent of the contact center decides not to service the item, and (v) an event in which the system clock of the contact center is changed;
    (b) a wait time estimation method of at least one of events in which a time associated with the item exceeds a predetermined time.
    [17" claim-type="Currently amended] The method of claim 13,
    And setting an indicator when the predetermined criterion is retrieved as present.
    [18" claim-type="Currently amended] A method for servicing real-time and non-real-time contact in a common contact center,
    Determining a first advancing time in a first queue for a real time contact using a first algorithm,
    Determining a second advancing time in a second queue for a non-real time contact using a second algorithm,
    The first algorithm is different from the second algorithm.
    Real-time and non-real-time contact service method.
    [19" claim-type="Currently amended] The method of claim 18,
    Wherein the first algorithm includes ordering information associated with the first queue and the second algorithm excludes ordering information associated with the second queue.
    [20" claim-type="Currently amended] The method of claim 18,
    Excluding timing information of an item in at least one of the first and second queues from a latency prediction associated with at least one of the first and second queues, wherein the item follows an event of a predetermined type. Real-time and non-real-time contact service method.
    [21" claim-type="Currently amended] An apparatus for estimating a waiting time of a specific item of an item in a queue,
    Indicating means for indicating a first item in the queue for service by the work agent at a first time, the following condition with respect to the first item, namely:
    (a) the first time is an event of a predetermined type, the event of the predetermined type being (i) an event in which the queue does not have a work agent, (ii) an event in which the queue is empty, (iii) At least one of an event in which an associated contact center does not operate and (iv) an event in which the work agent of the contact center decides not to service the item;
    (b) there is at least one of the events in which the time associated with the first item exceeds a predetermined time,
    Filtering means for excluding timing information associated with the first item from an estimate of latency associated with the queue;
    Latency estimation device.
    [22" claim-type="Currently amended] The method of claim 21,
    And the first item is a non-real time contact.
    [23" claim-type="Currently amended] The method of claim 21,
    Indicating means for indicating a second item in the queue for serving at a second time, the filtering means wherein any of the conditions (a) and (b) is present with respect to the second item; A waiting time estimating apparatus for determining whether or not to do so.
    [24" claim-type="Currently amended] The method of claim 21,
    And the timing information is determined independently of the order of the second item in the queue.
    [25" claim-type="Currently amended] The method of claim 21,
    Setting means for setting the indicator when at least one of the following conditions is found to be present for an item in the queue.
    [26" claim-type="Currently amended] An apparatus for estimating the waiting time of a particular item in a queue of items, the apparatus comprising:
    Instructing means for indicating a first item which is a queue for servicing at a first time, the queue having first and second items;
    Filtering means for including a time interval associated with the second item in the estimate of the waiting time associated with the queue except for a time interval associated with the first item from the estimation of the waiting time associated with the queue;
    Determining means for determining a time interval associated with the second item by subtracting a service time associated with a preceding item at the first time;
    Latency estimation means for estimating latency in the queue for a particular item using the time interval;
    Latency estimation device.
    [27" claim-type="Currently amended] The method of claim 26,
    Wait time estimation apparatus, wherein each of the items in the queue is a non-real time contact.
    [28" claim-type="Currently amended] The method of claim 26,
    The filtering means is then ie
    (a) the service time of the item is an event of a predetermined type, wherein the event of the predetermined type is (i) an event in which the queue does not have a work agent, (ii) an event in which the queue is empty, (iii) At least one of an event in which an associated contact center does not operate, (iv) an event in which the work agent of the contact center decides not to service the item, and (v) an event in which the system clock of the contact center is changed. event,
    (b) a latency estimation apparatus for filtering based on one or more predetermined criteria, wherein at least one of the events in which the time associated with the item exceeds a predetermined time.
    [29" claim-type="Currently amended] An apparatus for estimating the waiting time of a particular item in a queue of items, the apparatus comprising:
    Includes items in at least one queue, the first and second items are held in a first queue,
    A memory holding an item in said at least one queue,
    A processor in communication with the memory for advancing an item queued in an item in the at least one queue;
    A filter for filtering items serviced in the first queue to form a set of filtered items, the filter comprising at least one of the following types of events, i.e. the first queue did not have a work agent; An event in which (ii) the first queue was empty, (iii) the contact center associated with the first queue was inoperative, and (iv) an event agent at the contact center decided not to service the item. Based on whether the item was queued in the first queue after is generated.
    A latency predictor for predicting latency of a particular item in the first queue based on a time associated with the item in the set of filtered items
    Latency estimation device.
    [30" claim-type="Currently amended] The method of claim 29,
    And the first and second items are non-real time contacts.
    [31" claim-type="Currently amended] The method of claim 29,
    The latency estimation apparatus calculates a progress time by subtracting a service time associated with the first item from a service time associated with the second item.
    [32" claim-type="Currently amended] The method of claim 29,
    At least one of the following conditions comprises only conditions ((i), (ii) and (iii)).
    [33" claim-type="Currently amended] The method of claim 29,
    And an indicator when the at least one next condition is detected to be present.
    类似技术:
    公开号 | 公开日 | 专利标题
    US20170195492A1|2017-07-06|Task Assignments to Workers
    US10536578B2|2020-01-14|Pacing in knowledge worker engagement
    US10291781B2|2019-05-14|Best match interaction set routing
    US9031087B2|2015-05-12|Method and apparatus for optimizing response time to events in queue
    US20160277574A1|2016-09-22|System for indicating priority levels for transaction and task engagement in a call center
    US8582752B1|2013-11-12|System and method for optimal outbound dialing in virtual contact center
    EP1107556B1|2007-01-24|System for automatically predicting call center agent work time in a multi-skilled agent environment
    US7689426B2|2010-03-30|Method and apparatus for distributed interactive voice processing
    US6535600B1|2003-03-18|System for automatically routing calls to call center agents in an agent surplus condition based on service levels
    US5214688A|1993-05-25|Method and apparatus for dynamic and interdependent processing of inbound calls and outbound calls
    US8660256B2|2014-02-25|Dynamic load balancing between multiple locations with different telephony system
    EP0985308B1|2005-11-09|Apparatus and methods enhancing call routing within and between call-centers
    US6347139B1|2002-02-12|System for automatically routing calls to call center agents in an agent surplus condition based on agent occupancy
    US6510221B1|2003-01-21|System for automatically routing calls to call center agents in an agent surplus condition based on delay probabilities
    US8180043B2|2012-05-15|Method and apparatus for customer key routing
    US20170214797A1|2017-07-27|System and method for providing call-back options
    EP0577340B1|2001-07-04|Telephone usage monitoring arrangement
    USRE37001E|2000-12-26|Interactive call processor to facilitate completion of queued calls
    AU729882B2|2001-02-15|Work at home ACD agent network with cooperative control
    US4881261A|1989-11-14|Method for predictive pacing of calls in a calling system
    US7715332B2|2010-05-11|Determining presence for interaction routing
    US8781092B2|2014-07-15|Systems and methods for callback processing
    EP1168791B1|2005-05-11|Callback to a call center
    US8705725B1|2014-04-22|Member-initiated outbound call
    US6937715B2|2005-08-30|Contact center management
    同族专利:
    公开号 | 公开日
    JP4312970B2|2009-08-12|
    EP1150236B1|2012-06-13|
    CA2344767C|2006-08-15|
    JP4386901B2|2009-12-16|
    US6754333B1|2004-06-22|
    CA2344767A1|2001-10-27|
    JP2002051149A|2002-02-15|
    EP1150236A3|2002-09-25|
    EP1150236A2|2001-10-31|
    JP2006271004A|2006-10-05|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    2000-04-27|Priority to US20052000P
    2000-04-27|Priority to US60/200,520
    2000-08-17|Priority to US09/641,403
    2000-08-17|Priority to US09/641,403
    2001-04-26|Application filed by 빈 토마스 제이, 아바야 테크놀러지 코퍼레이션
    2001-11-08|Publication of KR20010098915A
    优先权:
    申请号 | 申请日 | 专利标题
    US20052000P| true| 2000-04-27|2000-04-27|
    US60/200,520|2000-04-27|
    US09/641,403|2000-08-17|
    US09/641,403|US6754333B1|2000-04-27|2000-08-17|Wait time prediction arrangement for non-real-time customer contacts|
    [返回顶部]