Data processing: speech signal processing – linguistics – language – Speech signal processing – Synthesis
Reexamination Certificate
2000-11-13
2002-05-28
Dorvil, Richemond (Department: 2641)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Synthesis
C704S258000, C704S271000
Reexamination Certificate
active
06397183
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a document reading system for reading aloud texts in a document inputted to a computer, and to a read control method.
2. Description of the Related Arts
A system disclosed in, e.g., Japanese Patent Application Laying-Open Publication No. 8-272388 is known as a document reading system in the prior art.
This document reading system includes, as a voice synthesizing module for vocally synthesizing text data in kanji characters mixed with kana characters, an embedding module for embedding control data into text data, and an output module for synthesizing and outputting the voices based on the text data with a voice quality corresponding to the control data.
This type of system, however, needs to embed the control data into the text data beforehand in order to read a certain portion with some quality of voice.
Accordingly, if a scheme is that, in the case of reading the HTML texts on the Internet, some proportion of texts are read in a male voice and the rest of those texts are read in a female voice, it is required that the control data be written into the desired portion in the document each time, which is considered laborious.
The present invention, which was devised under such circumstances, aims at providing a technology capable of eliminating a necessity for embedding the control data for giving a reading condition into the document each time.
SUMMARY OF THE INVENTION
According to the present invention, focusing on such a point that an HTML (Hyper Text Mark-up Language) text and an RTF (Rich Text File) text previously contain attribute data (which will hereinafter be called tags as the case may be) for defining a decoration condition, etc. with respect to the text in the document irrespective of reading aloud in voice, the tag is utilized as control data for reading aloud.
A document reading system for analyzing contents of a document with attributes and reading aloud texts in the document by use of a voice synthesizing module, comprises a basic reading condition setting module for, the attributes being set irrespective of reading aloud conditions, setting the reading aloud conditions for the entire document, an individual reading condition setting module for setting the reading aloud condition for every attribute, and a selective reading module for, when reading aloud the document, in principle, reading aloud the text with reference to the basic reading aloud condition set by the basic reading condition setting module, and selectively reading aloud the text having an individual reading aloud condition with a higher priority than the basic reading aloud condition.
Herein, the reading aloud condition is any one of at least a quality (e.g., a voice pitch, a male or female voice), a volume (loudness of voice) and an accent (voice intonations and dialects) of a reading aloud voice, and a selection of whether to read aloud or not. For example, supposing that there is a text such as (<h2> introduction of this homepage </h2>) in the HTML document, tags <h2></h2> specify a size of characters interposed between these tags when displayed. Then, in relation to these tags <h2></h2>, the characters therebetween is read aloud in a male voice.
In particular, it is preferable that the reading aloud condition set for every attribute by the individual reading condition setting module be related to an essential meaning of the attribute, and an essential meaning designated by the attribute can be associated from the reading aloud voice.
To be specific, the tags <h2></h2> indicate the size of the characters, h2 represents a display larger than h3 but smaller than h1. Then, a document specified by h2 is read louder than by h3 but fainter than by h1, whereby the reading aloud can be done in a mode based on the HTML protocol and the document can be visually associated simply by hearing the reading aloud voice.
The document reading system may further comprise a reading condition storage module for storing the reading aloud conditions.
In the document reading system according to the present invention, the basic reading condition setting module sets the reading aloud conditions for the entire document, and subsequently the individual reading condition setting module sets the reading aloud condition for every attribute.
When reading aloud the document, the text with no particular designation is read aloud referring in principle to the basic reading aloud condition set by the basic reading condition setting module.
Various items of data can be known from the tags in the document. In the case of the HTML, the various tags of a title of the page, a heading thereof, content texts, links, mail addresses and so on, are described in the document, and, on a screen, character sizes and colors are selectively set corresponding to the tags. In the prior art reading system, however, all the portions are read aloud in the same voice, with the result that these items of data are lost. According to the present invention, the tag data can be recognized in voice by setting the reading conditions corresponding to the essential data contents of the tags.
Note that the attribute given to the document serves to control how the document is displayed. For example, when the attribute given to the document is tag data, the document is displayed in HTML format.
A document reading method of reading aloud texts in a document by use of a voice synthesizing module according to the present invention, comprises a step of judging an attribute in the document, for controlling display of the document, and a step of changing a condition of reading aloud the text of which the display is controlled by the attribute on the basis of a result of the judgment.
In this document reading method, the reading aloud condition may be changed corresponding to a category of the attribute.
According to the present invention, there is also provided a readable-by-computer recording medium recorded with a program for making a computer, for reading aloud texts in a document by use of a voice synthesizing module, then executing a step of judging an attribute in the document for controlling display of the document, and finally a step of changing a condition of reading aloud the text of which the display is controlled by the attribute on the basis of a result of the judgement.
The readable-by-computer recording medium may be recorded with a program for making the computer further execute a step of changing the reading aloud condition corresponding to a category of the attribute.
REFERENCES:
patent: 5500919 (1996-03-01), Luther
patent: 5850629 (1998-12-01), Holm et al.
patent: 5924068 (1999-07-01), Richard et al.
patent: 6006187 (1999-12-01), Tanenblatt
patent: 6216104 (2001-04-01), Moshfeghi et al.
patent: 0 848 373 (1997-12-01), None
patent: A 01204100 (1989-05-01), None
patent: 01-204100 (1989-08-01), None
patent: 02-289071 (1990-11-01), None
patent: A 02289071 (1990-11-01), None
patent: 06-337876 (1994-12-01), None
patent: A 06337876 (1994-12-01), None
patent: A 08272388 (1996-10-01), None
patent: A 09244869 (1997-09-01), None
TextAssist® (“User's Guide,” Creative Labs ©Feb., 1994).*
Cover (“The XML Cover Pages,” OASIS © 1994).*
“Emacspeak—Speech Interface,” Raman, T.V., Common Ground, Chi '96 Conference Proceedings, Conference on Human Factors in Computing Systems, Vancouver, British Columbia, Apr. 13-18, 1996, pp. 66-71, XP 000657803.
“V-Lynx: Bringing the World Wide Web to Sigh Impaired Users,” Krell Michael et al., Assets '96, British Columbia, Apr. 11-12, 1996, vol. Conference 2, pp. 23-26, XP000683135.
Communication from the European Patent Office (3 pages) dated Sep. 17, 2001, including Supplemental European Search Report dated Sep. 10, 2001.
Baba Hitomi
Fukushima Takahiro
Kanada Momoko
Nakao Makiko
Dorvil Richemond
Fujitsu Limited
Nolan Daniel A.
Staas & Halsey , LLP
LandOfFree
Document reading system, read control method, and recording... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Document reading system, read control method, and recording..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Document reading system, read control method, and recording... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2894911