Information system

Abstract

An information system includes an output section to read a character string in received contents aloud, and a memory section to store a rule set to convert input contents to modified contents adapted to text reading. An optimizing section is arranged to covert the input contents to the modified contents according to the rule set, and to cause the output section to read out the modified contents.

Description

BACKGROUND OF THE INVENTION

The present invention relates to information system and method for producing speech of input text files and thereby providing audible information.

U.S. Pat. No. 6,226,614 B1 (corresponding to IP 11-202884A) shows a system for producing synthetic speech messages.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide information system and method for modifying text contents for production of speech without requiring much of human resources.

According to one aspect of the present invention, an information system comprises: an output section to read out a character string in received contents; a memory section to store a rule set to convert input contents to modified contents adapted to text reading; and an optimizing section to covert the input contents to the modified contents according to the rule set, and to cause the output section to read out the modified contents.

According to another aspect of the invention, an information supplying process comprises: receiving input contents; converting the input contents to modified contents adapted to text reading automatically according to a preliminarily stored rule set; and supplying the modified contents to speak the modified contents aloud.

The other objects and features of this invention will become understood from the following description with reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic view showing an information system according to one embodiment of the present invention. embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

An information system of earlier technology is arranged to supply audible information to drivers and passengers in vehicles by using Text-To-Speech (TTS). TTS is technology for creating audible speech from computer readable text file inputted to TTS application. Input text files are prepared by the following ways. (1) Text contents are prepared from data base in consideration of speech production by TTS. (2) Text contents available on the Web are diverted to the preparation. Text contents prepared by (1) are adequate for speech production mostly. On the other hand, text contents prepared by (2) are adequate for presentation of Web contents, but often inadequate for speech production by TTS.

When, for example, Web contents in the Japanese language (English translation is provided in braces { }) “[ 20 ]{London on the twentieth} -2004 (-□ 2004) 3 {Three games of qualifying round of Soccer European championships (Euro 2004) were played in various locations}, 9--2-0, 32. {Walse at the top in the ninth group won an away game easily against Azerbaijan by 2−0, and increased the point difference from the second and subsequent teams by an undefeated score of three wins}” are inputted to TTS, speech spoken aloud is: ┌-, , .┘ {parenthesis London the 20th parenthesis three games of qualifying round of Soccer European championships parenthesis Euro 2004 parenthesis were played in various locations, Walse at the top in the ninth group won an away game easily against Azerbaijan by two minus zero, and increased the point difference from the second and subsequent teams by an undefeated score of three wins}┘ in this case, “[“and”]” are pronounced as “” {parenthesis} (pronunciation of symbol character strings inadequate for sound production). Moreover, “2−0” is pronounced as “” {two minus zero}(inappropriate pronunciation of predetermined formats).

Thus, most texts contents used in Web contents are not adequate for speech production by TTS. Therefore, the information system of earlier technology requires modification by human operators to adapt text contents to the speech production. The modification includes operations of checking inappropriately pronounced portions by performing speech production of text contents by TTS, and operations of modifying text contents to modify the detected inappropriate portions. Thus, the information system of the earlier technology requires lots of human resources

FIG. 1 shows an information system 1 according to one embodiment of the present invention. This information system 1 is a system for providing information to vehicles. Information system 1 includes a center system 2 for transmitting and supplying information to driver and/or passenger of at least one vehicle, and at least one in-vehicle apparatus 9 mounted in a vehicle 10, for receiving the information from center system 2.

Center system 2 includes a category discriminating section 4 determining a category of input text contents inputted to center system 2, a memory section 3 storing a collection of rules for modifying character strings in the input contents to modified character strings suitable for text reading of TTS, an optimizing section 5 to optimize the text contents, and a TTS text contents transmitting section 8 to transmit the optimized or modified text contents to in-vehicle apparatus 9. Optimizing section 5 of this example includes a character string replacement section 6 and a character string format conversion section 7.

Memory section 3 of this example includes a rule memory section 13 storing various rules for modifying character strings as one of first rule group (or first rule set) or second rule group (or second rule set), and an optimization information memory section 14 storing optimization information including information on the order of rules and operations used for optimization in each of categories.

When text contents are inputted to center system 2 from a content provider, category discriminating section 4 identifies the category of the input contents. Then, the character string replacement section 6 and format conversion section 7 in optimizing section 5 select rules for optimization in accordance with information on the category identified by category discrimination section 4, referring to the optimization information stored in optimization information memory section 14; and thereby carry out optimizing operations on the text contents to achieve adequate sound reproduction.

The thus-optimized (modified) text contents for TTS are transmitted by TTS text contents transmitting section 8, to in-vehicle apparatus 9 of vehicle 10. In-vehicle apparatus 9 includes an sound output section 11. This sound output section 11 converts the optimized text contents to audible sound by using TTS, and thereby supplies information to the driver and passengers in vehicle 10 through a loudspeaker system.

Tables I, II and III (at the end of the detailed description of the invention) show rules stored in rule memory section 13 of memory section 3 of this example. By using these rules, character string replacement section 6 and format conversion section 7 perform optimizing operations of text contents in the following manner. Character string replacement section 6 replaces character or charters in text contents, with other character or characters or NULL according to the first rule group stored in rule memory section 13. Format conversion section 7 converts character or characters in text contents, into a form in a character string format suitable for the contents according to the second rule group stored in rule memory section 13. In these tables, <number> and <number string> denote “number” and “number string”, respectively (not “<number>” and “<number string>”).

Operation No. 1 in Table I provides one rule of the first rule group. The optimizing operation of Operation No. 1 is to replace a character string (such as symbol) undesirable for pronunciation, with NULL, to optimize text contents for automatic production of speech by TTS, as in the following example.

EXAMPLE 1

Replace each of character strings “<” and “>” in text contents, with NULL.

For example, a character string “<{today's weather}>” is pronounced by TTS as ┌{parenthesis today's weather parenthesis}┘ The undesired character ┌{parenthesis}┘. is read aloud. This character string is modified by the optimizing operation of Operation No. 1, into a character string “{today's weather}”, and pronounced by TTS as {today's weather}┘.

Other examples of character strings which are not desired for pronunciation are: “(“, ”)”, and “▪” (which is used, for example, to emphasize subsequent sentences). By an operation similar to the operation of Example 1, the information system replaces each of “(“, ”)”, and “▪” with NULL, and thereby removes “(“, ”)”, and “▪” from text contents, for production of adequate speech by TTS.

As one rule of the first rule group, the optimizing operation of Operation No. 4 is to replace linefeed (or line break)(for starting a new line or a new paragraph) in input text contents, with a pause symbol, and thereby to insert a pause (silent pause) in a sentence for optimum production of speech by TTS, as in the following example.

EXAMPLE 2

Replace a linefeed code in text contents, with a pause symbol.

For example, a character string “ . . . . (linefeed code)- . . . ” {“ . . . occurring in Kawasaki City. (linefeed code) Purchase a regular high-priced card . . . ”} is modified by the optimizing operation of Operation No. 4, into a character string “ . . . . (pause symbol)- . . . ”{“occurring in Kawasaki City. (pause) Purchase a regular high-priced card . . . ”}, and pronounced by TTS as ┌ . . . (pause)- . . . ┘ {┌ . . . occurring in Kawasaki City. (pause) Purchase a regular high-priced card . . . ┘}, with a pause in a proper portion in the speech.

As rule of the first rule group, the optimizing operations of Operation Nos. 11 and 14˜24 shown in Table II are to replace abbreviated expressions and symbols in input text contents, with unabbreviated forms for optimum production of speech by TTS, as in the following example.

EXAMPLE 3

Replace “({Sun})”, “({Mon})” and so fourth with the adequate full forms.

For example, a character string “({Sun})” is pronounced by TTS as ┘ {parenthesis Sun parenthesis}, including the unwanted pronunciation of parenthesis. By performing the optimizing operation of No. 14, the information system according to this embodiment rewrites “({Sun})” into “ {Sunday}”, and pronounces, with TTS, as ┌┘ {Sunday}.

EXAMPLE 4

Replace a wave dash “˜” used for indicating a range as in “10˜12”, with an adequate expression.

For example, a character string “10˜12” is pronounced by TTS improperly as ┘ {ten wave dash twelve}, including the inadequate pronunciation of wave dash. By performing the optimizing operation of No. 24, the information system according to this embodiment rewrites “10˜12” into “10 12” {from 10 to 12}, and pronounces, with TTS, as ┌┘ {from ten to twelve}. This replacement corresponds to replacement of “10˜12” with “10 to 12” in the case of English.

As rule of the second rule group, the optimizing operations of Operation Nos. 26, 27 and 35 shown in Table III are performed by format converting section 7, to convert a character string in a prescribed format in text contents into a character string in a modified format suitable for optimum production of speech by TTS, in accordance with a category of the text contents determined by category discriminating section 4, as in the following example.

EXAMPLE 5

Convert a character string such as “0.368” indicating a batting average in contents about baseball, into a character string in an adequate format.

For example, a character string “0.368” is pronounced by TTS as ┌┘ {point three hundred sixty eight}, in a format inadequate to indicate the batting average. By performing the optimizing operation of No. 27, the information system according to this embodiment rewrites the character string into “3 6 8”, and pronounces, with TTS, as ┌┘ {point three sixty eight}.

EXAMPLE 6

Convert a character string such as “2−0” in a format representing a score in contents relating to sports, into a character string in an adequate format.

For example, a character string “2−0” is pronounced by TTS improperly as ┌┘ {two minus zero} in a format inadequate to indicate the score. By performing the optimizing operation of No. 28, the information system according to this embodiment rewrites the character string into “2 0” {2 to 0}, and pronounces, with TTS, as ┌┘ {two to zero}.

EXAMPLE 7

Convert a character string in a format representing an amount of money, into a character string in an adequate format.

For example, a character string “¥1000” is pronounced by TTS improperly as ┌I┘ {yen one thousand}. By performing the optimizing operation of No. 36, the information system according to this embodiment rewrites the character string into “1000 I” {1000 yen}, and pronounces, with TTS, as ┌I┘ {a thousand yen}.

When input text contents are HTML contents, the optimizing operations of Operation Nos. 31˜34 shown in Table III as rules of the second rule group are performed by format converting section 7, to convert a character string such as a tag character string in HTML contents into a prescribed character string or to remove the character string in HTML contents, for optimum production of speech by TTS, as in the following example.

EXAMPLE 8

Remove a tag character string in HTML contents.

When, for example, HTML contents includes tags, such as “<head>”, “<a>”, “<img>”, “<center>”, or a portion enclosed by these tags or a portion (comment portion) enclosed by comment tags, tag character strings are pronounced by TTS uselessly. As to these character strings, by performing the optimizing operations of No. 32˜34, the information system according to this embodiment removes tags, portions enclosed by tags and portions enclosed by comment tags, for adequate sound reproduction by TTS.

EXAMPLE 9

Convert “<br>” tag inserting a line break in HTML contents, into a pause symbol by the optimizing operation of Operation No. 35. Therefore, a pause is inserted in a proper position in speech produced by TTS.

By performing these optimizing operations on input HTML contents as shown in Table IV, the information system according to this embodiment converts the input contents into modified text contents suitable for sound reproduction by TTS as shown in Table V, by removing unwanted tag character strings and converting “<br>” tag to a pause symbol. Consequently, the modified text contents are read aloud by TTS as shown in Table VI, in an adequate manner excluding unwanted characters. Other operations in Tables I, II and III are carried out in the same manner.

Character string replacement section 6 and format converting section 7 in optimizing section 5 make reference to the optimization information in optimization information memory section 14 in accordance with the category identified by category discriminating section 4; select rules prescribed in the optimization information from rule memory section 13; and optimize the input text contents by using the selected rules. Optimization information memory section 14 stores optimization information of each category.

When, for example, the category of the input text contents is “general” as in examples 1˜4, optimizing section 5 selects rules, such as rules of Operation Nos. 1˜26, from rules stored in rule memory section 13 by referring to the optimization information of the category “general”; and optimize the input text contents according to the selected rules. When the category of the input text contents is “sports” as in examples 5 and 6, optimizing section 5 selects rules, such as rules of Operation Nos. 26 and 27, stored in rule memory section 13, and optimize the input text contents.

In the thus-constructed information system according to this embodiment, optimizing section 5 carries out optimizing operations on input text contents (or before-process text contents) inputted to center system 2, according to the first rules and second rules stored in rule memory section 13 of memory section 3, to modify the input contents into modified contents adapted to the speech reproduction by TTS. On the basis of the modified contents (or after-process contents), sound output section 11 in in-vehicle apparatus 9 produces sound by TTS, and thereby supply audible information to driver and/or passenger in the vehicle. This information system converts the input text contents to the modified-contents automatically without need for human resources.

Character string replacement section 6 of optimizing section 5 selects rules from the rules in rule memory section by referring to the optimization information corresponding to the category identified by category discriminating section 4, and performs optimizing operation according to the selected rules. Therefore, a character string inadequate for audible speech reproduction in the before-process contents is replaced with a prescribed adequate character string as in example 1, so that the speech spoken aloud includes no pronunciation of unwanted symbols, and the naturalness is increased.

Furthermore, format converting section 7 in optimizing section 5 refers to the optimization information for the category of the input contents identified by category discriminating section 4, and selects rules from the rules in rule memory section 13. Therefore, a character string format or form of expression in the before-process contents is converted into an adequate character string format or expression form adequate in the category of the contents for adequate speech production by TTS.

Optimization information memory section 14 of memory section 3 stores information on the order of application of rules on text contents for optimization, and information on different optimizing operations. By using such optimization information selected in accordance with the category of contents, optimizing section 5 converts before-process text contents into modified text contents adequate for the contents category.

The information system according to the embodiment is a system for providing various information to vehicles. However, the present invention is not limited to such an information system for vehicles. The illustrated example treats a Japanese speaking information system. However, the present invention is not limited to the Japanese language system. The information system and method may be for English or other languages.

In the illustrated embodiment, section 3 can serve as means for storing a rule set to convert input contents to modified contents adapted to text reading; section 5 can serve as means for converting the input contents to the modified contents according to the rule set; and section 8 can serve as means for supplying the modified contents to speak the modified contents aloud.

This application is based on a prior Japanese Patent Application No. 2003-015170 filed on Jan. 23, 2003. The entire contents of this Japanese Patent Applications No. 2003-015170 are hereby incorporated by reference.

Although the invention has been described above by reference to certain embodiments of the invention, the invention is not limited to the embodiments described above. Modifications and variations of the embodiments described above will occur to those skilled in the art in light of the above teachings. The scope of the invention is defined with reference to the following claims.

TABLE 1
RULES
NO.	OPERATION	PRONUNCIATION
1	Remove prescribed symbol string in	(Not
	text contents.	pronounced)
2	Remove prescribed symbol string	(Not
	from text contents if similar symbols	pronounced)
	follow successively.
3	Remove linefeed	(Not
		pronounced)
4	Replace linefeed with pause symbol	(Pause)
5	Remove half-size space	(Not
	(corresponding to the size of en	pronounced)
	quad) and full-size space
	(corresponding to the size of em)
6	Replace half-size space and full-size	(Pause-Pattern
	space with pause symbol (A)	A)
7	Replace half-size space and full-size	(Pause-Pattern
	space with pause symbol (B)	B)
8	Remove “{”, “}”, “(”, “)”, “ ┌”, “┘ ”,	(Not
	“[” and “]”	pronounced)
9	Replace “{”, “}”, “(”, “)”, “ ┌”, “┘ ”,	(Pause-Pattern
	“[” and “]”, with pause symbol (A)	A)

TABLE II
NO.	OPERATION	PRONUNCIATION
10	Replace “{”, “}”, “(”, “)”, “ ┌”, “┘ ”,	(Pause.Pattern B)
	″[″ and ″]″, with pause symbol (B)
11	Replace “˜” with “{corresponding	{kara in
	to “to” in English}”	romaji}
12	Replace “<” and “<” with “·”.	(Pause)
13	Replace “>” and “>” with “·”.	(Pause)
14	Replace “( {Sun})” with “z,804
	{Sunday}”	{nichlyoubi in
		romaji}
15	Replace “( {Mon})” with “
	{Monday}”	{getsuyoubi in
		romaji}
16	Replace “( {Tue})” with “	{kayoubi
	{Tuesday}”	in romaji}
17	Replace “( {Wed})” with “
	{Wednesday}”	[suiyoubi in
		romaji]
18	Replace “( {Thu})” with “
	{Thursday}”	{mokuyoubi in
		romaji}
19	Replace “( {Fri})” with “
	{Friday}”	{kinyoubi in
		romaji}
20	Replace “( {Sat})” with “	{doyoubi
	{Saturday}”	in romaji}
21	Replace “( {Hol})” with “	{saijitu in
	[Holiday]”	romaji}
22	Replace “( {Off})” with “	{kyujitu
	{Offday}”	in romaji}
23	Replace “( {Fet})”
	with “{Fete	[shukujitu in
	day}”	romaji]
24	Replace “<number>˜<number>” with	<number>
	“<number>{corresponding to “to”	{kara in
	in English}<number>”	romaji}<number>
25	Replace “=” and “=” with “·”.	(Pause)

TABLE III
NO.	OPERATION	PRONUNCIATION
26	Replace	<number>)
	“<number> <number> <number>”	{wa{grave over (r)}i in
	with “<number>{wari in romaji	romaji}<number>
	meaning one tenth}<number>{bu	{bu in
	in romaji meaning one	romaji}<number >
	hundredth}<number>{rin in	{rin in
	romaji meaning one thousandth}”.	romaji}
27	Replace “<number>-<number>” with	<number>{tai
	“<number>{tai in romaji meaning	In romaji}
	versus}<number>”.	<number>
28	Replace “. (half-size and full-size)”	(Pause)
	with “, ”.
29	Replace linefeed with “∘ (punctuation	(Pause)
	mark)”.
30	Replace “! (half-size and full-size)”	(Pause)
	with “∘ (punctuation mark)”.
31	Remove prescribed HTML tags.	(Not pronounced)
32	Remove a portion between <head>	(Not pronounced)
	and </head>.
33	Remove a portion between <a> and	(Not pronounced)
	</a>.
34	Replace <br> tag with pause symbol.	(Pause)
35	Replace “¥<number string>” with	<number string>
	“<number string>{yen}”.	{yen}
36	Replace “$<number string>” with	<number string>
	“<number string>{doru In romaji	{doru in
	meaning dollar}”.
37	Replace “<number string>° C.” with “	<number
	{Celsius}<number string>	string>
	{degrees}”
38	Replace “<number string	<number string
	A>/<number string B> with	B> {bunno
	”<number string B>{bunno in	in romaji}
	romaji}<number string A>.	<number string
		A>

TABLE IV
Input HTML contents
<HTML>
<HEAD>
<TITLE></TITLE>
<META http-equiv=“content-type”content=“text/html;
charset=Shift_JIS”>
</HEAD>
<BODY bgcolor=“#FFFFFF” text=“#000000” link=“#0000FF”
alink=“#FF0000” vlink=“#FF0000”
<CENTER>
<BR>
<HR color=“#24B6FF”>
<IMG src=“IMG/01_1.gif”><BR>
<BR>
<BR>
<BR>
<BR><BR>
</CENTER>
<BR>
<CENTER>(1/3)<BR>
<A href=“voice01_2.html”></A>><BR>
<A href=“index.html”>TOP</A><BR>
</CENTER>
<HR color=“#24B6FF>
<A href=“../../index.html”accesskey=“#”>#cubcTOP</A><BR>
<A href=“../../../../index.html”accesskey=“*”>
*TOP</A>
</BODY>
</HTML>
The following is the above HTML contents including English
translation in braces { }.
{<HTML>
<HEAD>
<TITLE>{developer's voice}</TITLE>
<META http-equiv=“content-type”content=“text/html;
charset=Shift_JIS”>
</HEAD>
<BODY bgcolor=“#FFFFFF” text=“#000000” link=“0000FF”
alink=“#FF0000” vlink=“#FF0000”
<CENTER>
{developer's voice}<BR>
<HR color=“#24B6FF”>
<IMG src=“IMG/01_1.gif”><BR>
{Nissan Taro}<BR>
{Nissan Motor CO., Ltd.}<BR>
{New Model CUBE}<BR>
{Chief Product Specialist}<BR><BR>
</CENTER>
{For
CUBE, this is the first full model change.}
{First
generation Model CUBE opened up a new category of a compact, tall
wagon.}<BR>
{Now, CUBE has received favorable
appraisal from customers of young male mostly and wide varieties
of other classes beyond ages and gender, and we think what can
be called CUBE DNA can be established during these four years
though the history is still short.} <BR>
|As CUBE DNA, there are
simple and casual styling, ease in maneuvering, good interior
comfort, and ease in getting on and off.}
{In the full model change of this time,
we have aimed to further enhance these features and to produce a
car more suitable to sense of value and life style of young
people.}
<CENTER>(1/3)<BR>
<A href=“voice01_2.html“>{Next}</A><BR>
<A href=“index.html”>{Developer's
Voice}TOP</A><BR>
</CENTER>
<HR color=“24B6FF>
<A href=“../../index.html.html”accesskey=“#”>#cubcTOP</A><BR>
<A href=“../../../../index.html”accesskey=“*”>
*{RASHINBAN, Nissan official web site}TOP</A>
</BODY>
</HTML>}

TABLE V
Modified text contents
1/3
The following is English translation of the above modified text
contents.
{Developer's voice (pause)
(pause)
Nissan Taro
Nissan Motor CO., Ltd. (pause)
New Model CUBE (pause)
Chief Product Specialist (pause) (pause)
For CUBE, this is the first full model change. First generation
Model CUBE opened up a new category of a compact, tall wagon.
(pause) Now, CUBE has reveived favorable appraisal from customers
of young male mostly and wide varieties of other classes
beyond ages and gender, and we think what can be called
CUBE DNA can be established during these four years through the
history is still short. (pause) As CUBE DNA, there are simple and
casual styling, ease in maneuvering, good interior comfort, and
ease in getting on and off. In the full model change of this time,
we have aimed to further enhance these features and to produce a
care more suitable to sense of value and life style of young people.
1/3 (pause)
(pause)
(pause)}

TABLE VI
Pronunciation of modified text contents by TTS
{in romaji: kaihatusha no koe (pause)(pause) Nissan Taro (pause)
Nissan jidousha kabushiki kaisha (pause) singata kyubu (pause)
chifu purodakuto supeshiarisuto (pause)(pause) kyubu ni
totte konkaiga hajimeteno furumoderuchenji to narimasu. shodai
kyubu ha konpakuto de haitona wagon toiu atarashii kategori wo
kaitakushimasita. (pause) imadeha wakai dansel wo chushinn ni
nenrei seibetu wo koeta habahiroi souno okyakusama kara hyouka
wo itadaki, rekishi ha mada asai nodesuga, kono yonenkan de
kyubu DNA toyoberumonoga kakuritsu dekita to omotteimasu.
(pause) kyubu DNA toshiteha sinpurude sarigenai sutairingu
konpakuto ka toshiteno unten no shiyasusa kyojuuseinoyosa
joukousei noyosa ga ageraremasu. konkaino
furumoderuchenjideha sorerawo issou kyoukashinagara,
yoriimanowakaihitono kachikan, raifusutairu ni au kuruma wo
mezashimasita sanbunnoichi (pause)(pause)(pause)}

Claims

1. An information system comprising: an output section to read out a character string in received contents; a memory section to store a rule set to convert input contents to modified contents adapted to text reading; and an optimizing section to covert the input contents to the modified contents according to the rule set, and to cause the output section to read out the modified contents.

2. The information system as claimed in claim 1, wherein the memory section includes a rule memory section to store a character string replacement rule group to replace an input character string in the input contents with a modified character string for the modified contents; and the optimizing section includes a character string replacing section to replace the input character string in the input contents with the modified character string for the modified contents according to the character string replacement rule group, and to cause the output section to read out the modified character string.

3. The information system as claimed in claim 1, wherein the memory section includes a rule memory section to store a format conversion rule group to convert an input format in the input contents to a modified format for the modified contents; and the optimizing section includes a character string format converting section to convert the input format in the input contents to the modified format for the modified contents according to the format conversion rule group, and to cause the output section to read out the modified contents in the modified format.

4. The information system as claimed in claim 1, wherein the information system further comprises a category discriminating section to discriminate a category of the input contents; the memory section comprises an optimization information memory section to store optimization information including information on rules selected from the rule set in dependence on the category, and an order of application of the rules selected from the rule set; and the optimizing section is configured to covert the input contents to the modified contents according to rules selected from the rule set in dependence on the category of the input contents by using the optimization information stored in the optimization information memory section.

5. The information system as claimed in claim 1, wherein the output section is mounted on a vehicle.

6. The information system as claimed in claim 5, wherein the information system comprises a center system which includes the memory section; the optimizing section; and a transmitting section to transmit the modified contents to vehicles each equipped with the output section.

7. An information supplying process comprising: receiving input contents; converting the input contents to modified contents adapted to text reading automatically according to a preliminarily stored rule set; and supplying the modified contents to speak the modified contents aloud.

8. The information supplying process as claimed in claim 7, wherein the information supplying process further comprises: identifying a category of the input contents; and selecting rules from the rule sets in accordance with the category, for use for converting the input contents to the modified contents.

9. The information supplying process as claimed in claim 7, wherein the modified contents is supplied to speak the modified contents aloud by transmitting the modified contents to remote terminals.

10. An information system comprising: means for storing a rule set to convert input contents to modified contents adapted to text reading; means for converting the input contents to the modified contents according to the rule set; and means for supplying the modified contents to speak the modified contents aloud.