JP2023161739A - Telephone conversation recording system and telephone conversation recording method - Google Patents

Abstract

To acquire an outside line telephone number and an inside line telephone number of a telephone conversation to be recorded.SOLUTION: A telephone conversation recording system 100 comprises a detection device and a recording device. The detection device includes an outline processing part for generating first feature information by processing sound data on a telephone conversation obtained by collecting a packet of a first session flowing through an outline network 150 to which an outline telephone machine is connected in a predetermined algorithm. The recording device includes a storage part that stores recording data obtained by corresponding a mating station telephone number acquired from a calling control packet of the first session and an inside line telephone number acquired on the basis of the calling control packet of the second session to the sound data on the telephone conversation transmitted in the first session or the second session, in the case where the sound data on the telephone conversation obtained by collecting the packet of the second session flowing through an own station inside network 110 to which an inside telephone machine and a private branch exchange are connected is processed by a predetermined algorism, and an inside processing part generating second feature information and the first feature information and second feature information are matched while satisfying a predetermined condition.SELECTED DRAWING: Figure 1

Description

The present invention relates to a call recording system and a call recording method.

In companies, etc., when multiple telephones are used within the premises, a Private Branch eXchange (PBX) is used to manage calls within the premises, such as connecting extension telephones to each other and connecting extension telephones to outside lines connected to the public telephone network. will be installed.

Also, a call recording device for recording the contents of a call is known. In addition to the contents of the call, the call recording device also records information such as the telephone number of the originator of the call, the telephone number of the destination, the call start time, the call end time, and the like.

In this regard, techniques related to recording of telephone calls are known (for example, Patent Document 1). Furthermore, techniques related to voice recognition processing during a call are also known (for example, Patent Document 2).

Japanese Patent Application Publication No. 2004-328277 Japanese Patent Application Publication No. 2001-094665

However, when a call is made between a telephone connected to an internal network and a telephone connected to an external network, the management of telephone numbers changes when passing through a gateway or private branch exchange. Sometimes it is difficult to obtain both sides of the destination phone number. In order to obtain information on the two telephone numbers of the originator and destination in a call, a call recording device is linked with a private branch exchange to obtain telephone number information from the private branch exchange (so-called CTI cooperation). , private branch exchange). However, there are cases where there is no existing call recording device that can cooperate with a private branch exchange. When creating a new call recording device that can be linked to a private branch exchange used by the company, high costs are unavoidable. Therefore, it is desired to provide a further technique for obtaining the external phone number and extension number of the call to be recorded.

In one aspect, the present invention aims to obtain an external phone number and an extension number of a call to be recorded.

A call recording system according to one aspect of the present invention processes voice data of a call obtained by collecting packets of a first session flowing through an outside line network to which an outside line telephone is connected using a predetermined algorithm. An outside line processing unit that generates information, and a predetermined algorithm processes the voice data of the call obtained by collecting packets of the second session flowing through the extension network of the local office to which the extension telephone and private branch exchange are connected. If the first characteristic information and the second characteristic information meet and meet a predetermined condition, the extension processing unit generates the second characteristic information, and the extension processing unit generates the second characteristic information. Recording data in which the telephone number and the extension number obtained based on the call control packet of the second session are associated with the audio data of the call transmitted in the first session or the second session is stored in the storage unit. and a storage section for storing the information.

You can obtain the external telephone number and extension number of the call to be recorded.

1 is a diagram illustrating a system configuration of a call recording system according to an embodiment. FIG. 1 is a diagram illustrating a block configuration of a recording device according to an embodiment. 1 is a diagram illustrating a block configuration of a detection device according to an embodiment. FIG. 3 is a diagram illustrating an operational flow of recording data storage processing according to the embodiment. FIG. 3 is a diagram illustrating exemplary extension number information. FIG. 3 is a diagram illustrating an operation flow of continuous feature information generation processing according to the embodiment. FIG. 6 is a diagram illustrating generation of continuous feature information according to the embodiment. FIG. 3 is a diagram illustrating call channel information according to the embodiment. FIG. 3 is a diagram illustrating an operational flow of inquiry response processing according to the embodiment. FIG. 6 is a diagram illustrating determination of matching of feature information. FIG. 3 is a diagram illustrating an operational flow of feature information generation processing according to an embodiment. FIG. 3 is a diagram illustrating generation of feature information from audio data. FIG. 3 is a diagram illustrating an operational flow of continuous feature information generation processing according to an embodiment. FIG. 3 is a diagram illustrating chunk hash information according to an embodiment. FIG. 3 is a diagram illustrating a sequence for explaining the flow of recorded data storage processing according to the embodiment. 1 is a diagram illustrating a hardware configuration of a computer for realizing a recording device and a detection device according to an embodiment. FIG. 7 is a diagram showing another example of an outside line network in the call recording system according to the embodiment.

Hereinafter, some embodiments of the present invention will be described in detail with reference to the drawings. Note that the same reference numerals are given to corresponding elements in a plurality of drawings.

As mentioned above, there is a gateway (GW) between the external line network and the internal line network. Furthermore, in the extension network, a private branch exchange intervenes in calls between each extension telephone. Due to the functions of gateways and private branch exchanges, the telephone numbers of calls between external telephones connected to the external network and internal telephones connected to the local office's own internal network are hidden, making it impossible to obtain them. be.

For example, it is possible for the gateway to recognize the telephone number distributed on an external network, both the local station and the other party's station, but in normal cases, the information on the extension telephone that is actually making the call on the internal network, including the extension number, is It cannot be recognized by the gateway. Furthermore, when making a call from an extension telephone, the telephone number of the other party on the outside line is specified, so it is possible to obtain the telephone number of the extension telephone and the telephone number of the other party on the outside line through the extension network. However, when a call is received from an outside line, the gateway or private branch exchange manages the outside line, so the telephone number of the outside line is not distributed over the extension network.

Therefore, as mentioned above, information on extension numbers and external telephone numbers is obtained using technologies such as so-called CTI coordination and private branch exchange coordination, but call recording that can be linked with the private branch exchange being used In some cases, the device does not already exist, and it is costly to create a new call recording device that can be linked with the private branch exchange used by the company.

Therefore, it is desired to provide a technology that enables acquisition of external phone numbers and extension numbers for individual calls without relying on private branch exchanges.

The inventor of the present application has focused on the fact that the voice heard during a call is common between an internal line network and an external line network. Therefore, in the embodiment described below, for example, a device for collecting voice data of telephone calls is installed in each of the internal line network and the external line network. Then, if the voice data of the call obtained from the packets collected by the device installed on the internal line network and the voice data of the call collected from the packets collected by the device installed on the external line network meet the predetermined conditions, , consider them the same call. Next, the extension number obtained based on the packet information collected by the device installed on the internal line network and the external phone number obtained from the packet information collected by the device installed on the external line network are used to identify the originating and incoming call in the call. Identify as the previous two phone numbers. That is, according to the embodiment, it is possible to associate an extension number and an outside telephone number without depending on a private branch exchange. As a result, it becomes possible to store the recorded data of a telephone call in association with the two telephone numbers of the calling party and the receiving party. Hereinafter, embodiments will be described in more detail.

FIG. 1 is a diagram illustrating a system configuration of a call recording system 100 according to an embodiment. As shown in FIG. 1, in the call recording system 100, an internal network 110, which is a local communication network, and an external network 150, which is an external communication network, are connected via a gateway (GW) 101. . The extension network 110 includes a plurality of extension telephones 111 (for example, extension telephones 111a to 111c) and a private branch exchange 120. Private branch exchange 120 may be, for example, an IP-PBX (Internet Protocol Private Branch eXchange). The private branch exchange 120 has, for example, a function of connecting an incoming outside line to the extension telephone 111, and a function of connecting each extension telephone 111 in the premises to an outside line, and controls the incoming and outgoing calls of the plurality of extension telephones 111a to 111c. It's fine.

Further, in the outside line network 150 of FIG. 1, the gateway 101 is connected to an IP (Internet Protocol) telephone network 160 constructed by a telephone company via an optical network unit (ONU) 151. The optical line termination device 151 is, for example, a device that converts an electrical signal and an optical signal.

For example, when an external telephone 170 (hereinafter referred to as an outside line telephone 170) connected to the IP telephone network 160 makes a call to its own representative telephone number set in the private branch exchange 120, Private branch exchange 120 is connected to extension telephone 111.

Further, in FIG. 1, a detection device 180 is connected to an outside line network 150 via a hub (HUB) 152. The detection device 180 monitors packets flowing through the external line network 150, for example. When the detection device 180 detects a packet of a call between the extension telephone 111 and the extension telephone 170 flowing through the outside line network, the detection device 180 collects voice data from the packet and extracts first characteristic information from the voice data according to a predetermined algorithm. generate. Note that SIP messages and RTP packets on the outside line may be encrypted. For example, if the SIP message is encrypted using TLS (Transport Layer Security), etc., or the voice packet is encrypted using sRTP (Secure Real time Transport Protocol), etc., the B2BUA (Back to back User Agent) method is used. A session voter controller (SBC) may be placed in place of hub 152. Then, the SBC may transmit the contents of the decoded SIP message (external telephone number) and the decoded voice packet (RTP) to the detection device.

Further, in FIG. 1, the recording device 130 is connected to the extension network 110 via the hub 122. For example, the recording device 130 monitors packets flowing through the extension network. When the recording device 130 detects a packet of a call between the extension telephone 111 and the outside telephone 170 flowing through the extension network, the recording device 130 collects audio data of the call from the packet, and uses the same predetermined algorithm as that used in the detection device 180. Second feature information is generated from the audio data according to an algorithm. Note that the packet of the call between the extension telephone 111 and the outside telephone 170 includes, for example, a call control packet containing call control information for call control. Therefore, based on information such as the packet type and the destination address included in the call control information, the recording device 130 and the detection device 180 can detect the packet of the call between the extension telephone 111 and the outside telephone 170. can.

The recording device 130 is communicably connected to the detection device 180 via a communication line 155, and transmits an inquiry including the generated second characteristic information to the detection device 180 via the communication line 155. to inquire about outside line information. The outside line information may include, for example, the telephone number of the outside line telephone 170 of the other party in the call.

Upon receiving the second feature information from the recording device 130, the detection device 180 compares the received second feature information with the first feature information generated by the detection device 180. It is assumed that the first feature information and the second feature information match by satisfying a predetermined condition. In this case, each packet that transmits the audio data used to generate the first feature information and the second feature information can be considered to be a packet that transmits the audio of the same call. Therefore, the detection device 180 acquires the telephone number of the outside line telephone 170 from the call control packet of the session transmitting the voice data used to generate the first characteristic information, and extracts the outside line information including the telephone number of the outside line telephone 170. The recording device 130 is notified via the communication line 155.

Upon receiving the outside line information from the detection device 180, the recording device 130 acquires the extension number of the extension telephone 111 based on the call control information of the call control packet of the session that transmits the audio data used to generate the second characteristic information. . Then, the recording device 130 generates recorded data that associates the audio data transmitted in the session, the telephone number of the outside telephone 170 included in the outside line information, and the extension number of the acquired extension telephone 111, and stores it. Store in the device. Therefore, the user who uses the call recording system 100 can know, for example, which telephones the voice of the call in the recorded data was made between. Note that the recorded data may further include other information such as the start time and end time of the call.

The recording data storage process performed by the detection device 180 and the recording device 130 will be described in more detail below. Note that the call control packet may be, for example, a SIP (Session Initiation Protocol) packet, or an H. The packet may be a packet based on the H.323 protocol or the like. In the following, an explanation will be given using a SIP packet as an example.

FIG. 2 is a diagram illustrating a block configuration of the recording device 130 according to the embodiment. The recording device 130 includes, for example, a control section 201, a storage section 202, and a communication section 203. The control unit 201 of the recording device 130 includes, for example, an extension processing unit 211 and a storage unit 212. The storage unit 202 of the recording device 130 may store, for example, extension number information 500, a call extension number, outside line information, and recorded data, which will be described later. The communication unit 203 connects to the extension network according to instructions from the control unit 201, for example, and collects packets flowing through the extension network. Further, the communication unit 203 communicates with the detection device 180, for example, according to instructions from the control unit 201. The details of each of these units and the information stored in the storage unit 202 will be described later.

FIG. 3 is a diagram illustrating a block configuration of the detection device 180 according to the embodiment. The detection device 180 includes, for example, a control section 301, a storage section 302, and a communication section 303. The control unit 301 of the detection device 180 includes, for example, an outside line processing unit 311. The storage unit 302 of the detection device 180 may store, for example, call channel information 800 and chunk hash information 1400, which will be described later. For example, the communication unit 303 connects to an external line network according to instructions from the control unit 301, and collects packets flowing through the external line network. Further, the communication unit 303 communicates with the recording device 130, for example, according to instructions from the control unit 301. The details of each of these units and the information stored in the storage unit 302 will be described later.

Next, a recording data storage process executed by the control unit 201 of the recording device 130 according to the embodiment will be described.

FIG. 4 is a diagram illustrating an operational flow of recording data storage processing executed by the control unit 201 of the recording device 130 according to the embodiment. The control unit 201 of the recording device 130 may start the operation flow shown in FIG. 4 when an instruction to execute the recording process is input.

In S401, the control unit 201 of the recording device 130 captures packets flowing through the extension network via the hub 122.

In S402, the control unit 201 determines whether the captured packet is a SIP packet. Note that the type of the packet can be determined from information such as the packet header, for example. If the captured packet is a SIP packet (YES in S402), the flow advances to S403.

In S403, the control unit 201 determines whether the SIP packet is a session start packet. If the SIP packet is not a session start packet in S403 (NO in S403), the flow advances to S405. On the other hand, if the SIP packet is a session start packet (YES in S403), the flow advances to S404.

In S404, the control unit 201 acquires the extension number based on the address information included in the SIP packet. As an example, the storage unit 202 of the recording device 130 may store extension number information 500 that associates an IP address assigned to a private telephone with an extension number. FIG. 5 is a diagram illustrating example extension number information 500. For example, the control unit 201 may refer to the extension number information 500 to identify the extension number corresponding to the IP address included in the SIP packet. Note that the embodiment is not limited to this. For example, in another embodiment, recording device 130 may obtain extension number information from private branch exchange 120 based on address information in the SIP packet. Then, the control unit 201 stores the extension number in the storage unit 202, and the flow advances to S405.

In S405, the control unit 201 determines whether the SIP packet is a session end packet. If the SIP packet is not a session end packet (NO in S405), the flow returns to S401 and repeats the process for a new packet.

Further, if the captured packet in S402 is not a SIP packet (NO in S402), the flow advances to S407.

In S407, the control unit 201 determines whether the captured packet is an RTP (Real-time Transport Protocol) packet that transmits audio data of a session established using a SIP packet. Note that the above-mentioned session start SIP packet includes address information of an RTP packet to be transmitted within the session using, for example, SDP (Session Description Protocol) protocol. In one example, the control unit 201 may determine, based on the information, whether the captured packet is an RTP packet that transmits audio data of a session established using a SIP packet. If the captured packet is not an RTP packet that transmits audio data for a session established using a SIP packet (S407 is NO), the flow returns to S401 and the process is repeated for a new packet. On the other hand, if the captured packet is an RTP packet that transmits audio data of a session established using a SIP packet (YES in S407), the flow advances to S408. Although an RTP packet is exemplified here as a packet for transmitting audio data, the embodiment is not limited to this, and may be a packet for transmitting other audio data.

In S408, the control unit 201 collects audio data from the captured RTP packets. Then, in S409, the control unit 201 stores the collected audio data in the storage unit 202.

In S410, the control unit 201 determines whether the storage unit 202 has already stored outside line information including the telephone number of the outside line telephone 170 of the partner station corresponding to the session currently being processed. Note that the outside line information is stored in the storage unit 202 in the process of S414, which will be described later. If the outside line information has already been stored in the storage unit 202 (S410: YES), the flow returns to S401. On the other hand, if the outside line information has not been stored in the storage unit 202 (NO in S410), the flow advances to S411.

In S411, the control unit 201 processes the audio data stored in the storage unit 202 in S409 using a predetermined algorithm, and executes feature information generation processing to generate feature information representing the characteristics of the audio data. Note that as the characteristic information representing the characteristics of the audio data, for example, a hash value generated from the audio data or an audio fingerprint (for example, landmark-based fingerprinting) can be used. By using the hash value of audio data for comparison, it is possible to reduce the amount of data and calculation amount for comparing audio. However, in another embodiment, the feature information may be audio data extracted from a part of the audio data stored in the storage unit 202. Further, the feature information may be voiceprint data generated from audio data. Furthermore, voiceprint data may be used in combination with other voice comparisons (for example, comparisons of hash values, etc.). Further examples of generating feature information will be described below with reference to FIGS. 11-14.

In S412, the control unit 201 uses the generated feature information to inquire of the detection device 180 about outside line information. Note that the outside line information may include, for example, the telephone number of the outside line telephone 170 of the partner station, which cannot be obtained inside the gateway 101. Furthermore, the outside line information may include, for example, the telephone number (external line) of the own station that cannot be obtained inside the gateway 101.

Note that the detection device 180, for example, monitors packets in an outside line network, and generates characteristic information representing the characteristics of the voice data from the voice data collected from the packets. Then, when the generated feature information includes feature information that satisfies predetermined conditions and matches the feature information notified in response to the inquiry from the recording device 130, the detection device 180 detects the audio data used to generate the feature information. The telephone number of the outside line telephone 170 of the other station and the telephone number of the own station (external line) are acquired from the call control packet of the session in which the ) may be sent back to the recording device 130. Furthermore, if there is no feature information that satisfies and matches the feature information notified by the inquiry from the recording device 130, the detection device 180 may return an abnormal end. Further details of the processing executed by the detection device 180 will be described later with reference to FIG.

In S413, the control unit 201 determines whether or not outside line information has been received from the detection device 180 as a result of the inquiry. If a notification of abnormal termination is received from the detection device 180 (NO in S413), the flow returns to S401. On the other hand, if outside line information is received from the detection device 180 (S413: YES), the flow advances to S414.

In S414, the control unit 201 stores the outside line information in the storage unit 202, and the flow returns to S401.

If the SIP packet is a session end packet in S405 (YES in S405), the flow advances to S406. In S406, the control unit 201 retrieves the extension number acquired in S404, the telephone number of the outside line telephone 170 of the partner station and the telephone number (external line) of the own station included in the outside line information stored in the storage unit 202 in S414, and the telephone number (external line) of the own station in S409. Recorded data is generated in association with the voice data of the call stored in the storage unit 202, the recorded data is stored in the storage unit 202, and the flow returns to S401.

As described above, by comparing feature information representing the characteristics of voice data, a session in which voice data of the same call is transmitted between an extension network and an external network can be identified, and the extension The number can be associated with the telephone number of the outside telephone 170. As a result, information on the extension number where the call was made and the telephone number of the outside line telephone 170 can be added to the recorded data without depending on the private branch exchange. Furthermore, in the above-described embodiment, since the telephone number of the own station in the outside line network is also obtained, for example, if a plurality of representative numbers are set, it is also possible to specify which representative number was used. However, in the embodiment, the telephone number of the own station in the outside line network may not be acquired.

Next, the processing executed by the detection device 180 will be explained.

FIG. 6 is a diagram illustrating an operational flow of feature information generation processing executed by the control unit 301 of the detection device 180 according to the embodiment.

Note that the processes in S601 to S603, S605, and S607 to S608 in FIG. 6 correspond to the processes in S401 to S403, S405, and S407 to S408 in FIG. Processing similar to may be performed. However, the detection device 180 is installed in the outside line network 150 as shown in FIG. 1, and in S601, the control unit 301 captures a packet flowing through the outside line network. Then, in S604, the control unit 301 stores outside line information (for example, the telephone number of the outside line telephone 170 of the partner station and the telephone number (external line) of the own station) from the call control packet.

Furthermore, in S609, the control unit 301 acquires a plurality of pieces of audio data by shifting the starting position from the audio data collected in S608, and executes continuous feature information generation processing that generates continuous feature information from each piece of audio data. .

FIG. 7 is a diagram illustrating generation of continuous feature information according to the embodiment. FIG. 7(a) shows a stream of RTP packets flowing through a session. Then, the control unit 301 collects audio data from the RTP packets in the process of S608, divides the audio data collected in S609 into chunks of a predetermined length (for example, 10 ms), and divides the audio data into consecutive chunks. (Figure 7(b)). Then, the control unit 301 generates a plurality of pieces of feature information using a predetermined number of chunks (five in FIG. 7) of audio data while shifting the starting positions of the chunks used to generate the feature information. Note that by generating feature information at a plurality of starting positions in this way, for example, it is possible to determine from which position the audio data used to generate the feature information notified in response to an inquiry from the recording device 130 starts. Even if you do not know whether there is a user, you can find the corresponding feature information by searching to see if each of the generated feature information matches the feature information notified in the inquiry. In addition, settings such as the number of feature information generated by the detection device 180 and the range of audio data to be used for generating the feature information can be changed, for example, by determining whether a packet transmitting the same data as a packet detected by the recording device 130 is The time required for detection by the detection device 180 can be determined depending on the configuration and environment of the call recording system 100, etc. Then, in the process of S609, the control unit 301 may register a plurality of pieces of characteristic information generated while shifting the starting positions in the call channel information 800.

FIG. 8 is a diagram illustrating call channel information 800 according to the embodiment. In the call channel information 800, for example, call channels, outside line information, and a plurality of pieces of characteristic information are registered. The call channel is information that identifies the line used for the call. Further, as the outside line information, outside line information (for example, the telephone number of the outside line telephone 170 of the partner station and the telephone number (external line) of the own station) may be registered in the process of S604. A plurality of pieces of feature information may be registered in the feature information according to the order of the RTP packet streams. In the example of FIG. 8, the feature information registered in the upper row is generated from the audio data of the RTP packet located at the beginning, and the feature information registered in the lower row is generated from the audio data of the RTP packet located at the end. It is assumed that it is generated from audio data. The call channel information 800 may be generated for each call session detected in the process of S603.

According to the operation flow shown in FIG. 8 described above, the control unit 301 of the detection device 180 can generate call channel information 800 for each detected session from packets flowing through the outside line network 150.

Next, inquiry response processing executed by the detection device 180 according to the embodiment will be described. FIG. 9 is a diagram illustrating an operational flow of inquiry response processing executed by the control unit 301 of the detection device 180 according to the embodiment. For example, the control unit 301 of the detection device 180 may start the operation flow of FIG. 9 upon receiving the inquiry for outside line information from the recording device 130 through the process of S412 described above.

In S901, the control unit 301 of the detection device 180 selects unselected call channel information 800 from the call channel information 800 generated for the currently ongoing session.

In S902, the control unit 301 selects unselected feature information from the selected call channel information 800.

In S903, the control unit 301 determines whether the feature information included in the inquiry from the recording device 130 and the feature information selected from the call channel information 800 match by satisfying a predetermined condition. The predetermined condition may be set, for example, depending on the value used as the feature information. For example, when using a hash value acquired based on audio data as the feature information, the predetermined condition may be that two pieces of feature information match. Further, for example, when using audio data of a predetermined length cut out from audio data as the feature information, the predetermined condition may be that the two pieces of feature information are similar within a predetermined error range. Further, the predetermined condition may be, for example, that voiceprint data are similar within a predetermined error range by comparison. In this way, in the process of S903, depending on the nature of the feature information to be used, the determination may be made to allow a predetermined error, or it may be determined whether there is a complete match. If the feature information does not match (NO in S903), the flow advances to S905.

In S905, the control unit 301 determines whether or not confirmation of matching with all feature information included in the selected call channel information 800 has been completed. If there is still unselected feature information (NO in S905), the flow returns to S902, selects the unselected feature information, and repeats the process. On the other hand, if the confirmation of matching with all the characteristic information included in the call channel information 800 is completed (S905 is YES), the flow advances to S906.

In S906, the control unit 301 determines whether there is any unselected call channel information 800 among the call channel information 800 generated for the currently ongoing session. If there is still unselected call channel information 800 (NO in S906), the flow returns to S901, selects the unselected call channel information 800, and repeats the process. On the other hand, if all of the call channel information 800 generated for the currently ongoing session has been processed (YES in S906), the flow advances to S907.

In step S907, since the control unit 301 could not identify the call channel information 800 with which the characteristic information matches, it returns abnormal termination to the recording device 130, and this operation flow ends.

Further, if the feature information matches in S903 (YES in S903), the flow advances to S904. In this case, the voice data of the call used to generate the feature information notified in the inquiry and the voice data of the call used to generate the feature information of the call channel information 800 can be considered to be the same call. . Therefore, in S904, the control unit 301 responds with a response that includes the outside line information (for example, the telephone number of the outside line telephone 170 of the other party and the telephone number (external line) of the own station) of the call channel information 800 in which the matching feature information is registered. is transmitted to the recording device 130, and this operation flow ends.

According to the operation flow shown in FIG. 9, the control unit 301 of the detection device 180 transmits outside line information to the recording device 130 by specifying the call channel information 800 that includes feature information that matches the feature information notified in the inquiry. can be provided to Furthermore, when the matching feature information cannot be identified, the control unit 301 of the detection device 180 can notify that matching feature information was not found by returning an abnormal end.

Note that FIG. 10 is a diagram illustrating the determination of whether the feature information matches, which is executed in the operation flow of FIG. 9 above. FIG. 10 shows a stream of RTP packets (RTP1 to RTPn) that transmit a session detected by the control unit 301 of the detection device 180 (FIG. 10(a)). In the embodiment, the control unit 301 divides the audio data transmitted in the stream of RTP packets into chunks of a predetermined length (for example, 10 ms), and creates a plurality of consecutive chunks (FIG. 10( b)).

Then, the control unit 301 generates a plurality of pieces of feature information from the audio data that is a predetermined number of chunks while shifting the starting position (FIG. 10(c)). For example, in the example shown in FIG. 10, feature information is generated from audio data that is a collection of five chunks, and multiple pieces of feature information (for example, feature information 1 to feature information 5) are generated by shifting the starting position one chunk at a time. has been done. Then, the control unit 301 compares the generated plural pieces of feature information with the feature information X received from the recording device 130. Here, it is assumed that any one of the plurality of generated feature information and the feature information X received from the recording device 130 meet a predetermined condition. For example, in the example of FIG. 10, feature information 4 and feature information X match. In this case, the control unit 301 can identify the calls of the voice data for which the feature information 4 and the feature information X have been acquired as the same call.

Note that in FIG. 10C, a plurality of consecutive pieces of feature information are generated while gradually shifting the start position of the audio data used to generate the feature information. Thereby, even if it is not known which position of the audio data collected by the detection device 180 corresponds to the starting position of the audio data used to generate the feature information X in the recording device 130 of FIG. 10(d), By searching for matching positions, it is possible to determine whether the calls are the same or not.

As described above, according to the embodiment, the extension number obtained from the extension network, the telephone number of the external telephone 170 obtained from the external network, and the telephone number of the own station (external line) are included in the voice data of the call. You can save the recorded data by attaching it. Therefore, according to the embodiment, it is possible to identify the originator and destination of the call from the recorded data.

(Example of feature information generation)
Next, an example of generation of feature information according to the embodiment will be shown below. In the example below, the hash value of the chunk obtained by dividing the audio data described above is calculated, and the stream hash obtained by further obtaining hash values from the hash values of the obtained multiple chunks is used as the feature information. use

FIG. 11 is a diagram illustrating an operational flow of feature information generation processing performed by the recording device 130 according to an embodiment. Further, FIG. 12 is a diagram illustrating generation of feature information from audio data. In one embodiment, the control unit 201 of the recording device 130 may execute the operation flow of FIG. 11 in S411, for example.

In S1001, the control unit 201 reads audio data for a predetermined period corresponding to the RTP packet from the audio data stored in the storage unit 202 in the process of S409, and divides the audio data into a plurality of chunks (Fig. 12 (a) and (b)). In one embodiment, the audio data for the predetermined period that is read may include audio data included in the RTP packet received in the most recent execution of the process of S407.

In S1002, the control unit 201 generates hash values for a predetermined number of chunks from among the plurality of chunks obtained by the division. In the example of FIG. 12(c), five hash values are generated from each of five consecutive chunks. Note that hereinafter, a hash value generated from a chunk may be referred to as a chunk hash. Also, generation of hash values can be performed using various hash algorithms such as, for example, FNV-1 and CRC-32.

In S1003, the control unit 201 generates a stream hash using the generated predetermined number of chunk hashes. In the example of FIG. 12(d), the control unit 201 applies a hash algorithm to five consecutive chunk hashes to generate a hash value. Note that the hash value generated from the plurality of consecutive chunk hashes may hereinafter be referred to as a stream hash. Once the stream hash is generated, the operational flow of FIG. 11 ends, and the flow may proceed to S412. Note that in S412, the control unit 201 may inquire about outside line information by using the generated stream hash as feature information.

Next, continuous feature information generation processing by the detection device 180, which corresponds to the feature information generation processing by the recording device 130 in FIG. 11, will be described.

FIG. 13 is a diagram illustrating an operational flow of continuous feature information generation processing according to an embodiment. For example, the control unit 301 of the detection device 180 may start the operation flow of FIG. 13 after proceeding to S609 of FIG.

In S1301, the control unit 301 divides the audio data collected in the process of S608 into multiple chunks. In S1302, the control unit 301 selects one chunk sequentially from the top of the plurality of generated chunks.

In S1303, the control unit 301 generates a chunk hash from the selected chunk. In S1304, the control unit 301 updates the chunk hash information 1400 with the generated chunk hash. Note that in the process of S1303, for example, if the number of chunk hashes registered in the chunk hash information 1400 is less than a predetermined number, the control unit 301 adds the generated chunk to the end of the registered chunk hashes. You can register hashes. On the other hand, if a predetermined number of chunk hashes are already registered in the chunk hash information 1400, the control unit 301 deletes the first chunk hash that was registered in the chunk hash information 1400 the oldest. Then, the control unit 301 may update the chunk hash information 1400 by adding the newly generated chunk hash in S1303 to the end of the chunk hash information 1400.

FIG. 14 is a diagram illustrating chunk hash information 1400 according to one embodiment. In the chunk hash information 1400, a call channel, outside line information, and a predetermined number of chunk hashes are registered. The call channel is information that identifies the line used for the call. In addition, outside line information (for example, the telephone number of the outside line telephone 170 of the partner station and the telephone number (external line) of the own station) may be registered in the outside line information in the process of S604. A predetermined number of chunk hashes may be registered in the chunk hash according to the order of the RTP packet streams that transmit the audio data used to generate the chunk hashes. In the example of FIG. 14, the chunk hash registered in the upper row is generated from the audio data of the RTP packet located at the beginning, and the chunk hash registered in the lower row is generated from the audio data of the RTP packet located at the end. It is assumed that it is generated from audio data. The chunk hash information 1400 may be generated for each call session detected in S603.

In the process of S1305, the control unit 301 determines, for example, whether the number of chunk hashes registered in the chunk hash information 1400 has reached a predetermined number. Note that in one embodiment, the predetermined number may be set to the number of chunk hashes used to generate the stream hash. If the number of chunk hashes registered in the chunk hash information 1400 is less than the predetermined number (NO in S1305), the flow advances to S1308. On the other hand, if a predetermined number of chunk hashes are registered in the chunk hash information 1400 (YES in S1305), the flow advances to S1306.

In S1306, the control unit 301 arranges the chunk hashes registered in the chunk hash information 1400 in order from the first chunk hash and applies a hash algorithm to generate a stream hash.

In S1307, the control unit 301 registers the generated stream hash as feature information in the call channel information 800, and updates the call channel information 800. Note that the generated stream hashes may be sequentially added to the call channel information 800. Alternatively, as described in the example of chunk hash information 1400 in FIG. 14, updating may be performed such that a predetermined number of stream hashes are registered in call channel information 800 in order from the newly generated stream hash.

In S1308, the control unit 301 determines whether all the chunks generated by the division in S1301 have been processed. If there is an unprocessed chunk (NO in S1308), the flow returns to S1302. On the other hand, if the processing of all chunks has been completed (YES in S1308), this operational flow may end and the flow may return to S601.

With the above-described operational flow of FIG. 13, it is possible to store the stream hash as feature information in the call channel information 800. Note that the stream hash described with reference to FIGS. 11 to 14 is an example of feature information, and the feature information according to the embodiment is not limited to this. For example, an audio fingerprint can be used as the feature information, or a predetermined number of chunk hashes generated in the processes of FIGS. 11 to 14 may be arranged and used as the feature information.

FIG. 15 is a diagram illustrating a sequence for explaining the flow of recorded data storage processing according to the embodiment. As shown in FIG. 15, when the call starts, the recording device 130 starts recording the voice of the call, and also acquires the extension number based on the packet of the call (1501a in FIG. 15). Note that since the recording device 130 is connected to the extension network, the telephone number of the other party's office and the telephone number of the own office's external line, which are distributed over the external line network, are not acquired.

Subsequently, the recording device 130 generates characteristic information from the audio of the call (1502a in FIG. 15). The recording device 130 transmits an inquiry including the generated feature information to the detection device 180 (1503a in FIG. 15).

On the other hand, when the call starts, the detection device 180 also acquires the telephone number of the other party and the outside line telephone number of the own station from the packet of the call (1501b in FIG. 15). Note that since the detection device 180 is connected to the external line network, the extension number distributed on the internal line network is not acquired.

Subsequently, the detection device 180 generates a plurality of consecutive pieces of feature information using a plurality of pieces of audio data having different starting positions obtained from the speech of the call (1502b in FIG. 15).

When receiving an inquiry from the recording device 130, the detection device 180 searches whether the characteristic information of the inquiry matches any of the plurality of characteristic information obtained from the voice data during the call. For example, in FIG. 15, the feature information obtained from call A does not match the feature information of the inquiry (1503b in FIG. 15). However, the characteristic information of the inquiry matches the characteristic information obtained from call B (1504b in FIG. 15).

Therefore, the detection device 180 notifies the recording device 130 of the outside line information (for example, the telephone number of the partner station and the outside line telephone number of the own station) of the call B (1505b in FIG. 15). Therefore, the recording device 130 can acquire the outside line information of the inquired call (for example, the telephone number of the partner station and the outside line telephone number of the own station).

Subsequently, when the call ends, the recording device 130 adds the extension number of the call, the telephone number of the partner station included in the outside line information received from the detection device 180, and the outside line telephone number of the own station to the recording file in which the call was recorded. The recorded data associated with the numbers is output (1504a in FIG. 15). For example, the recording device 130 may store recorded data in the storage unit 202. This makes it possible to specify information on the originator and destination of the call, as well as information on the representative number of the own station used for the call, from the recorded data stored in the storage unit 202.

Although the embodiments have been illustrated above, the embodiments are not limited thereto. For example, the above-described operational flow is an example, and the embodiments are not limited thereto. If possible, the operational flow may be executed by changing the order of processing, may include additional processing, or may omit some processing.

Further, in the above-described embodiment, an IP-PBX is shown as an example of the private branch exchange 120, and an example is shown in which connection with an external telephone is via the IP telephone network 160. However, embodiments are not limited thereto. For example, even when the external line network is an ISDN (Integrated Services Digital Network), telephone number information can be obtained from ISDN call data based on protocol regulations. Furthermore, for example, the above-described embodiments can be applied even to a phone call via an analog line by performing analog-to-digital conversion and processing it as a digital signal.

Furthermore, in the above-described embodiment, a case is exemplified in which the recording device 130 is placed in an extension network, and the detection device 180 is placed in an outside line network. In this case, the recording device 130 can record all extension calls between the extension telephones 111 and calls with outside parties. However, embodiments are not limited thereto. For example, in another embodiment, the recording device 130 may be located in the external network and the detection device 180 may be located in the internal network. In this case, the detection device 180 may acquire the extension number and notify it to the recording device 130, so that the recording device 130 can match the received extension number to the acquired partner station number and own station number (external line). You can save recorded data with Furthermore, for example, by connecting the recording device 130 to an outside line network, it is possible to efficiently record only telephone calls connected to outside lines.

Further, in the above-described embodiment, an example is shown in which the recording device 130 and the detection device 180 are directly connected through the communication line 155, and an example is shown in which characteristic information and outside line information are exchanged via the communication line 155. ing. However, embodiments are not limited thereto. For example, in another embodiment, feature information, outside line information, etc. may be included in the audio data of the RTP packet and notified using audio watermarking technology or the like.

Moreover, although the detection device 180 is shown as one device in the above-mentioned embodiment, the embodiment is not limited to this. For example, if the recording device 130 is installed in an extension network, the detection device 180 may be incorporated into the gateway 101. Furthermore, for example, when the recording device 130 is installed in an outside line network, the detection device 180 may be incorporated into the private branch exchange 120.

For example, as a method for acquiring an extension number that is distributed over the extension network 110 across the gateway 101 and the private branch exchange 120, or a telephone number of a partner station that is distributed over the external network 150, it is possible to transmit the extension number to the extension telephone 111. It is also possible to connect an adapter that transmits voice signals and notify the extension number by voice. However, in this case, it is necessary to connect an adapter to each of the extension telephones 111, and to configure the adapter to transmit audio signals for individual extension numbers. However, according to the above-described embodiment, there is no need to separately install an adapter or the like in the extension telephone. For example, if you monitor the packets flowing through the extension network and the packets flowing through the external network, you can check the extension numbers circulating on the extension network without having to configure the adapter to send audio signals for individual extension numbers. You can also obtain the telephone number of the other party's station on the outside line network.

Also, as described above, for example, SIP messages and RTP packets on the outside line may be encrypted. For example, if the SIP message is encrypted using TLS (Transport Layer Security), etc., or the voice packet is encrypted using sRTP (Secure Real time Transport Protocol), etc., the B2BUA (Back to back User Agent) method is used. A session voter controller (SBC) may be placed in place of hub 152.

FIG. 17 is a diagram showing another example of the outside line network 150 in the call recording system 100 of FIG. 1. In FIG. 17, an SBC 1701 is placed in place of the hub 152. The SBC 1701 is, for example, a B2BUA session voter controller, and can decrypt encrypted SIP messages (TLS, etc.) and voice packets (sRTP, etc.). Then, the SBC transmits the contents of the decoded SIP message (external telephone number) and the decoded voice packet (RTP) to the detection device. For example, when a call is made from an external telephone to the private branch exchange 120 such as an IP-PBX, the SBC 1701 operates as a user agent server, and the user agent 1 Terminates RTP and passes it to User Agent 2. That is, the SBC 1701 simply pretends to be an IP-PBX (GW in this configuration). Furthermore, when a call is made from the private branch exchange 120 side such as an IP-PBX, the user agent 2 similarly terminates the call and hands it over to the user agent 1.

If communication with an external telephone is encrypted, the user agent 1 of the SBC 1701 performs encryption and decryption. Therefore, SBC 1701 allows unencrypted SIP messages and voice packets to be distributed to detection devices.

Note that in the above-described embodiment, the control unit 201 of the recording device 130 operates as the extension processing unit 211, for example, in S411, the process of FIG. 11, and 1502a. Further, for example, in the process of S406, the control unit 201 of the recording device 130 operates as the storage unit 212. For example, in S609, the process of FIG. 13, and 1502b, the control unit 301 of the detection device 180 operates as the outside line processing unit 311.

FIG. 16 is a diagram illustrating the hardware configuration of a computer 1600 for realizing the recording device 130 and the detection device 180 according to the embodiment. A hardware configuration for realizing the recording device 130 and the detection device 180 in FIG. 16 includes, for example, a processor 1601, a memory 1602, a storage device 1603, a communication interface 1606 (for example, 1606a and 1606b), and an input/output interface 1607. . Note that the processor 1601 is connected to a memory 1602, a storage device 1603, a communication interface 1606, and an input/output interface 1607 via, for example, a bus 1610.

The processor 1601 may be, for example, a single processor, a multiprocessor, or a multicore. The processor 1601 provides some or all of the functions of each unit described above by using the memory 1602 to execute, for example, a program that describes the procedure of the operation flow described above. For example, the processor 1601 of the recording device 130 operates as the control unit 201, the extension processing unit 211, and the storage unit 212 by reading and executing a program stored in the storage device 1603. Further, for example, the processor 1601 of the detection device 180 operates as the control unit 301 and the outside line processing unit 311 by reading and executing a program stored in the storage device 1603.

Memory 1602 is, for example, a semiconductor memory and may include a RAM area and a ROM area. The storage device 1603 is, for example, a hard disk, a semiconductor memory such as a flash memory, or an external storage device. Note that RAM is an abbreviation for Random Access Memory. Further, ROM is an abbreviation for Read Only Memory.

The storage unit 202 and storage unit 302 described above include, for example, a memory 1602 and a storage device 1603. For example, the storage device 1603 of the recording device 130 may store information such as extension number information 500, a call extension number, outside line information, and recorded data. Further, for example, the storage device 1603 of the detection device 180 may store information such as call channel information 800 and chunk hash information 1400. Note that the recording device 130 preferably includes a storage device 1603 in order to accumulate recorded data of telephone calls. On the other hand, detection device 180 may not store call recordings and, in one embodiment, may not include storage device 1603.

Communication interface 1606 transmits and receives data to and from other devices according to instructions from processor 1601. Note that the communication interface 1606a may be an interface for connecting to the hub 122 or the hub 152, for example. Additionally, the communication interface 1606b may be used for communication between the detection device 180 and the recording device 130. Note that in one embodiment, the communication interface 1606b may not be included. In that case as well, the feature information and outside line information can be notified by, for example, generating DTMF (Dual-Tone Multi-Frequency) representing the feature information, outside line information, etc., and including it in the audio data of the RTP packet.

Input/output interface 1607 may be, for example, an interface between an input device and an output device. The input device is, for example, a device such as a keyboard or mouse that receives instructions from a user. The output device is, for example, a display device such as a display, and an audio device such as a speaker.

Each program according to the embodiment is provided to the recording device 130 and the detection device 180 in the following format, for example.
(1) Installed in the storage device 1603 in advance.
(2) Provided by a server such as a program server.

Note that the hardware configuration of the computer 1600 for realizing the recording device 130 and the detection device 180 described with reference to FIG. 16 is an example, and the embodiment is not limited thereto. For example, some of the functional units described above may be replaced or deleted, or new components may be added. In another embodiment, for example, some or all of the functions of the above-described units may be implemented as hardware such as FPGA and SoC. Note that FPGA is an abbreviation for Field Programmable Gate Array. SoC is an abbreviation for System-on-a-chip.

Above, several embodiments are described. However, the embodiments are not limited to the embodiments described above, but should be understood to include various modifications and alternative forms of the embodiments described above. For example, it will be understood that the various embodiments can be embodied by changing the components without departing from the spirit and scope thereof. Furthermore, it will be understood that various embodiments can be implemented by appropriately combining the plurality of components disclosed in the embodiments described above. Furthermore, various embodiments can be implemented by deleting or replacing some components from all the components shown in the embodiments, or adding some components to the components shown in the embodiments. Those skilled in the art will understand that this can be done.

100: Call recording system 101: Gateway 110: Extension network 111: Extension telephone 120: Private branch exchange 122: Hub 130: Recording device 150: External line network 151: Optical line termination device 152: Hub 155: Communication line 160: IP telephone network 170 : External line telephone 180 : Detection device 201 : Control unit 202 : Storage unit 203 : Communication unit 211 : Extension processing unit 212 : Storage unit 301 : Control unit 302 : Storage unit 303 : Communication unit 311 : External line processing unit 1600 : Computer 1601 : Processor 1602: Memory 1603: Storage device 1606: Communication interface 1607: Input/output interface 1610: Bus

Claims (5)

an outside line processing unit that processes voice data of a call obtained by collecting packets of a first session flowing through an outside line network to which the outside line telephone is connected using a predetermined algorithm to generate first characteristic information;
an extension that generates second feature information by processing voice data of a call obtained by collecting packets of a second session flowing through the extension network of the local office to which the extension telephone and the private branch exchange are connected, using the predetermined algorithm; a processing section;
If the first feature information and the second feature information match by satisfying a predetermined condition, the telephone number of the partner station acquired from the call control packet of the first session and the second session a storage unit that stores in a storage unit recorded data in which the extension number obtained based on the call control packet is associated with voice data of the call transmitted in the first session or the second session;
Call recording system, including: 2. The call recording system according to claim 1, wherein the recorded data is further associated with a telephone number of the own station acquired from a call control packet of the first session. The outside line processing unit extracts a plurality of voice data having different starting positions from the voice data of the call obtained by collecting packets of the first session, and extracts the first characteristic from each of the plurality of voice data. generate information,
The storage unit stores the recorded data in the storage unit when any of the first feature information generated from each of the plurality of audio data and the second feature information meet a predetermined condition. 3. The call recording system according to claim 1 or 2, wherein the call recording system stores the call recording system. The extension processing unit extracts a plurality of voice data having different starting positions from the voice data of the call obtained by collecting packets of the second session, and extracts the second characteristic from each of the plurality of voice data. generate information,
The storage unit stores the recorded data in the storage unit when any of the second feature information generated from each of the plurality of audio data and the first feature information meet a predetermined condition. 3. The call recording system according to claim 1 or 2, wherein the call recording system stores the call recording system. generating first feature information by processing voice data of a call obtained by collecting packets of a first session flowing through an outside line network to which the outside line telephone is connected using a predetermined algorithm;
generating second feature information by processing voice data of a call obtained by collecting packets of a second session flowing through the extension network of the local office to which the extension telephone and the private branch exchange are connected, using the predetermined algorithm;
If the first feature information and the second feature information match by satisfying a predetermined condition, the telephone number of the partner station acquired from the call control packet of the first session and the second session storing in a storage device recorded data in which the extension number obtained based on the call control packet is associated with voice data of the call transmitted in the first session or the second session;
How to record calls, including:

Images