TWI569619B - Message recording server, message data management system and method thereof - Google Patents
Message recording server, message data management system and method thereof Download PDFInfo
- Publication number
- TWI569619B TWI569619B TW104108398A TW104108398A TWI569619B TW I569619 B TWI569619 B TW I569619B TW 104108398 A TW104108398 A TW 104108398A TW 104108398 A TW104108398 A TW 104108398A TW I569619 B TWI569619 B TW I569619B
- Authority
- TW
- Taiwan
- Prior art keywords
- call
- data
- voice
- voice recognition
- server
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 25
- 238000013523 data management Methods 0.000 title claims description 21
- 238000004891 communication Methods 0.000 claims description 61
- 238000007726 management method Methods 0.000 claims description 56
- 238000012806 monitoring device Methods 0.000 claims description 54
- 230000005540 biological transmission Effects 0.000 claims description 25
- 239000000463 material Substances 0.000 claims description 14
- 230000004044 response Effects 0.000 claims 1
- 238000004458 analytical method Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 238000003491 array Methods 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/22—Arrangements for supervision, monitoring or testing
- H04M3/2281—Call monitoring, e.g. for law enforcement purposes; Call tracing; Detection or prevention of malicious calls
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/1066—Session management
- H04L65/1083—In-session procedures
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/42221—Conversation recording systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/53—Centralised arrangements for recording incoming messages, i.e. mailbox systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M7/00—Arrangements for interconnection between switching centres
- H04M7/006—Networks other than PSTN/ISDN providing telephone service, e.g. Voice over Internet Protocol (VoIP), including next generation networks with a packet-switched transport layer
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M2201/00—Electronic components, circuits, software, systems or apparatus used in telephone systems
- H04M2201/40—Electronic components, circuits, software, systems or apparatus used in telephone systems using speech recognition
Description
本發明是關於一種記錄並管理網際網路協定(Internet Protocol，IP)電話網中的通話聲音的資料的通話錄音伺服器(sever)、通話資料(data)管理系統(system)、及通話資料管理方法。 The present invention relates to a call recording server (sever), a call data management system (system), and a call data management for recording and managing data of a call voice in an Internet Protocol (IP) telephone network. method.
先前，出於提昇客服中心(call center)的服務(service)品質等各種目的，而對通話內容進行確認或監視(以下稱「監視」)。另外，近年來，使用網際網路語音通訊協定(Voice over Internet Protocol，VoIP)技術的IP電話正在不斷普及。因此，存在各種用以記錄並管理IP電話網中的通話聲音的資料的技術(例如參照專利文獻1)。 Previously, the contents of the call were confirmed or monitored for various purposes such as improving the quality of the service of the call center (hereinafter referred to as "monitoring"). In addition, in recent years, IP phones using Voice over Internet Protocol (VoIP) technology are spreading. Therefore, there are various techniques for recording and managing data of a call voice in an IP telephone network (for example, refer to Patent Document 1).
於專利文獻1所記載的技術(以下稱為「現有技術」)中，客服中心的操作員(operator)終端於電話對應結束之後，將藉由操作員製作成的業務履歷資訊、及對通話聲音的聲音資料進行的聲音辨識處理的結果即聲音辨識資訊發送至伺服器。管理者 的終端獲取業務履歷資訊及聲音辨識資訊並提示給管理者。 In the technique described in Patent Document 1 (hereinafter referred to as "prior art"), the operator terminal of the call center, after the telephone correspondence is completed, the business history information created by the operator and the voice of the call. The result of the voice recognition processing performed by the voice data is that the voice recognition information is transmitted to the server. Manager The terminal obtains business history information and voice recognition information and presents it to the manager.
根據該現有技術，於通話結束後管理者可確認對IP電話網中的通話聲音的聲音辨識結果。即，藉由使用現有技術，可監視IP電話的通話內容。 According to this prior art, the manager can confirm the voice recognition result of the call voice in the IP telephone network after the call ends. That is, by using the prior art, the content of the call of the IP telephone can be monitored.
[先前技術文獻] [Previous Technical Literature]
[專利文獻] [Patent Literature]
專利文獻1：日本專利特開2008-211271號公報 Patent Document 1: Japanese Patent Laid-Open Publication No. 2008-211271
然而，於各通話結束後，提取(pickup)需要確認的通話、且檢索所提取的通話的聲音資料並確認所儲存的聲音辨識結果或聲音資料的作業耗費工時。於如大規模的客服中心般IP電話機的個數多的情況下，儲存於伺服器中的上述業務履歷資訊及聲音辨識資訊的量亦變多，該作業變得非常煩雜。因此，現有技術難以應用於成為監視對象的IP電話機個數多的情況(case)。 However, after the end of each call, the call that needs to be confirmed is picked up, and the sound data of the extracted call is retrieved, and the stored work of the voice recognition result or the voice data is confirmed. In the case of a large number of IP telephones such as a large-scale customer service center, the amount of business history information and voice recognition information stored in the server is also increased, and the operation becomes very complicated. Therefore, the prior art is difficult to apply to the case where the number of IP telephones to be monitored is large.
本發明的目的在於提供一種可更簡單地監視多個IP電話機的通話內容的通話錄音伺服器、通話資料管理系統及通話資料管理方法。 An object of the present invention is to provide a call recording server, a call data management system, and a call data management method that can more easily monitor the contents of a call of a plurality of IP telephones.
本發明的通話錄音伺服器包括：聲音記錄控制部，依序獲取以IP電話網傳輸的通話的聲音資料並記錄於記憶體(memory)中；通話開始獲取部，根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序(timing)；以及聲音辨識控制部，於所獲取的上述通話開始時序後立即使對所記錄的上述聲音資料的聲音辨識處理開始。 The call recording server of the present invention comprises: a sound recording control unit that sequentially acquires sound data of a call transmitted by the IP telephone network and records it in a memory; the call start acquisition unit is provided according to the attached The voice information control information acquires a call start timing for starting the call, and the voice recognition control unit starts the voice recognition process for the recorded voice data immediately after the acquired call start sequence.
本發明的通話資料管理系統包括：通話錄音伺服器，記錄以IP電話網傳輸的通話的聲音資料；聲音辨識伺服器，對所記錄的上述聲音資料進行聲音辨識處理，而產生作為上述聲音辨識處理的結果的文字資料(text data)；以及監視(monitoring)裝置，將所記錄的上述聲音資料與所產生的上述文字資料建立對應而進行提示，其中上述通話錄音伺服器包括：聲音記錄控制部，自IP電話網依序獲取上述聲音資料並記錄於記憶體；通話開始獲取部，根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序；以及聲音辨識控制部，將所記錄的上述聲音資料輸出至上述聲音辨識伺服器，使上述聲音辨識伺服器於所獲取的上述通話開始時序後立即開始對上述聲音資料的聲音辨識處理。 The call data management system of the present invention comprises: a call recording server for recording the voice data of the call transmitted by the IP telephone network; and a voice recognition server for performing voice recognition processing on the recorded voice data, and generating the voice recognition processing as the voice recognition processing a result of the text data; and a monitoring device, wherein the recorded audio data is associated with the generated text data, wherein the call recording server includes: a sound recording control unit, The voice data is sequentially acquired from the IP telephone network and recorded in the memory; the call start acquisition unit acquires a call start sequence for starting the call according to the control information attached to the acquired voice data; and the voice recognition control unit And outputting the recorded sound data to the sound recognition server, so that the sound recognition server starts sound recognition processing on the sound data immediately after the acquired call start timing.
本發明的通話資料管理方法包括如下步驟：依序獲取以IP電話網傳輸的通話的聲音資料並記錄於記憶體中；根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序；以及於所獲取的上述通話開始時序後立即使對所記錄的上述聲音資料的聲音辨識處理開始。 The call data management method of the present invention comprises the steps of: sequentially acquiring the sound data of the call transmitted by the IP telephone network and recording it in the memory; and obtaining the start of the call according to the control information attached to the acquired sound data. The call start sequence; and the voice recognition process for the recorded voice material is started immediately after the acquired call start sequence.
根據本發明，由於自通話開始時序後立即開始對以IP電話網傳輸的通話的聲音資料的聲音辨識處理，因此可於通話中幾乎實時(realtime)地提示聲音辨識結果。因此，根據本發明，可更簡單地監視多個IP電話機的通話內容。 According to the present invention, since the voice recognition processing of the voice data of the call transmitted by the IP telephone network is started immediately after the call start sequence, the voice recognition result can be presented in real time in the call. Therefore, according to the present invention, it is possible to more easily monitor the contents of the call of a plurality of IP telephones.
100‧‧‧通訊系統 100‧‧‧Communication system
200‧‧‧外線網 200‧‧‧External network
300‧‧‧內線網 300‧‧‧Internal network
310‧‧‧電話機 310‧‧‧ telephone
320‧‧‧網路機器 320‧‧‧Network Machine
330‧‧‧PBX裝置 330‧‧‧PBX device
3101~310N‧‧‧第1~第N電話機 310 1 ~ 310 N ‧‧‧1st to Nth telephones
400‧‧‧通話管理網 400‧‧‧Call Management Network
410‧‧‧通話錄音伺服器 410‧‧‧Call Recording Server
411‧‧‧電話網通訊部 411‧‧‧Telephone Network Communications Department
412‧‧‧管理網通訊部 412‧‧‧Management Network Communications Department
413‧‧‧記憶體 413‧‧‧ memory
414‧‧‧聲音記錄控制部 414‧‧‧ Sound Recording Control Department
415‧‧‧通話開始獲取部 415‧‧‧ Call Start Acquisition Department
416‧‧‧聲音辨識控制部 416‧‧‧Sound Identification Control Department
420‧‧‧管理伺服器 420‧‧‧Management Server
430‧‧‧聲音辨識伺服器 430‧‧‧Sound Identification Server
440‧‧‧監視裝置 440‧‧‧Monitor
圖1是表示包含本發明的一實施方式的通話資料管理系統的通訊系統的構成的一例的系統構成圖。 1 is a system configuration diagram showing an example of a configuration of a communication system including a call data management system according to an embodiment of the present invention.
圖2是表示本實施方式的通話錄音伺服器的構成的一例的方塊圖(block diagram)。 FIG. 2 is a block diagram showing an example of a configuration of a call recording server according to the present embodiment.
圖3是表示本實施方式的通話錄音伺服器的動作的一例的流程圖(flow chat)。 FIG. 3 is a flowchart (flow chat) showing an example of the operation of the call recording server according to the present embodiment.
圖4是表示本實施方式的通訊系統的動作的流程的一例的序列(sequence)圖。 4 is a sequence diagram showing an example of a flow of an operation of the communication system according to the embodiment.
以下，參照圖式對本發明的一實施方式進行詳細說明。本實施方式為將本發明應用於配置有多個IP電話機的客服中心的通話監視系統的情況的具體形態的一例。 Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. The present embodiment is an example of a specific form of the case where the present invention is applied to a call monitoring system of a customer service center in which a plurality of IP telephones are arranged.
〈系統構成〉 <System Components>
首先，對包含本實施方式的通話資料管理系統的通訊系統的構成進行說明。 First, the configuration of a communication system including the call material management system of the present embodiment will be described.
圖1是表示包含本實施方式的通話資料管理系統的通訊系統的構成的一例的系統構成圖。 FIG. 1 is a system configuration diagram showing an example of a configuration of a communication system including the call data management system according to the present embodiment.
於圖1中，通訊系統100具有外線網200、內線網300、及通話管理網400。 In FIG. 1, the communication system 100 has an external network 200, an internal network 300, and a call management network 400.
外線網200為網際網路(internet)等公共網，其是連接有客服中心的顧客所使用的IP終端(未圖示)的通訊網路(network)。即，外線網200構成客服中心所形成的IP電話網的一部分。 The external network 200 is a public network such as the Internet, which is a communication network of an IP terminal (not shown) used by customers connected to the customer service center. That is, the external network 200 constitutes a part of the IP telephone network formed by the customer service center.
內線網300為構築於客服中心的區域網路(Local Area Network，LAN)等通訊網路的一部分。內線網300具有第1電話機3101~第N電話機310N、網路機器320、及專用交換分機(Private Branch eXchange，PBX)裝置330。 The internal network 300 is a part of a communication network such as a local area network (LAN) built in a customer service center. The extension network 300 has a first telephone 310 1 to an Nth telephone 310 N , a network device 320, and a Private Branch eXchange (PBX) device 330.
各電話機310是進行顧客對應的操作員所使用的IP電話機。第1電話機3101~第N電話機310N經由網路機器320而分別連接於PBX(Private Branch eXchange)裝置330。 Each telephone 310 is an IP telephone used by an operator corresponding to the customer. The first telephone 310 1 to the Nth telephone 310 N are respectively connected to a PBX (Private Branch eXchange) device 330 via the network device 320.
網路機器320是於各電話機310與PBX裝置330之間進行IP封包(packet)的傳輸的中繼裝置，例如為交換式集線器(switching hub)、接線盒(TAP box)、或路由器(router)。其中，網路機器320藉由端口鏡像(port mirroring)等功能，而將傳送的IP封包的複製件發送至通話管理網400。 The network device 320 is a relay device that performs IP packet transmission between each of the telephones 310 and the PBX device 330, such as a switching hub, a TAP box, or a router. . The network device 320 sends a copy of the transmitted IP packet to the call management network 400 by a function such as port mirroring.
PBX裝置330為專用交換機，連接於外線網200。PBX裝置330自外線網200接收以第1電話機3101~第N電話機310N為目的地的IP封包，並將該IP封包傳送至網路機器320。另外，PBX裝置330自網路機器320接收以外線網200的IP電話機(未圖示)為目的地的IP封包，並將該IP封包傳送至外線網200。 The PBX device 330 is a private switch connected to the external network 200. The PBX device 330 receives an IP packet destined for the first telephone 310 1 to the Nth telephone 310 N from the external network 200, and transmits the IP packet to the network device 320. Further, the PBX device 330 receives an IP packet destined for the IP telephone (not shown) of the external network 200 from the network device 320, and transmits the IP packet to the external network 200.
即，內線網300構成IP電話網的一部分，一面傳輸於客服中心進行的多個通話的IP封包，一面將傳輸的IP封包的複製件發送至通話管理網400。 That is, the extension network 300 constitutes a part of the IP telephone network, and transmits a copy of the transmitted IP packet to the call management network 400 while transmitting the IP packet of the plurality of calls performed by the call center.
通話管理網400例如為構築於客服中心的LAN等通訊網路的一部分，且為與本發明的通話資料管理系統對應的部分。通話管理網400具有通話錄音伺服器410、管理伺服器420、聲音辨識伺服器430、及監視裝置440。 The call management network 400 is, for example, a part of a communication network such as a LAN built in a customer service center, and is a part corresponding to the call material management system of the present invention. The call management network 400 has a call recording server 410, a management server 420, a voice recognition server 430, and a monitoring device 440.
此外，各裝置的連接關係並不限定於圖1所示的連接線。各裝置例如分別連接於LAN，於任何裝置間均可進行通訊。 Further, the connection relationship of each device is not limited to the connection line shown in FIG. 1. Each device is, for example, connected to a LAN, and can communicate with any device.
通話錄音伺服器410連接於內線網300的網路機器320。通話錄音伺服器410接收自網路機器320發送的IP封包，且自所接收到的IP封包提取通話的聲音資料並加以記錄。即，通話錄音伺服器410記錄以IP電話網傳輸的通話的聲音資料。 The call recording server 410 is connected to the network machine 320 of the internal network 300. The call recording server 410 receives the IP packet sent from the network device 320, and extracts and records the voice data of the call from the received IP packet. That is, the call recording server 410 records the voice data of the call transmitted on the IP telephone network.
圖2是表示通話錄音伺服器410的構成的一例的方塊圖。 FIG. 2 is a block diagram showing an example of the configuration of the call recording server 410.
於圖2中，通話錄音伺服器410具有電話網通訊部411、管理網通訊部412、記憶體413、聲音記錄控制部414、通話開始獲取部415、及聲音辨識控制部416。 In FIG. 2, the call recording server 410 has a telephone network communication unit 411, a management network communication unit 412, a memory 413, a voice recording control unit 414, a call start acquisition unit 415, and a voice recognition control unit 416.
電話網通訊部411是用以連接於內線網300的通訊網路的通訊介面(interface)，且連接於網路機器320。電話網通訊部411接收自網路機器320發送的IP封包，並將所接收到的IP封包逐次輸出至聲音記錄控制部414及通話開始獲取部415。 The telephone network communication unit 411 is a communication interface for connecting to the communication network of the extension network 300, and is connected to the network device 320. The telephone network communication unit 411 receives the IP packet transmitted from the network device 320, and sequentially outputs the received IP packet to the voice recording control unit 414 and the call start acquisition unit 415.
管理網通訊部412是用以連接於通話管理網400的通訊網路的通訊介面，且連接於管理伺服器420、聲音辨識伺服器430、及監視裝置440。 The management network communication unit 412 is a communication interface for connecting to the communication network of the call management network 400, and is connected to the management server 420, the voice recognition server 430, and the monitoring device 440.
記憶體413為硬碟(hard disk)等記錄媒體，且可讀出地保持自聲音記錄控制部414儲存的資訊。 The memory 413 is a recording medium such as a hard disk, and readable and retains information stored in the sound recording control unit 414.
聲音記錄控制部414對所輸入的IP封包進行解析，並自各個IP封包提取聲音資料(通話聲音信號)及控制資訊(通訊控制信號)。然後，聲音記錄控制部414將提取的聲音資料與控制資訊等確定聲音資料的資訊建立對應而逐次儲存於記憶體413。 即，聲音記錄控制部414自IP電話網依序獲取聲音資料並記錄於記憶體413。 The voice recording control unit 414 analyzes the input IP packet, and extracts voice data (call voice signal) and control information (communication control signal) from each IP packet. Then, the sound recording control unit 414 sequentially stores the extracted sound data in the memory 413 in association with the information of the determined sound data such as the control information. That is, the sound recording control unit 414 sequentially acquires sound data from the IP telephone network and records it in the memory 413.
聲音資料為包含通話中的雙方的說話人的說話聲音的聲頻資料。控制資訊為隨附於聲音資料的資訊，包含通話識別資訊、說話人識別資訊、及時刻資訊。通話識別資訊為用以識別通話的資訊(例如雙方的電話號碼)。說話人識別資訊為用以識別聲音資料中所包含的說話聲音的說話人(IP電話機)的資訊。時刻資訊為表示聲音資料所對應的時刻的資訊。控制資訊可自IP封包的標頭(header)部分獲取，亦可自IP封包的有效負載(payload)部分獲取。 The sound data is audio data containing the voice of the speaker of both parties in the call. The control information is information attached to the sound data, including call identification information, speaker identification information, and time information. The call identification information is information for identifying the call (for example, the telephone numbers of both parties). The speaker identification information is information for identifying a speaker (IP telephone) of the voice included in the voice material. The time information is information indicating the time corresponding to the sound data. Control information can be obtained from the header portion of the IP packet or from the payload portion of the IP packet.
通話開始獲取部415對輸入的IP封包進行解析，並自各個IP封包提取包含通話識別資訊的控制資訊。通話開始獲取部415根據所提取的控制資訊，而針對每一通話確定電話網通訊部411最初接收到該通話的IP封包的時序。通話開始獲取部415獲取所確定的上述時序作為開始該通話的時序(以下稱為「通話開始時序」)。而且，通話開始獲取部415每當獲取通話開始時序時，將通話開始時序與所對應的通話的控制資訊一併通知給聲音辨識控制部416。 The call start acquisition unit 415 analyzes the input IP packet and extracts control information including the call identification information from each IP packet. The call start acquisition unit 415 determines, based on the extracted control information, the timing at which the telephone network communication unit 411 initially receives the IP packet of the call for each call. The call start acquisition unit 415 acquires the determined timing as the timing at which the call is started (hereinafter referred to as "call start timing"). Further, each time the call start acquisition unit 415 acquires the call start timing, the call start timing is notified to the voice recognition control unit 416 together with the control information of the corresponding call.
此外，通話開始獲取部415於控制資訊中包含表示通話開始時刻的資訊等直接表示通話開始時序的資訊的情況下，亦可自該資訊獲取通話開始時序。 Further, when the control information includes the information indicating the call start time or the like, and the information indicating the call start timing is directly included in the control information, the call start acquisition unit 415 may acquire the call start timing from the information.
另外，自IP封包提取聲音資料及控制資訊亦可利用電話網通訊部411進行。 In addition, the voice data and the control information can be extracted from the IP packet by the telephone network communication unit 411.
聲音辨識控制部416當被通知通話開始時序時，經由管 理網通訊部412而對管理伺服器420發送表示通話開始時序的通話開始通知。於通話開始通知中包含例如控制資訊。 The voice recognition control unit 416, when notified of the call start timing, via the pipe The network communication unit 412 transmits a call start notification indicating the call start timing to the management server 420. For example, control information is included in the call start notification.
另外，聲音辨識控制部416當經由管理網通訊部412而接收到發送聲音資料的請求(以下稱為「聲音發送請求」)時，將記憶體413中所記錄的成為請求對象的聲音資料回復至請求源。於聲音發送請求中包含控制資訊等確定聲音資料的資訊。另外，聲音發送請求例如是自聲音辨識伺服器430及監視裝置440發送。聲音發送請求例如為如下內容，即一指定通話識別資訊並儲存所對應的通話的聲音資料，就立即請求逐次回復該聲音資料。 When the voice recognition control unit 416 receives the request to transmit the voice material (hereinafter referred to as "sound transmission request") via the management network communication unit 412, the voice recognition control unit 416 restores the voice data to be requested in the memory 413 to Request source. The sound transmission request includes information for determining sound information such as control information. Further, the voice transmission request is transmitted from the voice recognition server 430 and the monitoring device 440, for example. The voice transmission request is, for example, a message specifying the call identification information and storing the voice data of the corresponding call, and immediately requests the voice data to be replied successively.
於下文敍述，當發送通話開始通知時，藉由管理伺服器420的管理功能，聲音辨識伺服器430對通話錄音伺服器410請求聲音資料而開始對所回復的聲音資料進行聲音辨識處理。 As will be described later, when the call start notification is transmitted, by the management function of the management server 420, the voice recognition server 430 requests the voice recording server 410 to perform voice recognition processing on the voice data to be recovered.
即，聲音辨識控制部416將發送通話開始通知的結果所記錄的聲音資料輸出至聲音辨識伺服器430，於通話開始時序後立即使聲音辨識伺服器430開始對記憶體413中所記錄的聲音資料進行聲音辨識處理。 That is, the voice recognition control unit 416 outputs the voice data recorded as a result of the transmission of the call start notification to the voice recognition server 430, and causes the voice recognition server 430 to start the voice data recorded in the memory 413 immediately after the call start timing. Perform sound recognition processing.
圖1的管理伺服器420藉由接收自通話錄音伺服器410發送的通話開始通知而獲取通話開始時序。然後，管理伺服器420根據所獲取的通話開始時序，而控制通話錄音伺服器410、聲音辨識伺服器430、及監視裝置440各自的動作時序。 The management server 420 of FIG. 1 acquires the call start timing by receiving the call start notification transmitted from the call recording server 410. Then, the management server 420 controls the respective operation timings of the call recording server 410, the voice recognition server 430, and the monitoring device 440 based on the acquired call start timing.
更具體而言，管理伺服器420當接收到通話開始通知時，根據通話開始通知中所包含的控制資訊，而決定是否對通話開始通知所顯示的通話聲音資料進行聲音辨識處理。 More specifically, when receiving the call start notification, the management server 420 determines whether or not to perform voice recognition processing on the call voice data displayed by the call start notification based on the control information included in the call start notification.
而且，管理伺服器420於判斷為進行聲音辨識的情況 下，對聲音辨識伺服器430發送開始對通話錄音伺服器410中所記錄的聲音資料進行聲音辨識處理的請求(以下稱為「辨識開始請求」)。於辨識開始請求中包含控制資訊等確定聲音資料的資訊。 Moreover, the management server 420 determines that the voice recognition is performed. Next, the voice recognition server 430 transmits a request to start voice recognition processing on the voice data recorded in the call recording server 410 (hereinafter referred to as "identification start request"). The identification start request includes information for determining sound information such as control information.
另外，管理伺服器420於判斷為進行聲音辨識的情況下，對監視裝置440傳送通話開始通知。進而，管理伺服器420當自聲音辨識伺服器430接收到開始聲音辨識處理的意旨的通知(以下稱為「辨識開始通知」)時，將該辨識開始通知傳送至監視裝置。於辨識開始通知中包含控制資訊等確定聲音資料的資訊。 Further, when it is determined that the voice recognition is performed, the management server 420 transmits a call start notification to the monitoring device 440. Further, when the management server 420 receives a notification (hereinafter referred to as "identification start notification") for starting the voice recognition processing from the voice recognition server 430, the management server 420 transmits the identification start notification to the monitoring device. The identification start notification includes information for determining sound information such as control information.
聲音辨識伺服器430當接收到辨識開始請求時，對通話錄音伺服器410發送以與辨識開始請求相同的聲音資料為對象的聲音發送請求。然後，聲音辨識伺服器430對自通話錄音伺服器410回復的聲音資料進行聲音辨識處理，產生作為聲音辨識處理的結果的文字資料，並儲存於聲音辨識伺服器430的記憶體(未圖示)。 Upon receiving the recognition start request, the voice recognition server 430 transmits a voice transmission request to the call recording server 410 for the same voice material as the recognition start request. Then, the voice recognition server 430 performs voice recognition processing on the voice data returned from the call recording server 410, generates text data as a result of the voice recognition process, and stores the voice data in the voice recognition server 430 (not shown). .
聲音辨識伺服器430進行藉由公知的聲音辨識技術的聲音辨識處理。例如，聲音辨識伺服器具有聲音辨識資料庫(database)、聲頻分析部、及辨識解碼器(decoder)部(均未圖示)。 The voice recognition server 430 performs voice recognition processing by a well-known voice recognition technique. For example, the voice recognition server has a voice recognition database, an audio analysis unit, and a decoder portion (none of which are shown).
聲音辨識資料庫預先儲存有聲頻示範(model)、字典、及語言示範。聲頻示範是將聲音的特徵量與音標(phonetic symbol)的概率性的關聯資料化所得者。字典記述有多個文字排列作為聲音辨識處理的聲音辨識結果的候補群。語言示範是針對字典中所記述的各個文字排列將出現概率或連接概率資料化而得到的。 The voice recognition database is pre-stored with audio models, dictionaries, and language demonstrations. The audio demonstration is obtained by correlating the feature quantity of the sound with the probability of the phonetic symbol. The dictionary describes a plurality of character arrays as candidate groups for the sound recognition result of the voice recognition processing. The language demonstration is obtained by digitizing the probability of occurrence or the probability of connection for each character arrangement described in the dictionary.
聲頻分析部對聲音信號進行訊框(frame)處理，對每一訊框進行包含傅立葉(Fourier)解析的規定處理而提取聲音特 徵量。然後，聲頻分析部自解析結果檢測包含說話聲音的聲音區間，而產生僅包含聲音區間的聲音特徵量的時間序列資料。 The audio analysis unit performs frame processing on the sound signal, and performs predetermined processing including Fourier analysis on each frame to extract the sound. Signs. Then, the audio analysis unit detects the sound section including the speech sound from the analysis result, and generates time-series data including only the sound feature amount of the sound section.
辨識解碼器部根據聲頻分析部所產生的聲音特徵量的時間序列資料，並參照聲音辨識資料庫的聲頻示範、字典、及語言示範而決定聲音辨識結果。 The recognition decoder unit determines the sound recognition result based on the time series data of the sound feature amount generated by the audio analysis unit, and refers to the audio demonstration, the dictionary, and the language demonstration of the voice recognition database.
此外，於聲音辨識成功的情況下，聲音辨識結果中包含將說話聲音文字化所得的文字資料。即，聲音辨識伺服器430對通話錄音伺服器410中所記錄的聲音資料進行聲音辨識處理，而產生作為聲音辨識處理的結果的文字資料。 In addition, in the case where the voice recognition is successful, the voice recognition result includes the text data obtained by texting the voice. That is, the voice recognition server 430 performs voice recognition processing on the voice data recorded in the call recording server 410, and generates text data as a result of the voice recognition processing.
另外，聲音辨識伺服器430當自監視裝置440接收到文字資料發送的請求(以下稱為「辨識結果發送請求」)時，將所保存的成為請求對象的聲音資料的聲音辨識結果回復至監視裝置440。於辨識結果發送請求中包含原聲音資料的控制資訊等確定聲音資料的資訊。辨識結果發送請求例如為如下內容，即一指定通話識別資訊並產生所對應的通話的聲音辨識結果，就立即請求逐次回復該聲音辨識結果。 When the voice recognition server 430 receives a request for text data transmission (hereinafter referred to as "identification result transmission request") from the monitoring device 440, the voice recognition server 430 returns the voice recognition result of the stored voice data to the monitoring device. 440. The information of the sound data is determined by the control information including the original sound data in the identification result transmission request. The identification result transmission request is, for example, a message specifying the call identification information and generating a voice recognition result of the corresponding call, and immediately requests the voice recognition result to be successively replied.
監視裝置440是監視客服中心的各通話的管理者使用的個人電腦(personal computer)中作為網頁瀏覽器(web browser)發揮功能的部分。監視裝置440當自管理伺服器420接收到通話開始通知時，對通話錄音伺服器410發送聲音發送請求，並對聲音辨識伺服器430發送辨識結果發送請求。 The monitoring device 440 is a portion that functions as a web browser in a personal computer used by a manager who monitors each call of the customer service center. When receiving the call start notification from the management server 420, the monitoring device 440 transmits a voice transmission request to the call recording server 410, and transmits a recognition result transmission request to the voice recognition server 430.
然後，監視裝置440將自通話錄音伺服器410回復的聲音資料、與自聲音辨識伺服器430回復的聲音辨識結果中的至少文字資料建立對應而顯示於液晶顯示器(display)等顯示部。即， 監視裝置440自通話開始時序後立即將聲音資料與聲音辨識結果(文字資料)建立對應而提示給管理者。 Then, the monitoring device 440 displays the sound data returned from the call recording server 410 in association with at least the text data of the voice recognition result returned from the voice recognition server 430, and displays it on a display unit such as a liquid crystal display. which is, The monitoring device 440 promptly associates the sound data with the sound recognition result (text data) immediately after the start of the call, and presents it to the manager.
此外，雖未圖示，通話錄音伺服器410、管理伺服器420、聲音辨識伺服器430、及監視裝置440各自具有例如中央控制單元(Central Processing Unit，CPU)、儲存有控制程式(program)的唯讀記憶體(Read Only Memory，ROM)等記憶媒體、隨機存取記憶體(Random Access Memory，RAM)等作業用記憶體、及通訊電路等。於該情況下，上述各裝置及各部的功能是藉由CPU執行控制程式而實現。 Further, although not shown, the call recording server 410, the management server 420, the voice recognition server 430, and the monitoring device 440 each have, for example, a central processing unit (CPU) and a control program (program). A memory such as a read only memory (ROM), a memory for a work such as a random access memory (RAM), and a communication circuit. In this case, the functions of each of the above devices and the respective units are realized by the CPU executing the control program.
於此種通訊系統100中，通話錄音伺服器410可將以IP電話網傳輸的通話聲音資料以串流(streaming)提供至聲音辨識伺服器430。而且，聲音辨識伺服器430可將針對聲音資料的聲音辨識結果以串流提供至監視裝置440。 In such a communication system 100, the call recording server 410 can provide the voice data transmitted by the IP telephone network to the voice recognition server 430 in a stream. Moreover, the voice recognition server 430 can provide the sound recognition result for the sound material to the monitoring device 440 in a stream.
即，由於通訊系統100可自通話開始時序後立即開始聲音辨識處理及聲音辨識結果的提示，因此可於通話中幾乎實時地提示聲音辨識結果。 That is, since the communication system 100 can start the voice recognition processing and the voice recognition result prompt immediately after the start of the call sequence, the voice recognition result can be presented in the near real time during the call.
〈通話錄音伺服器的動作〉 <Action of the call recording server>
其次，對通話錄音伺服器的動作進行說明。 Next, the operation of the call recording server will be described.
圖3是表示通話錄音伺服器410的動作的一例的流程圖。 FIG. 3 is a flowchart showing an example of the operation of the call recording server 410.
首先，於步驟S1100中，電話網通訊部411判斷是否自外線網200接收到IP封包。電話網通訊部411於接收到IP封包的情況下(S1100：是)，使處理前進至步驟S1200。另外，電話網通訊部411於未接收到IP封包的情況下(S1100：否)，使處理前進 至下述的步驟S1500。 First, in step S1100, the telephone network communication unit 411 determines whether or not an IP packet has been received from the external network 200. When the telephone network communication unit 411 receives the IP packet (S1100: YES), the process proceeds to step S1200. In addition, when the telephone network communication unit 411 does not receive the IP packet (S1100: NO), the processing advances. Go to step S1500 described below.
於步驟S1200中，聲音記錄控制部414自IP封包提取聲音資料，並將聲音資料與控制資訊等確定聲音資料的資訊建立對應而記錄於記憶體413。另外，通話開始獲取部415自IP封包提取控制資訊。 In step S1200, the sound recording control unit 414 extracts the sound data from the IP packet, and records the sound data in association with the information of the determined sound data such as the control information, and records the data in the memory 413. Further, the call start acquisition unit 415 extracts control information from the IP packet.
然後，於步驟S1300中，通話開始獲取部415根據控制資訊而判斷是否為通話開始時序。通話開始獲取部415於為通話開始時序的情況下(S1300：是)，使處理前進至步驟S1400。另外，通話開始獲取部415於並非為通話開始時序的情況下(S1300：否)，使處理前進至下述的步驟S1500。 Then, in step S1300, the call start acquisition unit 415 determines whether or not the call start sequence is based on the control information. When the call start acquisition unit 415 is in the call start sequence (S1300: YES), the process proceeds to step S1400. When the call start acquisition unit 415 is not the call start sequence (S1300: NO), the process proceeds to step S1500 described below.
於步驟S1400中，聲音辨識控制部416經由管理網通訊部412而將通話開始通知發送至管理伺服器420。 In step S1400, the voice recognition control unit 416 transmits a call start notification to the management server 420 via the management network communication unit 412.
然後，於步驟S1500中，聲音辨識控制部416判斷是否經由管理網通訊部412而接收到聲音發送請求。聲音辨識控制部416於接收到聲音發送請求的情況下(S1500：是)，使處理前進至步驟S1600。另外，聲音辨識控制部416於未接收到聲音發送請求的情況下(S1500：否)，使處理前進至下述的步驟S1700。 Then, in step S1500, the voice recognition control unit 416 determines whether or not the voice transmission request is received via the management network communication unit 412. When the voice recognition control unit 416 receives the voice transmission request (S1500: YES), the voice recognition control unit 416 advances the processing to step S1600. When the voice recognition control unit 416 has not received the voice transmission request (S1500: NO), the voice recognition control unit 416 advances the processing to the following step S1700.
於步驟S1600中，聲音辨識控制部416開始對聲音發送請求的發送源(請求源)傳送聲音資料。 In step S1600, the voice recognition control unit 416 starts transmitting the voice material to the transmission source (request source) of the voice transmission request.
然後，於步驟S1700中，聲音辨識控制部416判斷是否已藉由管理者的操作等而指示結束對通話資料進行監視的處理。聲音辨識控制部416在未指示結束處理的情況下(S1700：否)，使處理返回至步驟S1100。另外，聲音辨識控制部416在已指示結束處理的情況下(S1700：是)，結束一連串的處理。 Then, in step S1700, the voice recognition control unit 416 determines whether or not the process of monitoring the call data has been instructed by the operation of the manager or the like. When the voice recognition control unit 416 has not instructed to end the processing (NO in S1700), the voice recognition control unit 416 returns the processing to step S1100. When the voice recognition control unit 416 has instructed to end the processing (S1700: YES), the voice recognition control unit 416 ends the series of processing.
藉由此種動作，通話錄音伺服器410可獲取以IP電話網傳輸的通話的通話開始時序，並於通話開始時序後立即開始對以IP電話網傳輸的通話的聲音資料進行聲音辨識處理。 By such an action, the call recording server 410 can acquire the call start timing of the call transmitted by the IP telephone network, and immediately start voice recognition processing on the voice data of the call transmitted by the IP telephone network immediately after the call start timing.
〈系統全體的動作〉 <Actions of the whole system>
其次，對通訊系統100全體的動作的流程的一例進行說明。 Next, an example of the flow of the operation of the entire communication system 100 will be described.
圖4是表示通訊系統100的動作的流程的一例的序列圖。 FIG. 4 is a sequence diagram showing an example of the flow of the operation of the communication system 100.
首先，監視裝置440將成為聲音辨識處理的對象的聲音資料的條件發送至管理伺服器420而預先設定聲音資料的條件(S2010)。該條件例如為說話人識別資訊、通話的時間段、通話中包含規定的單詞。即，監視裝置440將聲音辨識處理的對象預先登錄於管理伺服器420。而且，當通話開始時，網路機器320開始對通話錄音伺服器410發送IP封包(S2020)。 First, the monitoring device 440 transmits the condition of the sound data to be subjected to the sound recognition processing to the management server 420 to set the condition of the sound data in advance (S2010). The condition is, for example, speaker identification information, a time period of a call, and a predetermined word included in the call. That is, the monitoring device 440 registers the object of the voice recognition processing in advance in the management server 420. Moreover, when the call starts, the network device 320 starts transmitting an IP packet to the call recording server 410 (S2020).
通訊錄音伺服器410開始分別自發送而來的IP封包提取聲音資料及控制資訊的處理與聲音資料的記錄(S2030)，並且將通話開始通知發送至管理伺服器420(S2040)。於該時間點，於通話錄音伺服器410中至少儲存有通話的最初部分的聲音資料。 The communication recording server 410 starts to extract the processing of the sound data and the control information and the recording of the sound data from the transmitted IP packets (S2030), and transmits the call start notification to the management server 420 (S2040). At this point in time, at least the voice data of the first part of the call is stored in the call recording server 410.
管理伺服器420根據通話開始通知中包含的控制資訊、及於S2010中設定的條件，而判斷是否對聲音資料進行聲音辨識(S2050)。管理伺服器420於判斷為進行聲音辨識的情況下，將辨識開始請求發送至聲音辨識伺服器430(S2060)，並且將通話開始通知發送至監視裝置440(S2070)。聲音辨識伺服器430接收辨識開始請求，而將聲音發送請求發送至通話錄音伺服器410(S2080)。 The management server 420 determines whether or not the voice data is voice-recognized based on the control information included in the call start notification and the condition set in S2010 (S2050). When the management server 420 determines that voice recognition is to be performed, the management start request is transmitted to the voice recognition server 430 (S2060), and the call start notification is transmitted to the monitoring device 440 (S2070). The voice recognition server 430 receives the recognition start request and transmits a voice transmission request to the call recording server 410 (S2080).
如上所述，於通話錄音伺服器410中至少儲存有通話的 最初部分的聲音資料。因此，通話錄音伺服器410接收聲音發送請求，而將儲存的聲音資料回復至聲音辨識伺服器430(S2090)。此外，為了獲得高精度的聲音辨識結果，理想的是發送至聲音辨識伺服器430的聲音資料維持自IP封包提取的聲音資料的品質。 As described above, at least the call recording server 410 stores the call. The initial part of the sound data. Therefore, the call recording server 410 receives the sound transmission request and returns the stored sound data to the voice recognition server 430 (S2090). Further, in order to obtain a highly accurate sound recognition result, it is desirable that the sound data transmitted to the voice recognition server 430 maintains the quality of the sound material extracted from the IP packet.
如此，聲音辨識伺服器430開始對通話錄音伺服器410中儲存的聲音資料進行聲音辨識處理(S2100)。於該時間點，於聲音辨識伺服器430中至少儲存有通話的最初部分的聲音辨識結果。另外，聲音辨識伺服器430將辨識開始通知發送至管理伺服器420(S2110)。 In this manner, the voice recognition server 430 starts voice recognition processing on the voice material stored in the call recording server 410 (S2100). At this point in time, at least the voice recognition result of the initial portion of the call is stored in the voice recognition server 430. In addition, the voice recognition server 430 transmits an identification start notification to the management server 420 (S2110).
藉由進行該辨識開始通知，即便為如網頁瀏覽器般藉由拉拽式(pull type)動作而獲取顯示對象的監視裝置440，亦可實時地獲取並顯示聲音資料及聲音辨識結果。 By performing the identification start notification, even if the monitoring device 440 acquires the display object by a pull type operation like a web browser, the sound data and the sound recognition result can be acquired and displayed in real time.
管理伺服器420將自聲音辨識伺服器430接收到的辨識開始通知傳輸至監視裝置440(S2120)。此外，理想的是於該辨識開始通知、或於步驟S2070中發送的通話開始通知中，作為表示聲音辨識結果的獲取目的地的資訊而包含有聲音辨識伺服器430的識別資訊。監視裝置440接收辨識開始通知，而將辨識結果發送請求發送至聲音辨識伺服器430(S2130) The management server 420 transmits the recognition start notification received from the voice recognition server 430 to the monitoring device 440 (S2120). Further, it is preferable that the identification start notification or the call start notification transmitted in step S2070 includes the identification information of the voice recognition server 430 as the information indicating the acquisition destination of the voice recognition result. The monitoring device 440 receives the recognition start notification and transmits the identification result transmission request to the voice recognition server 430 (S2130)
如上所述，於聲音辨識伺服器430中至少儲存有通話的最初部分的聲音辨識結果。因此，聲音辨識伺服器430接收辨識結果發送請求，而將所儲存的聲音辨識結果發送至監視裝置440(S2140)。 As described above, at least the voice recognition result of the initial portion of the call is stored in the voice recognition server 430. Therefore, the voice recognition server 430 receives the identification result transmission request, and transmits the stored voice recognition result to the monitoring device 440 (S2140).
監視裝置440進而將聲音發送請求發送至通話錄音伺服器410(S2150)，而自通話錄音伺服器410接收聲音資料(S2160)。 此外，理想的是通話錄音伺服器410於聲音辨識控制部416中，將向監視裝置440發送的聲音資料轉換為能夠以網頁瀏覽器輸出的形式的聲音資料。而且，監視裝置440將接收到的聲音資料及聲音辨識結果建立對應而顯示(S2170)。 The monitoring device 440 further transmits a voice transmission request to the call recording server 410 (S2150), and receives the voice data from the call recording server 410 (S2160). Further, it is preferable that the call recording server 410 converts the sound data transmitted to the monitoring device 440 into sound data in a form that can be output by the web browser in the voice recognition control unit 416. Then, the monitoring device 440 displays the received sound data and the sound recognition result in association with each other (S2170).
監視裝置440例如於同時進行多個成為監視對象的通話的情況下，可根據各聲音資料的控制資訊中包含的通話識別資訊或說話人辨識資訊，而針對每一通話獲取關於該些多個通話的聲音辨識結果。於該情況下，理想的是監視裝置440將關於該些多個通話的聲音辨識結果同時顯示於一個網頁瀏覽器畫面上。 For example, when the plurality of calls to be monitored are simultaneously performed, the monitoring device 440 can acquire the plurality of calls for each call based on the call identification information or the speaker identification information included in the control information of each sound data. The sound recognition result. In this case, it is desirable that the monitoring device 440 simultaneously displays the sound recognition results for the plurality of calls on a web browser screen.
藉由此種動作，通訊系統100可將聲音辨識的對象集中於所需要者，並可自通話開始時序後立即開始聲音辨識處理及聲音辨識結果的提示。另外，通訊系統100可於網頁瀏覽器實時地顯示通話的聲音資料及聲音辨識結果。 By such an action, the communication system 100 can concentrate the object of the sound recognition on the person in need, and can start the sound recognition process and the sound recognition result prompt immediately after the start of the call. In addition, the communication system 100 can display the voice data and the voice recognition result of the call in real time in the web browser.
此外，以通訊系統100發送的各種請求，能夠以一次請求來請求對通話全體的資料進行處理，亦能夠以封包、訊框、或批次(batch)的聲音辨識結果等為單位，而請求對通話的一部分資料進行處理。於後者的情況下，例如可將訊框編號或聲音辨識結果的事件(event)編號等用作指定處理對象的識別資訊。 In addition, the various requests sent by the communication system 100 can request the processing of the data of the entire call in one request, and can also request the pair by using the packet, the frame, or the batch sound recognition result as a unit. A part of the call is processed. In the latter case, for example, the frame number or the event number of the sound recognition result or the like can be used as the identification information of the designated processing target.
〈本實施方式的效果〉 <Effect of the present embodiment>
如上所述，根據包含本實施方式的通話資料管理系統的通訊系統100，可自通話開始時序後立即開始對以形成IP電話網的內線網300傳輸的通話的聲音資料進行聲音辨識處理。藉此，通訊系統100可於通話中幾乎實時地提示針對IP電話網的通話的聲音資料的聲音辨識結果。 As described above, according to the communication system 100 including the call material management system of the present embodiment, the voice recognition processing of the voice data of the call transmitted by the internal network 300 forming the IP telephone network can be started immediately after the call start sequence. Thereby, the communication system 100 can present the voice recognition result of the voice data of the call for the IP telephone network in the near real time during the call.
如上所述，於各通話結束後，提取需要確認的通話、且檢索通話的聲音資料並確認所儲存的聲音辨識結果或聲音資料的作業，於IP電話機的個數龐大的情況下變得非常煩雜。 As described above, after the end of each call, the operation of extracting the call to be confirmed, and retrieving the voice data of the call and confirming the stored voice recognition result or voice data becomes very complicated when the number of IP telephones is large. .
就該方面而言，本實施方式的通訊系統100對管理者實時地提示各通話的內容，因此可回避此種煩雜的作業，從而可高效率地監視各通話。因此，根據本實施方式的通訊系統100，可更簡單地實時監視多個IP電話機的通話內容。 In this respect, the communication system 100 of the present embodiment promptly reminds the manager of the content of each call, so that such troublesome work can be avoided, and each call can be monitored efficiently. Therefore, according to the communication system 100 of the present embodiment, it is possible to more easily monitor the call contents of a plurality of IP telephones in real time.
另外，於管理者進行對操作員提出建議(advice)等行動(action)的情況下，若如現有技術般，於通話結束後確認通話內容，則該行動會遲於適當的時序。就該方面而言，本實施方式的通訊系統100可實時地監視各IP電話機的通話內容，因此能夠於恰當的時序進行與通話內容對應的行動。 Further, when the manager performs an action such as an advice to the operator, if the content of the call is confirmed after the call is completed as in the prior art, the action is later than the appropriate timing. In this respect, since the communication system 100 of the present embodiment can monitor the content of the call of each IP telephone in real time, the action corresponding to the content of the call can be performed at an appropriate timing.
另外，本實施方式的通訊系統100於管理伺服器420中，根據控制資訊而針對每一通話來控制通話錄音伺服器410、聲音辨識伺服器430、及監視裝置440各自的動作時序。藉此，即便本實施方式的通訊系統100為通話錄音伺服器410、聲音辨識伺服器430、及監視裝置440獨立的裝置，亦可藉由對該些裝置進行最小限度的改變而使該些裝置聯合動作，從而可獲得上述效果。 Further, in the communication system 100 of the present embodiment, the management server 420 controls the operation timing of each of the call recording server 410, the voice recognition server 430, and the monitoring device 440 for each call based on the control information. Therefore, even if the communication system 100 of the present embodiment is a device independent of the call recording server 410, the voice recognition server 430, and the monitoring device 440, the devices can be minimized by making minimal changes to the devices. The above actions are obtained by combining actions.
另外，於本實施方式的通訊系統100中，監視裝置440自聲音辨識伺服器430獲取聲音辨識伺服器430中所儲存的聲音辨識結果而進行提示。因此，本實施方式的通訊系統100即便於存在多個監視裝置440的情況下，亦可利用各監視裝置440獨立地提示聲音辨識結果。 Further, in the communication system 100 of the present embodiment, the monitoring device 440 acquires the voice recognition result stored in the voice recognition server 430 from the voice recognition server 430 and presents the voice recognition result. Therefore, even in the case where there are a plurality of monitoring devices 440, the communication system 100 of the present embodiment can independently present the sound recognition result by each monitoring device 440.
另外，本實施方式的通訊系統100可動態地選擇成為聲 音辨識的對象的聲音資料(通話、IP電話或說話人等)，因此可更高效率地監視多個IP電話機的通話內容。 In addition, the communication system 100 of the present embodiment can dynamically select a sound. The voice data of the object to be recognized (call, IP phone or speaker, etc.), so that the contents of the call of multiple IP phones can be monitored more efficiently.
另外，本實施方式的通訊系統100可自IP電話網獲取通話的聲音資料，因此可高品質且高效率地獲取各通話的聲音資料。例如，與於各IP電話機設置用以獲取聲音資料的設備的情況相比，本實施方式的通訊系統100可減少所需要的設備成本(cost)及設備空間(space)。另外，由於本實施方式的通訊系統100可獲取送話聲音與受話聲音被完全分離錄音的高品質的聲音資料，因此作為聲音辨識結果而可獲得高精度的文字資料，從而可實現更高的可靠性。 In addition, the communication system 100 of the present embodiment can acquire the voice data of the call from the IP telephone network, so that the voice data of each call can be acquired with high quality and high efficiency. For example, the communication system 100 of the present embodiment can reduce the required equipment cost and equipment space as compared with the case where each IP telephone sets a device for acquiring sound data. In addition, since the communication system 100 of the present embodiment can acquire high-quality sound data in which the transmitted voice and the received voice are completely separated and recorded, high-precision text data can be obtained as a sound recognition result, thereby achieving higher reliability. Sex.
此外，獲取以IP電話網傳輸的通話的聲音資料的方法並不限定於上述例。例如，通話錄音伺服器410於配置於各通話的聲音資料的傳輸路上的情況下，亦可於傳送聲音資料時獲取其複製件。 Further, the method of acquiring the voice data of the call transmitted by the IP telephone network is not limited to the above example. For example, when the call recording server 410 is disposed on the transmission path of the voice data of each call, the call recording server 410 may acquire the copy when the voice material is transmitted.
另外，管理伺服器420的功能、聲音辨識伺服器43及監視裝置440的功能的一部分或全部亦可配置於通話錄音伺服器410。 Further, some or all of the functions of the management server 420, the voice recognition server 43, and the monitoring device 440 may be disposed in the call recording server 410.
例如，通話錄音伺服器410亦可具有處理對象決定部，該處理對象決定部根據所獲取的控制資訊，而決定是否對所記錄的聲音資料進行聲音辨識處理。於該情況下，可利用通話錄音伺服器410進行聲音辨識對象的限定，從而可減少通話開始通知的發送次數。 For example, the call recording server 410 may have a processing target determining unit that determines whether or not to perform voice recognition processing on the recorded sound data based on the acquired control information. In this case, the call recording server 410 can be used to limit the voice recognition target, thereby reducing the number of times the call start notification is transmitted.
另外，本發明的應用並不限定於客服中心。本發明可應用於政府機關或企業中的接待及營業等各種窗口總機、或公司內 內線電話網等可進行多個通話的各種IP電話網。 In addition, the application of the present invention is not limited to the customer service center. The invention can be applied to various window switches or companies in the government office or enterprise, such as reception and business. Various IP telephone networks that can make multiple calls, such as an internal telephone network.
本發明的通話錄音伺服器包括：聲音記錄控制部，依序獲取以IP電話網傳輸的通話的聲音資料並記錄於記憶體中；通話開始獲取部，根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序；及聲音辨識控制部，於所獲取的上述通話開始時序後立即使對所記錄的上述聲音資料的聲音辨識處理開始。 The call recording server of the present invention comprises: a sound recording control unit that sequentially acquires the sound data of the call transmitted by the IP telephone network and records it in the memory; the call start acquisition unit is based on the sound data attached thereto Controlling the information and acquiring a call start sequence for starting the call; and the voice recognition control unit starts the voice recognition process for the recorded voice data immediately after the acquired call start sequence.
此外，上述通話錄音伺服器亦可包括處理對象決定部，該處理對象決定部根據所獲取的上述控制資訊，而決定是否對所記錄的上述聲音資料進行上述聲音辨識處理。 Further, the call recording server may include a processing target determining unit that determines whether or not to perform the voice recognition processing on the recorded voice data based on the acquired control information.
另外，上述通話錄音伺服器亦可包括電話網通訊部，該電話網通訊部自上述IP電話網，接收儲存上述聲音資料且被賦予有包含上述通話的識別資訊的上述控制資訊的封包，且上述通話開始獲取部根據上述控制資訊而確定上述電話網通訊部最初接收到上述通話的上述封包的時序，並獲取所確定的上述時序作為上述通話開始時序。 In addition, the call recording server may further include a telephone network communication unit, and the telephone network communication unit receives, from the IP telephone network, a packet storing the voice data and is provided with the control information including the identification information of the call, and the foregoing The call start acquisition unit determines the timing at which the telephone network communication unit first receives the packet of the call based on the control information, and acquires the determined timing as the call start sequence.
另外，在上述通話錄音伺服器中，上述聲音辨識控制部亦可根據上述控制資訊，而針對每一上述通話來將作為上述聲音辨識處理的結果的文字資料與所記錄的上述聲音資料建立對應而進行管理。 Further, in the call recording server, the voice recognition control unit may associate the character data as a result of the voice recognition processing with the recorded voice data for each of the calls based on the control information. Manage.
本發明的通話資料管理系統亦可為包括：通話錄音伺服器，記錄以IP電話網傳輸的通話的聲音資料；聲音辨識伺服器，對所記錄的上述聲音資料進行聲音辨識處理，而產生作為上述聲音辨識處理的結果的文字資料；及監視裝置，將所記錄的上述聲 音資料與所產生的上述文字資料建立對應而進行提示；且上述通話錄音伺服器包括：聲音記錄控制部，自IP電話網依序獲取上述聲音資料並記錄於記憶體；通話開始獲取部，根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序；及聲音辨識控制部，將所記錄的上述聲音資料輸出至上述聲音辨識伺服器，使上述聲音辨識伺服器於所獲取的上述通話開始時序後立即開始對上述聲音資料進行聲音辨識處理。 The call data management system of the present invention may further include: a call recording server that records voice data of the call transmitted by the IP telephone network; and a voice recognition server that performs voice recognition processing on the recorded voice data, and generates the above a textual data of the result of the voice recognition process; and a monitoring device that will record the above sound The sound data is associated with the generated text data and is presented; and the call recording server includes: a sound recording control unit that sequentially acquires the sound data from the IP telephone network and records the data in the memory; the call start acquisition unit, according to Acquiring the control information of the obtained sound data, and acquiring a call start sequence for starting the call; and the voice recognition control unit outputs the recorded voice data to the voice recognition server to make the voice recognition server The sound recognition processing of the sound data is started immediately after the acquired call start timing.
此外，於上述通話資料管理系統中，上述聲音辨識控制部亦可根據來自上述監視裝置的請求，而將所記錄的上述聲音資料輸出至上述監視裝置，上述聲音辨識伺服器根據來自上述監視裝置的請求，而將所產生的上述文字資料輸出至上述監視裝置，且上述通話資料管理系統包括管理伺服器，該管理伺服器自上述通話錄音伺服器獲取上述通話開始時序，並根據所獲取的上述通話開始時序而控制上述通話錄音伺服器、上述聲音辨識伺服器、及上述監視裝置各自的動作時序。 Further, in the call data management system, the voice recognition control unit may output the recorded voice data to the monitoring device based on a request from the monitoring device, and the voice recognition server may be based on the monitoring device. And outputting the generated text data to the monitoring device, and the call data management system includes a management server, and the management server acquires the call start timing from the call recording server, and according to the acquired call The sequence of operations is started to control the operation timing of each of the call recording server, the voice recognition server, and the monitoring device.
另外，於上述通話資料管理系統中，上述通話錄音伺服器亦可包括電話網通訊部，該電話網通訊部自上述IP電話網接收儲存上述聲音資料且被賦予有包含上述通話的識別資訊的上述控制資訊的封包，且上述管理伺服器根據上述控制資訊，而針對每一上述通話來控制上述通話錄音伺服器、上述聲音辨識伺服器、及上述監視裝置各自的動作時序。 In addition, in the above call data management system, the call recording server may further include a telephone network communication unit, and the telephone network communication unit receives the above-mentioned voice data from the IP telephone network and is given the above-mentioned identification information including the call. Controlling the packet of the information, and the management server controls the operation sequence of each of the call recording server, the voice recognition server, and the monitoring device for each of the calls according to the control information.
另外，於上述通話資料管理系統中，上述管理伺服器亦可根據上述控制資訊，而決定是否對通話記錄於上述通話錄音伺服器中的上述聲音資料進行上述聲音辨識處理。 Further, in the call data management system, the management server may determine whether to perform the voice recognition processing on the voice data recorded in the call recording server based on the control information.
本發明的通話資料管理方法亦可包括：依序獲取以IP電話網傳輸的通話的聲音資料並記錄於記憶體中的步驟；根據隨附於所獲取的上述聲音資料的控制資訊，而獲取開始上述通話的通話開始時序的步驟；及於所獲取的上述通話開始時序後立即使對所記錄的上述聲音資料的聲音辨識處理開始的步驟。 The call data management method of the present invention may further include: sequentially acquiring the sound data of the call transmitted by the IP telephone network and recording the data in the memory; and obtaining the start according to the control information attached to the acquired sound data. a step of a call start sequence of the call; and a step of starting a voice recognition process for the recorded voice material immediately after the acquired call start sequence.
2014年3月17日申請的日本專利特願2014-053355的日本申請案中所包含的說明書、圖式及摘要的揭示內容全部引用於本申請案中。 The disclosure of the specification, drawings and abstract included in the Japanese application of Japanese Patent Application No. 2014-053355, filed on Jan.
[工業上的可利用性] [Industrial availability]
本發明可用作能夠更簡單地監視多個IP電話機的通話內容的通話錄音伺服器、通話資料管理系統、及通話資料管理方法。 The present invention can be used as a call recording server, a call data management system, and a call data management method capable of more easily monitoring the contents of a call of a plurality of IP telephones.
410‧‧‧通話錄音伺服器 410‧‧‧Call Recording Server
411‧‧‧電話網通訊部 411‧‧‧Telephone Network Communications Department
412‧‧‧管理網通訊部 412‧‧‧Management Network Communications Department
413‧‧‧記憶體 413‧‧‧ memory
414‧‧‧聲音記錄控制部 414‧‧‧ Sound Recording Control Department
415‧‧‧通話開始獲取部 415‧‧‧ Call Start Acquisition Department
416‧‧‧聲音辨識控制部 416‧‧‧Sound Identification Control Department
Claims (9)
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2014053355A JP5860085B2 (en) | 2014-03-17 | 2014-03-17 | Call recording server, call data management system, and call data management method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TW201540041A TW201540041A (en) | 2015-10-16 |
| TWI569619B true TWI569619B (en) | 2017-02-01 |
Family
ID=54144173
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW104108398A TWI569619B (en) | 2014-03-17 | 2015-03-17 | Message recording server, message data management system and method thereof |
Country Status (5)
| Country | Link |
|---|---|
| JP (1) | JP5860085B2 (en) |
| KR (2) | KR20160100412A (en) |
| CN (1) | CN106068641B (en) |
| TW (1) | TWI569619B (en) |
| WO (1) | WO2015141189A1 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI672690B (en) * | 2018-03-21 | 2019-09-21 | 塞席爾商元鼎音訊股份有限公司 | Artificial intelligence voice interaction method, computer program product, and near-end electronic device thereof |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6767796B2 (en) | 2016-07-08 | 2020-10-14 | 株式会社日立情報通信エンジニアリング | Call management system and its voice recognition control method |
| JP6875905B2 (en) * | 2017-03-29 | 2021-05-26 | 株式会社日立情報通信エンジニアリング | Call control system and call control method |
| JP6927614B2 (en) * | 2017-04-07 | 2021-09-01 | 株式会社Ｗｉｄｓｌｅｙ | Call system, call method and program |
| JP6749645B2 (en) * | 2017-04-07 | 2020-09-02 | 株式会社Ｗｉｄｓｌｅｙ | Call system, call method and program |
| JP6433003B1 (en) * | 2017-09-04 | 2018-12-05 | Ｎｅｃプラットフォームズ株式会社 | Call recording system, call recording method and call recording program |
| KR101952730B1 (en) * | 2018-06-20 | 2019-05-22 | (주)넥타르소프트 | Radio Communication Systems capable of Voice Recognition with Voting Technology for Communication Contents |
| CN108833727A (en) * | 2018-07-04 | 2018-11-16 | 广东商路信息科技有限公司 | Sound record management method and system based on telephone exchange |
| JP7355996B2 (en) * | 2019-05-17 | 2023-10-04 | アイコム株式会社 | Relay equipment and voice communication recording method |
| CN112822249A (en) * | 2020-12-31 | 2021-05-18 | 盐城数智科技有限公司 | WebRTC-based real-time voice acquisition method |
| KR102241532B1 (en) | 2021-01-15 | 2021-04-16 | (주)두타위즈 | Intelligent callbot server and unmanned counsel systeim using thereof |
| JP7380620B2 (en) * | 2021-03-15 | 2023-11-15 | 株式会社ナカヨ | Call recording system, main device, voice recognition request terminal, program, and telephone call recording method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007212532A (en) * | 2006-02-07 | 2007-08-23 | Nec Corp | Monitoring device, evaluation data selection device, reception person evaluation device, and reception person evaluation system and program |
| CN101247432A (en) * | 2007-07-18 | 2008-08-20 | 北京高信达网络科技有限公司 | VoIP voice data real-time monitoring method and device |
| CN101834931A (en) * | 2010-05-06 | 2010-09-15 | 沈阳帝信通信电子工程有限公司 | IP (Internet Protocol) centralized dictation system and method |
| JP2012120120A (en) * | 2010-12-03 | 2012-06-21 | Toshiba Corp | Call center system and escalation notification method |
| JP2013187733A (en) * | 2012-03-08 | 2013-09-19 | Hitachi Information & Telecommunication Engineering Ltd | Speech communication recording system |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008060785A (en) | 2006-08-30 | 2008-03-13 | Oki Electric Ind Co Ltd | Ip telephone communication recording system |
| JP4773946B2 (en) * | 2006-12-28 | 2011-09-14 | 富士通株式会社 | MONITOR CONTROL SYSTEM, MONITOR DEVICE, MONITOR CONTROL METHOD, AND MONITOR CONTROL PROGRAM |
| JP2008211271A (en) | 2007-02-23 | 2008-09-11 | Nomura Research Institute Ltd | Call center management system, call center managing method, program, and recording medium |
| US8463606B2 (en) * | 2009-07-13 | 2013-06-11 | Genesys Telecommunications Laboratories, Inc. | System for analyzing interactions and reporting analytic results to human-operated and system interfaces in real time |
| JPWO2011074260A1 (en) * | 2009-12-17 | 2013-04-25 | 日本電気株式会社 | Speech recognition system |
| CN102364925A (en) * | 2011-08-31 | 2012-02-29 | 北京天和恒力科技发展有限公司 | Method for automatically recording call content and selectively storing content at mobile phone terminal and system |
| CN103260152B (en) * | 2012-02-20 | 2016-08-03 | 东南大学常州研究院 | The equipment of a kind of mobile terminal sound communication safety supervision and method |
| JP2014042128A (en) | 2012-08-22 | 2014-03-06 | Hitachi Ltd | Speech communication recording system |
-
2014
- 2014-03-17 JP JP2014053355A patent/JP5860085B2/en active Active
-
2015
- 2015-03-11 WO PCT/JP2015/001338 patent/WO2015141189A1/en active Application Filing
- 2015-03-11 CN CN201580012625.2A patent/CN106068641B/en active Active
- 2015-03-11 KR KR1020167022114A patent/KR20160100412A/en active Search and Examination
- 2015-03-11 KR KR1020177024362A patent/KR101826918B1/en active IP Right Grant
- 2015-03-17 TW TW104108398A patent/TWI569619B/en active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007212532A (en) * | 2006-02-07 | 2007-08-23 | Nec Corp | Monitoring device, evaluation data selection device, reception person evaluation device, and reception person evaluation system and program |
| CN101247432A (en) * | 2007-07-18 | 2008-08-20 | 北京高信达网络科技有限公司 | VoIP voice data real-time monitoring method and device |
| CN101834931A (en) * | 2010-05-06 | 2010-09-15 | 沈阳帝信通信电子工程有限公司 | IP (Internet Protocol) centralized dictation system and method |
| JP2012120120A (en) * | 2010-12-03 | 2012-06-21 | Toshiba Corp | Call center system and escalation notification method |
| JP2013187733A (en) * | 2012-03-08 | 2013-09-19 | Hitachi Information & Telecommunication Engineering Ltd | Speech communication recording system |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI672690B (en) * | 2018-03-21 | 2019-09-21 | 塞席爾商元鼎音訊股份有限公司 | Artificial intelligence voice interaction method, computer program product, and near-end electronic device thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20170102394A (en) | 2017-09-08 |
| JP2015177411A (en) | 2015-10-05 |
| JP5860085B2 (en) | 2016-02-16 |
| CN106068641A (en) | 2016-11-02 |
| CN106068641B (en) | 2018-05-29 |
| TW201540041A (en) | 2015-10-16 |
| KR101826918B1 (en) | 2018-02-07 |
| WO2015141189A1 (en) | 2015-09-24 |
| KR20160100412A (en) | 2016-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| TWI569619B (en) | Message recording server, message data management system and method thereof | |
| US10194029B2 (en) | System and methods for analyzing online forum language | |
| US10129394B2 (en) | Telephonic communication routing system based on customer satisfaction | |
| US7136478B1 (en) | Interactive voice response unit response display | |
| EP1889253A1 (en) | A method and system for analyzing separated voice data of a telephonic communication between a customer and a contact center by applying a psychological behavioral model thereto | |
| WO2006124945A1 (en) | A method and system for analyzing separated voice data of a telephonic communication between a customer and a contact center by applying a psychological behavioral model thereto | |
| JP6296821B2 (en) | Business support system | |
| US20080240374A1 (en) | Method and system for linking customer conversation channels | |
| JP2016143909A (en) | Telephone conversation content analysis display device, telephone conversation content analysis display method, and program | |
| JP7274315B2 (en) | Call center system and call monitoring method | |
| JP2019144400A (en) | Controller, control method and computer program | |
| CN113630309B (en) | Robot conversation system, method, device, computer equipment and storage medium | |
| JP6123533B2 (en) | Call center system, call monitoring method and program | |
| JP2005292476A (en) | Client response method and device | |
| JP6859702B2 (en) | Conversation recording device, conversation recording method and conversation recording program | |
| JP2020046634A (en) | Voice recognition system and voice recognition method | |
| JP2017163198A (en) | Voice recognition system, connection device, and voice recognition method |