Mail text-to-speech generation program

The email reading text generation program addresses the issue of redundant vocalization in email bodies by deleting unnecessary lines and extracting relevant text, resulting in a more efficient and concise voice output.

JP2025091812APending Publication Date: 2025-06-19PATLITE CORP
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023207280
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-12-07
Publication Date
2025-06-19

AI Technical Summary

Technical Problem

Existing email reading technologies vocalize redundant parts of email bodies, leading to increased processing load and time for speech synthesis, making them unsuitable for conveying key points quickly.

Method used

An email reading text generation program that deletes specific lines from email bodies, such as those containing URLs, email addresses, and redundant symbols, and extracts relevant lines with punctuation marks or polite language, to generate a concise reading text.

Benefits of technology

The program effectively reduces the amount of text to be vocalized, alleviating the processing load and time for speech synthesis, while ensuring key information is conveyed efficiently.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2025091812000001_ABST
    Figure 2025091812000001_ABST
Patent Text Reader

Abstract

To provide an email text-to-speech generation program capable of appropriately processing the text of an email message to generate a text-to-speech.SOLUTION: A computer for executing an email text-to-speech generation program executes a step (S14) of deleting lines containing URL descriptions from a text included in the body of an email, and a step (S13) of deleting lines immediately before the lines containing URL descriptions. The computer may perform a step (S21) of deleting lines whose first characters are predetermined symbols from the text contained in the body of the email. The computer may perform a step (S30) of extracting lines whose last characters are punctuation marks from the text contained in the body of the email and adding them to read-aloud texts. The computer may perform a step (S32) of extracting texts of sentences written in honorific style to add them to the read-aloud texts.SELECTED DRAWING: Figure 7
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to an e-mail reading text generation program for generating, by a computer, reading text used to read out text included in the body of an e-mail. The present invention also relates to a computer that executes the e-mail reading text generation program, an e-mail receiving / reading device including such a computer, and a notification device including the e-mail reading / receiving device.

Background Art

[0002] Patent Document 1 discloses an e-mail device that reads out text data of a received e-mail in voice. When a predetermined pattern exists in the received text data, this device stops the voice output of the portion of the pattern. Specifically, the citation marks (such as ">") that are conventional symbols of e-mails and represent the citation of another person's sentence in the body, the face marks that express emotions, and the repeated notation of arithmetic symbols for representing the separation of document contents are set as registered patterns. When the pattern at the beginning of a line matches the registered pattern, the matching portion is deleted from the voice output target. In this way, it attempts to avoid the vocalization of symbol portions that would make the meaning unclear and difficult for the user to understand when output in voice, and output only meaningful text data in voice.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, since the prior art of Patent Document 1 simply deletes the symbol part that does not make sense even when vocalized from the vocalization target, all parts that make sense when vocalized are targets for vocalization. Therefore, redundant parts in the body of the email also become targets for vocalization. Thus, it is not necessarily suitable for applications that convey key points to the user in a short time. Also, since redundant parts are also targets for vocalization, the load on the speech synthesis process is large and the process takes time.

[0005] Therefore, one embodiment of the present invention provides an email reading text generation program that can appropriately process the text of the body of an email to generate a reading text.

[0006] Also, one embodiment of the present invention provides a computer that executes the above-described email reading text generation program, an email receiving / reading device that includes such a computer, and a notification device that includes the email reading / receiving device.

Means for Solving the Problem

[0007] To solve the above problems, the present invention provides the following exemplary embodiments.

[0008] 1. A program for generating, by a computer, a reading text used for reading the text included in the body of an email, the program causing the computer to execute steps of deleting a line including a description of a URL (Uniform Resource Locator) (preferably a line terminated with a description of a URL) from the text included in the body of the email, and deleting a line immediately preceding the line including the description of the URL. A mail reading text generation program.

[0009] 2. The mail reading text generation program according to item 1, further causing the computer to execute a step of deleting a line whose line start is a predetermined symbol (for example, an annotation symbol, a quotation mark, etc.) from the text included in the body of the email.

[0010] 3. The mail reading text generation program according to claim 1 or 2, further causing the computer to extract lines whose ends are punctuation marks from the text included in the body of the email and add them to the reading text.

[0011] 4. The mail reading text generation program according to any one of claims 1 to 3, further causing the computer to extract the text of sentences described in polite form and add them to the reading text.

[0012] 5. The mail reading text generation program according to any one of claims 1 to 4, further causing the computer to delete lines including descriptions of email addresses (preferably lines ending with descriptions of email addresses) from the text included in the body of the email, and to delete the lines immediately preceding the lines including descriptions of email addresses.

[0013] 6. The mail reading text generation program according to any one of claims 1 to 5, further causing the computer to delete lines including strings in which a predetermined symbol continues for a predetermined number or more times from the text included in the body of the email, and to delete the subsequent lines.

[0014] 7. The mail reading text generation program according to any one of claims 1 to 6, further causing the computer to extract lines or sentences including strings subjected to a predetermined emphasis process and add them to the reading text.

[0015] 8. The mail reading text generation program according to any one of claims 1 to 7, further causing the computer to extract lines or sentences including a predetermined keyword and add them to the reading text.

[0016] 9. A computer that executes an email program for receiving emails, the email reading text generation program according to any one of items 1 to 8, and a voice synthesis program for synthesizing voice data for vocalizing the reading text generated by the email reading text generation program, the computer comprising: an email receiving step of executing the email program to receive an email; a reading text generation step of executing the email reading text generation program to generate a reading text used for reading the text included in the body of the email received in the email receiving step; and a voice synthesis step of executing the voice synthesis program to synthesize voice data of the reading text generated in the reading text generation step.

[0017] 10. An email receiving / reading device including the computer according to item 9 and a pronunciation means for acoustically outputting the voice data generated in the voice synthesis step.

[0018] 11. An informing device including the email receiving / reading device according to item 10 and an informing device controlled by the computer to generate a visual notification signal.

Brief Description of the Drawings

[0019]

Figure 1

Figures 2A - 2B

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Embodiments for Carrying Out the Invention

[0020] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0021] FIG. 1 shows an example of a system configuration including a notification device 1 according to an embodiment of the present invention. In this embodiment, the notification device 1 is configured to receive an electronic mail (hereinafter simply referred to as "mail") from the network 100 and read aloud the mail. The notification device 1 is connected to the network 100. The network 100 typically includes a wide area network (WAN) 101 such as the Internet or an intranet. The network 100 may further include a local area network (LAN) 102. In one specific example, the notification device 1 is connected to the local area network 102, and the local area network 102 is connected to the wide area network 101. The connection between the notification device 1 and the local area network 102 may be a wired connection or a wireless connection. Further, the network 100 may include a wireless communication network (not shown) such as a mobile communication network, and the wireless communication network may be connected to the wide area network 101. Therefore, the notification device 1 may be connected to the wide area network 101 via the wireless communication network.

[0022] In the network 100 (typically a wide area network 101), a mail server 103 (hereinafter referred to as the "sender mail server 103") used by a mail sender and a mail server 104 (hereinafter referred to as the "recipient mail server 104") used by a mail recipient are connected. Here, the recipient mail server 104 is a mail server in which a mail account to be received in the notification device 1 is set and a mailbox of the mail account is provided. A mail sent by a mail sender to the mail account of a mail recipient as the destination is stored in the mailbox of the recipient mail server 104 via the sender mail server 103 and the network 100. The notification device 1 periodically accesses the mailbox of the recipient mail server 104 to obtain mails addressed to the mail account to be received.

[0023] FIGS. 2A and 2B are a front view and a rear view for explaining an external configuration example of the notification device 1. In this example, the notification device 1 has a form of a signal lamp connectable to the network 100. The notification device 1 includes a base 11 and a body 12 provided on the upper part of the base 11. The body 12 includes a columnar (cylindrical in the illustrated example) notification device 13 extending upward from the base 11. The notification device 13 includes a plurality (three in the illustrated example) of light emitting parts 13a, 13b, 13c laminated in the vertical direction in this example. The light emitting parts 13a, 13b, 13c may be configured to generate lights of different colors (for example, red, yellow, green). The light emitting parts 13a, 13b, 13c may be, for example, LED units using LEDs as light sources. The body may also include, for example, a wireless communication unit 18 connectable to the network 100 in addition to the light emitting parts 13a, 13b, 13c (notification device 13).

[0024] On the base 11, a plurality of operation buttons 14 are arranged on the front, and a speaker 15 (built-in speaker) is arranged so as to be exposed on the outer surface. Further, on the base 11, connectors for a communication interface 16 and an external interface 17 are arranged on the back. The communication interface 16 provides an interface for communication with the network 100. The external interface 17 provides an interface for connection with an external device 107 (see FIG. 1) such as a production machine.

[0025] FIG. 3 is a block diagram for explaining an electrical configuration example of the notification device 1. The notification device 1 includes a computer 20. The computer 20 includes a control unit 21 (processing device) and a storage unit 22 (storage). The control unit 21 and the storage unit 22 can exchange data through data communication via a bus 23. Connected to the bus 23 are a communication interface (I / F) 16 and an external interface (I / F) 17.

[0026] The communication interface 16 is typically a wired LAN module, a wireless LAN module, etc., and is an interface for performing wired or wireless communication. The external interface 17 is an interface for connection with the external device 107 (see FIG. 1) and is appropriately configured according to the external device 107 to be connected. The external device 107 is, for example, a production machine, a printer, a server, a network camera, a disaster prevention system, a security system, an entrance and exit system, an environmental monitoring system, an order receiving and issuing system, etc. A light emission control signal or an audio control signal can be input from the external device 107 to the external interface 17. The light emission control signal can be used for controlling the light emitting units 13a, 13b, 13c, and the audio control signal can be used for audio output from the speaker 15. The audio control signal may be in the form of text data, and the audio of the text data may be synthesized and pronounced from the speaker 15.

[0027] The bus 23 is further connected with a drive unit 24 for driving the speaker 15. The drive unit 24 includes, for example, a digital / analog converter that converts audio data into an analog audio signal, and an amplifier that amplifies the analog audio signal to drive the speaker 15. Thereby, the drive unit 24 drives the speaker 15 based on the audio data provided from the control unit 21 to acoustically reproduce the audio data. That is, the drive unit 24 and the speaker 15 constitute a sounding means. The analog audio signal generated by the drive unit 24 is also led out to the line-out terminal 25. An external device such as an external speaker can be connected to the line-out terminal 25.

[0028] The bus 23 is further connected with a light-emitting drive unit 26 for driving the light-emitting parts 13a, 13b, 13c. The light-emitting drive unit 26 drives the light-emitting parts 13a, 13b, 13c based on the light-emitting command data provided from the control unit 21. Thereby, an optical signal, that is, a visual notification signal can be generated.

[0029] The control unit 21 includes a processor (CPU) 27 and a memory 28. The memory 28 is a main storage device and includes a RAM (Random Access Memory) 29 and a ROM (Read Only Memory) 30. The storage unit 22 is an auxiliary storage device such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). Various programs 40 executed by the control unit 21 and various data 50 are stored in the storage unit 22.

[0030] In this example, the program 40 includes a mail program 41, a mail filtering program 42, a mail reading text generation program 43, a voice synthesis program 44, a text reading program 45, etc. The data 50 includes data of the received mail 51, text data of the mail body text 52 extracted from the mail 51, text data of the reading text 53 generated by processing the mail body text 52, and voice data 54 corresponding to the text data.

[0031] The control unit 21 can execute the program 40 stored in the storage unit 22 by the functions of the processor 27 and the memory 28. When the mail program 41 is executed, the computer 20 functions as a mailer. Specifically, when the mail program 41 is executed, the computer 20 performs a receiving operation of periodically accessing the recipient mail server 104 to obtain new mails. The received mails are stored in the storage unit 22. When the mail filtering program 42 is executed, the computer 20 performs filtering to extract the mails to be read aloud from the received mails according to the preset filtering rules. When the mail read-aloud text generation program 43 is executed, the computer 20 extracts the text of the body of the mail to be read aloud and generates the text for reading aloud (read-aloud text) from the text. The generated read-aloud text is stored in the storage unit 22. When the voice synthesis program 44 is executed, the computer 20 synthesizes the read-aloud text into voice data. The voice data is stored in the storage unit 22. The voice synthesis program 44 may incorporate a voice synthesis engine or may use a voice synthesis engine (cloud-based voice synthesis service) arranged on the network 100. When the text reading program 45 is executed, the computer 20 supplies the voice data to the drive unit 24 to make the voice sound.

[0032] Needless to say, the processing of the computer 20 on the text is executed on the data representing the text, that is, the text data. In this specification, for the sake of brevity, the text data to be processed by the computer 20 may be simply expressed as "text".

[0033] The program 40 can be stored in the memory unit 22 by using the communication interface 16. For example, a setting PC 105 (personal computer, see FIG. 1) may be connected to a local area network 102 connected to the communication interface 16, and firmware including the program 40 may be supplied from the setting PC 105 to the notification device 1. The setting PC 105 may be directly connected to the notification device 1 via the communication interface 16. The firmware update can be performed in the same manner. Also, when the communication between the notification device 1 and the network 100 is established, the computer 20 may communicate with a firmware supply server (not shown) on the network 100 so that a new version of the firmware can be autonomously acquired.

[0034] FIG. 4 is a diagram for explaining the function of the mail filtering program 42, and shows an example of a setting screen 60 for filtering rules. For example, by connecting the setting PC 105 to the local area network 102 to which the notification device 1 is connected (or directly connecting to the communication interface 16) and starting a web browser, and accessing the address (IP address) set in the notification device 1, the setting screen 60 shown in FIG. 4 can be displayed on the display device of the setting PC 105.

[0035] In this example, the filtering rules can be set by one or more filtering conditions. The filtering condition setting unit 61 for setting each filtering condition includes a type setting unit 62, a search character string input unit 63, and a matching condition setting unit 64. That is, each filtering condition is set by a combination of a condition type, a search character string, and a matching condition. The type setting unit 62 is configured to be able to select one condition type from default options such as the sender (email address), subject, and body. Any character string to be included in the condition can be entered in the search character string input unit 63. The matching condition setting unit 64 is configured to be able to select one matching condition from default options such as "matches", "starts with", "includes", and "does not include".

[0036] Furthermore, in this example, there is an operation selection button 65 for selecting any one of "matches all of the following" for extracting emails that match all of the set filter conditions, "matches any of the following" for extracting emails that match any of the filter conditions, or "none" for extracting all emails regardless of the filter conditions.

[0037] By appropriately setting the filtering rules from such a setting screen 60, emails to be read aloud can be appropriately selected. For example, by setting as a filter condition that the sender (email address) includes the email address of a predetermined public office, emails sent from that public office can be extracted as read-aloud targets. Also, by setting as a filter condition that a specific word (for example, "urgent", "alert", "alarm", etc.) is included in the subject or the text, emails containing that specific word can be extracted as read-aloud targets. Therefore, appropriate emails can be extracted as read-aloud targets and inappropriate emails can be excluded from the read-aloud targets, so that notification by read-aloud voice can be appropriately performed. Also, since the amount of target data such as voice synthesis processing decreases, the processing load on the computer 20 can be reduced.

[0038] FIG. 5 is a diagram for explaining an example of the functions of the email reading text generation program 43. The email reading text generation program 43 may cause the computer 20 to execute a step of deleting a line 71 (more specifically, a line terminated by a URL description) including a URL (Uniform Resource Locator) description from the text included in the email body (email text) 70, and a step of deleting a line 72 immediately before the line 71 including the URL description. The URL description in the email text 70 represents the location of the destination web page, and the URL description itself has no meaning and does not form meaningful words, so there is little significance to be vocalized. Also, the line 72 immediately before the URL description in the email text 70 is, in most cases, a description introducing that supplementary information or the like exists in the URL. Therefore, if the URL description line 71 is deleted, it loses its meaning, so it is appropriate to delete it together. Thereby, the text to be read aloud, that is, the text to be speech-synthesized, can be reduced. In this specification, "deletion" means excluding from the reading text.

[0039] The email reading text generation program 43 recognizes a line break code as a line delimiter. That is, the computer 20 that executes the email reading text generation program 43 recognizes logical lines (logical lines) delimited by line break codes instead of the lines at the time of display (display lines), and performs processing on the email text. If necessary, before starting the processing of the text of the email body, a formatting process of deleting line break codes in the middle of a sentence may be executed.

[0040] The email reading text generation program 43 may also cause the computer 20 to execute a step of extracting a line whose end is a punctuation mark from the text included in the email body and adding it to the reading text. In the processing by the computer 20, the corresponding line can be extracted by searching for a punctuation mark and the subsequent line break code. Thereby, since a part suitable for reading aloud can be extracted from the email body, a part representing the gist of the email can be extracted, and the text to be read aloud can be reduced.

[0041] Note that punctuation marks include full stops and commas. A full stop is a symbol for separating sentences, such as "。", ".", "?", "!", ".", etc. A comma is a delimiter in a sentence (mainly a delimiter considering readability), such as "、", "·", ",", ";", ":", etc. A sentence is a series of character strings separated by full stops, and may include line break codes in the middle. In a normal email, the line break code in the middle of a sentence is placed immediately after a comma.

[0042] The email reading text generation program 43 may also cause the computer 20 to execute a step of extracting the text of a sentence described in polite form and adding it to the reading text. The polite form, also called the "desu-masu form", refers to the one that uses "desu" or "masu" in the expression specifying the end of a sentence. In the process by the computer 20, it is determined whether it is in the polite form by searching for character strings such as "です", "でした", "ます", "ました", "ません", "ください" (and further, if necessary, a full stop following the character string). Since the polite form is an oral style, by extracting the sentence described in the polite form, it is possible to extract a part suitable for reading from the email body. Thereby, it is possible to extract the part representing the gist of the email and reduce the text to be read.

[0043] In addition, the email reading text generation program 43 may cause the computer 20 to execute a step of deleting a line whose beginning is a predetermined symbol from the text included in the email body. The predetermined symbol typically corresponds to an annotation symbol such as "※" or a citation symbol such as ">" representing a cited portion of another person's text. The predetermined symbol may be settable by the user. For example, by accessing the notification device 1 with the setting PC 105, it may be possible to set the predetermined symbol. An annotation symbol is a symbol written prior to the description of an annotation, and since an annotation is a description of supplementary information, it is often appropriate to omit it when concisely representing the gist of an email and converting it into speech. For example, if there is a line in the email body that says "※Reply emails are not accepted.", this line is deleted. Also, since cited portions in the email body are often repetitions of information, it is appropriate to omit them when concisely representing the gist of the email and converting it into speech. Thus, by appropriately determining the predetermined symbol, the text to be read aloud can be reduced.

[0044] In addition, the email reading text generation program 43 may cause the computer 20 to execute a step of deleting a line containing the description of an email address (more specifically, a line that ends with the description of an email address) from the text included in the email body, and a step of deleting the line immediately preceding the line containing the description of the email address. The description of the email address in the email body represents the contact email address, and the description of the email address itself has no meaning and does not form meaningful words even when converted into speech. Also, the line immediately preceding the description of the email address in the email body is, in most cases, a description introducing that the contact is the email address, so if the line containing the description of the email address is deleted, it loses its meaning, so it is appropriate to delete it together. This can reduce the text to be read aloud.

[0045] For example, the following description lines in the email body are deleted. Inquiries can be made here. ****012@**345.com

[0046] In addition, the email reading text generation program 43 may cause the computer 20 to execute steps of deleting a line including a character string in which a predetermined symbol appears continuously a predetermined number or more (for example, 3 or more) times from the text included in the body of the email, and deleting subsequent lines. By this process, typically, a signature column appended to the end of the email body can be deleted. The predetermined symbol is, for example, a symbol used to delimit the signature column of the email, and typical examples include "+", "-", "*", " / ", which represent arithmetic operations, and other symbols such as "=", "#", etc. A combination of two or more predetermined symbols, such as "-*-*-*", may also be regarded as "continuous" of the predetermined symbol. By appropriately setting the predetermined symbol and the predetermined number (repetition number), the description of the signature column in the email body can be deleted from the reading target.

[0047] For example, the following description lines in the email body are deleted. ----------------------------------- Patent Co., Ltd. Development Headquarters, Development Section Fuku ○, Tai △ Mita City, Hyogo Prefecture ***** TEL / 07*-56*-7**7 e-mail / ****.******@****.co.jp -----------------------------------

[0048] In addition to the above, the email reading text generation program 43 may cause the computer 20 to execute a step of extracting a line or sentence including a string subjected to a predetermined highlighting process and adding it to the reading text. The highlighting process includes coloring with a color different from the normal character color (typically black), boldface processing, italic processing, underlining processing, using an emphasized font different from the normal font, using a font larger than the normal size, and the like. By searching for a string subjected to such a highlighting process and extracting a sentence or line including the highlighted string, important portions in the email body can be extracted and added to the reading text. In this way, a reading text that can convey the gist and has a small amount of text can be generated.

[0049] Further, the email reading text generation program 43 may cause the computer 20 to execute a step of extracting a line or sentence including a predetermined keyword and adding it to the reading text. The predetermined keyword may be settable by the user. For example, by connecting the setting PC 105 to the notification device 1 via the local area network 102, it may be possible to set a predetermined keyword. For example, keywords such as "emergency", "alert", and "heat stroke" may be set. By searching for such a keyword and extracting a sentence or line including the keyword, important portions in the email body can be extracted and added to the reading text. In this way, a reading text that can convey the gist and has a small amount of text can be generated.

[0050] Further, the email reading text generation program 43 may cause the computer 20 to perform a process of deleting a description portion of supplementary information. For example, the computer 20 searches for an opening parenthesis "(" and a corresponding closing parenthesis ") " used in the description of the supplementary information, and deletes the text of these parentheses and the description portion sandwiched between them. Thereby, the description portion of the supplementary information can be excluded from the reading target, so that the amount of the reading text can be reduced, and in addition, it becomes easier to understand when voice-activated.

[0051] FIG. 6 is a flowchart for explaining an example of processing executed by computer 20 at a predetermined control cycle. When a preset periodic reception cycle arrives (step S1: YES), computer 20 accesses recipient mail server 104 and executes mail reception processing to acquire mails that have arrived at the preset recipient mail accounts (step S2. Mail reception step. Function of mailer program 41). If there is no received mail (step S3: NO), the processing for the current cycle ends.

[0052] When there is received mail (step S3: YES), computer 20 executes filtering for sorting whether the received mail should be read aloud (step S4. Function of mail filtering program 42). If the received mail is to be excluded from the read-aloud target (step S4: NO), it further checks for the presence of other received mails (step S9). If there are other received mails, filtering is executed on the other received mails (step S4). If there is no received mail to be read aloud (step S4: NO and step S9: NO), the processing for the current control cycle ends. If there is received mail to be read aloud (step S4: YES), read-aloud text generation processing is executed on the received mail (step S5. Read-aloud text generation step. Function of mail read-aloud text generation program 43).

[0053] When read-aloud text is generated by the read-aloud text generation processing, specific message text is added to the read-aloud text (step S6). The specific message text may be, for example, the text of a message such as "Please check the mail for details". The specific message text may also be changeable by connecting setting PC 105 and accessing notification device 1. The process of adding the specific message text may be omitted.

[0054] By performing speech synthesis processing on the read text thus created, voice data is generated (step S7. Speech synthesis step. Function of speech synthesis program 44). Then, by supplying the voice data to the drive unit 24 (step S8), voice is emitted from the speaker 15, and the reading of the read text is executed (function of text reading program 45).

[0055] If there is another received email (step S9: YES), the same processing is repeated.

[0056] FIG. 7 is a flowchart for explaining an example of the processing of the computer 20 that executes the read text generation process (step S5 in FIG. 6), that is, the mail reading text generation program 43. The computer 20 acquires the text body (mail text body) from the mail selected as the reading target by filtering (step S11). The read text is generated from the acquired mail text body.

[0057] The computer 20 executes a deletion process (steps S12 to S24) for deleting unnecessary parts from the text of the acquired mail text body, and an extraction process (steps S25 to S32) for extracting necessary parts from the text of the mail text body and adding them to the read text.

[0058] FIG. 7 shows an example of executing a deletion process and then an extraction process. By executing the deletion process on the text of the email body and then executing the extraction process on the text of the email body after the deletion process, the amount of text to be processed in the extraction process is reduced, so the processing load on the computer 20 can be reduced and the processing time can be shortened. However, the order of the deletion process and the extraction process is not limited to that shown in FIG. 7. Also, the plurality of subordinate processes included in the deletion process can be in any order, and the plurality of subordinate processes included in the extraction process can be in any order. Briefly, the plurality of subordinate processes included in the deletion process and the plurality of subordinate processes included in the extraction process can be executed in any order determined so as to obtain an appropriate result. Also, when an appropriate read-aloud text can be obtained only by the deletion process, the extraction process may be omitted, and the text of the email body that has undergone the deletion process may be used as the read-aloud text. Also, when an appropriate read-aloud text can be obtained only by the extraction process, the deletion process may be omitted.

[0059] The subordinate processes included in the deletion process, in this example, include a URL deletion process (steps S12 to S14), an email address deletion process (steps S15 to S17), a signature field deletion process (steps S18 to S20), a specific symbol line deletion process (steps S21 to S22), and a supplementary information deletion process (steps S23 to S24). However, the deletion process does not necessarily have to include all of these subordinate processes, and may be appropriately selected according to the content to be notified by the notification device 1.

[0060] In the URL deletion process (steps S12 to S14), the computer 20 searches for the description of the URL from the text included in the email body (step S12). When the description of the URL is found, the computer 20 executes the step of deleting the line including the description of the URL (preferably, the line terminated by the description of the URL) (step S14), and the step of deleting the line immediately before the line including the description of the URL (step S13). These two steps S13 and S14 may be executed in either order or simultaneously.

[0061] In the email address deletion process (Steps S15 to S17), the computer 20 searches for the description of an email address (electronic mail address) from the text included in the email body (Step S15). For example, it may search for "@" (at sign), and regard a series of characters including that "@" (at sign) as an email address. When the description of the email address is found, the computer 20 executes the step of deleting the line including the description of the email address (preferably, the line ending with the description of the email address) (Step S17), and the step of deleting the line immediately before the line including the description of the email address (Step S16). These two steps S16 and S17 may be executed in either order or simultaneously.

[0062] In the signature field deletion process (Steps S18 to S20), the computer 20 searches for a character string in which a predetermined symbol continues for a predetermined number or more from the text included in the email body (Step S18). This is a process for finding the signature field included in the email body. Therefore, the character string to be searched is a character string that is often used for delimiting the signature field in the email body. When a character string in which a predetermined symbol continues for a predetermined number or more is found, the computer 20 executes the step of deleting the line including that character string (Step S19), and the step of deleting the subsequent lines (Step S20). These two steps S19 and S20 may be executed in either order or simultaneously.

[0063] In the specific symbol line deletion process (Steps S21 to S22), the computer 20 searches for a predetermined symbol (such as a comment symbol or a quotation mark) at the beginning of a line from the text included in the email body (Step S21). When the predetermined symbol at the beginning of the line is found, the computer 20 executes the step of deleting the line in which the predetermined symbol is arranged (Step S22).

[0064] In the supplementary information deletion process (steps S23 to S24), the computer 20 searches for the opening parenthesis "(" and the corresponding closing parenthesis ")" used for the description of the supplementary information from the text included in the email body (step S23). When the corresponding part is found, the computer 20 deletes those parentheses and the text of the description part sandwiched between them (step S24). Thereby, the description part of the supplementary information is deleted.

[0065] In one example, the subordinate processes included in the extraction process include an emphasized part extraction process (steps S25 to S26), a keyword part extraction process (steps S27 to S28), and a punctuation line extraction process (steps S29 to S30). In another example, the subordinate processes included in the extraction process include an emphasized part extraction process (steps S25 to S26), a keyword part extraction process (steps S27 to S28), and a polite sentence extraction process (steps S31 to S32). However, the extraction process does not necessarily have to include all of these subordinate processes, and may be appropriately selected according to the content to be notified by the notification device 1.

[0066] In this embodiment, the extraction process is executed on the text of the email body after the deletion process has been performed.

[0067] In the emphasized part extraction process (steps S25 to S26), the computer 20 searches for a character string (emphasis processing part) subjected to a predetermined emphasis process from the text of the email body (here, the email body after the deletion process) (step S25). For example, in the case of an HTML (Hypertext Markup Language) email, the emphasized character string may be searched by searching for the attributes (character color, font, character size, font, etc.) specified for each character of the text of the email body. When the emphasized character string is found, the computer 20 executes the step of extracting the line or sentence including the emphasized character string and adding it to the spoken text (step S26).

[0068] In the keyword location extraction process (steps S27 to S28), the computer 20 searches for a predetermined keyword from the text of the email body (here, the email body after the deletion process) (step S27). When a predetermined keyword is found, the computer 20 executes a step of extracting the line or sentence containing the keyword and adding it to the speech text (step S28).

[0069] In the punctuation mark line extraction process (steps S29 to S30), the computer 20 searches for the punctuation mark at the end of the line from the text of the email body (here, the email body after the deletion process) (step S29). When the punctuation mark at the end of the line is found, the computer 20 executes a step of adding the line containing the punctuation mark to the speech text (step S30).

[0070] Instead of the punctuation mark extraction process (steps S29 to S30), a polite sentence extraction process (steps S31 to S32) may be executed. Both the punctuation mark extraction process (steps S29 to S30) and the polite sentence extraction process (steps S31 to S32) may be executed, but it is considered that the processing results are not significantly different from those when either one is executed. In the polite sentence extraction process (steps S31 to S32), the computer 20 searches for the end of a polite sentence from the text of the email body (here, the email body after the deletion process) (step S31). When the end of a polite sentence is found, the computer 20 executes a step of extracting the text of the sentence containing the end of the polite sentence and adding it to the speech text (steps S31 to S32).

[0071] For example, by executing the speech text generation process of FIG. 7 for the sentence example shown in FIG. 5, the following speech text is generated. Note that this example shows the processing result when the process of deleting the line break code in the middle of the sentence is also executed in the deletion process. "In Saitama Prefecture, it is predicted that tomorrow will be a weather situation with an extremely high risk of heat stroke. Please avoid going out as much as possible and spend time in a cool environment indoors with an air conditioner or the like." Also, except in special cases, please refrain from exercising. Check the heat index in your vicinity and take actions to prevent heatstroke. This information is released when the heat index is predicted to be 33 or higher. It is released around 17:00 on the day before the target prediction date or around 5:00 on the prediction date. In prefectures where information was released on the day before the target prediction date, even if the prediction for the current day drops below 33, information will still be released around 5:00 to call attention and prevent any relaxation of vigilance against heatstroke. Please check the email for details.

[0072] In this way, a read-aloud text suitable for vocalizing the gist of an email and transmitting it in a short time is generated, enabling easy-to-understand information transmission. Moreover, since the amount of text for speech synthesis is reduced, the processing load for speech synthesis can be alleviated, and the time required for speech synthesis can be shortened.

[0073] Above, an embodiment of this invention has been described. However, this invention can also be implemented in other forms.

[0074] For example, the deletion process (steps S12 - S24) shown in the above-described embodiment may include subordinate processes other than those described above. Specifically, a delimiter front part deletion process that searches for a delimiter symbol in the middle of a line and deletes the delimiter symbol within the line and the string before it may be included in the subordinate processes of the deletion process. Examples of delimiter symbols include " " (space), ":" (colon), ";" (semicolon), etc. For example, if there is a line in the email body that says "Content: There is a risk of river flooding.", the front part before the delimiter symbol ":" is deleted, and the rear part "There is a risk of river flooding." remains in the text to be processed. This enables the generation of a read-aloud text that is even more suitable for voice notification. The same applies to the extraction process (steps S25 - S32), which may also include subordinate processes other than those described above.

[0075] In the above-described embodiment, the notification device 1 includes the notification device 13 having the light emitting units 13a, 13b, and 13c. However, the present invention may be applied to a notification device that does not include a light emitting unit and performs notification only by voice.

[0076] In the above-described embodiment, an example in which the mail reading text generation program 43 is incorporated in the notification device 1 has been described. However, the mail reading text generation program 43 may be incorporated in a device other than the notification device. For example, the mail reading text generation program 43 may be incorporated in the host computer 106 (see FIG. 1) that can communicate with the network 100 and the notification device 1. Specifically, the host computer 106 may receive a mail, extract the mail body from the mail, generate a reading text, and transmit the generated reading text to the notification device 1. In this case, the notification device 1 does not need to have a reading text generation function, and may have a function of voice-synthesizing the input reading text. Further, in the host computer 106, the reading text may be voice-synthesized to generate voice data, and the voice data may be supplied to the notification device 1. In this case, the notification device 1 does not need to have a voice synthesis function, and may have a pronunciation means for sounding the voice data.

[0077] In addition, various design changes can be made within the scope of the matters described in the claims.

Explanation of Reference Numerals

[0078] 1 Notification device 13 Notification device 15 Speaker 20 Computer 41 Mail program 42 Mail filtering program 43 Mail reading text generation program 44 Voice synthesis program 45 Text reading program 51 Mail 52 Mail body text 53 Read aloud text 54 Voice data

Claims

1. A program for a computer to generate speech text used to read aloud the text contained in the body of an email, causing the computer to execute the steps of deleting lines containing a description of a URL (Uniform Resource Locator) from the text contained in the body of the email, and deleting the line immediately preceding the line containing the description of the URL, a speech text generation program for email reading.

2. The email speech text generation program according to claim 1, further causing the computer to execute the step of deleting lines whose line start is a predetermined symbol from the text contained in the body of the email.

3. The email speech text generation program according to claim 1, further causing the computer to execute the step of extracting lines whose line end is a punctuation mark from the text contained in the body of the email and adding them to the speech text.

4. The email speech text generation program according to claim 1, further causing the computer to execute the step of extracting the text of a sentence described in polite form and adding it to the speech text.

5. The email speech text generation program according to claim 1, further causing the computer to execute the steps of deleting lines containing a description of an email address from the text contained in the body of the email, and deleting the line immediately preceding the line containing the description of the email address.

6. The email speech text generation program according to claim 1, further causing the computer to execute the steps of deleting lines containing a character string in which a predetermined symbol continues for a predetermined number or more consecutive times from the text contained in the body of the email, and deleting the subsequent lines.

7. The mail reading text generation program according to claim 1, further causing the computer to execute a step of extracting a line or sentence including a string subjected to a predetermined emphasis process and adding the same to the reading text.

8. The mail reading text generation program according to claim 1, further causing the computer to execute a step of extracting a line or sentence including a predetermined keyword and adding the same to the reading text.

9. A computer that executes a mail program for receiving an electronic mail, the mail reading text generation program according to any one of claims 1 to 8, and a voice synthesis program for synthesizing voice data for vocalizing the reading text generated by the mail reading text generation program, a mail receiving step of executing the mail program to receive an electronic mail; a reading text generation step of executing the mail reading text generation program to generate a reading text used for reading the text included in the body of the electronic mail received in the mail receiving step; a voice synthesis step of synthesizing voice data of the reading text generated in the reading text generation step by executing the voice synthesis program.

10. The computer according to claim 9, and a pronunciation means for acoustically outputting the voice data generated in the voice synthesis step.

11. The mail receiving / readout device according to claim 10, and a notification device that is controlled by the computer and generates a visual notification signal.

Citation Information

Patent Citations

  • Electronic mail device, reading out method for electronic mail in voice, and medium recording program

    JP1999085458A