FR2781629A1 - Automatic alphabet discrimination method for restoring HTML documents in original language, based on analysis of suffix component of URL address - Google Patents

Abstract

The method involves determining the address suffix of a URL and based on successful determination, ordering the reception of the document. At least one reference alphabet is determined for the document as a function of its header data describing the alphabet used, or if this is unsuccessful, a default alphabet is assigned. Independent claims are included for a system interfacing a terminal and server over an IP network, and for a gateway interfacing system between a videotex terminal and a server over the PSTN.

Description

Method of automatic discrimination of an alphabet for restitution

  visual or audio of an HTML document,

  and corresponding HTML document management device.

  The invention relates to a method for automatic discrimination of an alphabet, in particular for the visual or audible reproduction of an HTML document and a device for managing HTML documents. By "restitution

  visual or audible ", we mean viewing the document

and / or its voice reading.

  The considerable boom in data exchange from

  any type, data, text or image files, through

  of telecommunications networks, in particular the INTERNET network, poses the problem of the communication of

  their users from different cultures and languages

  férentes. At the present time the documents transmitted on the INTERNET network are transmitted in the form of documents

  HTML, for Hyper Text Mark-up Language in English language

  Saxon. HTML or compatible documents, such as SGML or XML, generally consist of an HTML file and associated elements, such as for example image and / or sound files, or other HTML files. The passage from one document to another is generally done from

  of a link, element pointing to a URL address. Lasso-

  Citation of HTML files to a first HTML file (called "frame") is generally done by reference to the URL address of these files. These URLs relate to a

  site, such as a WEB site installed on this network. WEB desi-

  short for WWW (World-Wide Web) hypermedia system

in Anglo-Saxon language.

  The above documents, commonly called docu-

  HTML elements, can therefore be written, by their authors, to

  from foreign alphabets for the user having sol-

  lawful and obtained access to these documents.

  These documents are usually accessible, through any access provider, from

  any terminal interconnected to the INTERNET network and authorized

you.

  In particular the access and display of these documents

  is currently obtained: - from a microcomputer, whose keyboard and display characters are particularly adapted to the alphabet of the country, this microcomputer being equipped with a

  specific software designated by browser in English language

  Saxon, allowing to treat, alphabets answering the standard W3C, as well as documents written in a way

  specific in the official or official national language

  officially recognized from this country; - by carrying out a manual operation, prior

  or after reading such a document, which

  is generally to choose, for the machine or the termi-

  nal, the type of alphabet the latter should use; - from a HTML 4.x standard decoder and

  able to decode documents written according to the standard

  UNICODE dart or to detect the markers placed at the head

  of the document for the purpose of specifying the alphabet that was initiated

  specially used to compose the latter. Remember that the UNICODE standard is defined by the ISO / IEC 10 646 standard.

  However, some countries have put in place a process

  specific die, generally based on existing standards

  in order to allow adaptation to the native language

  official or officially recognized nal of this country. This type of process consists of recovering documents

  of existing information, by implementing lo-

  software to compose documents in this

  language. Each of these processes in fact constitutes a stan-

  coding sting and high risk, in each country, of losing

  rer. In addition there are currently a relative variety of Internet access terminals, some of these

  only allowing display operations li-

  moths. This is in particular the case with the first versions of low-cost terminals, which allow the display of an alphabet satisfying the ISO 8859-1 standard, alphabet by default, but not accepting display at all.

  UNICODE standard or even other ISO definitions of alpha-

various bets.

  The object of the present invention is to remedy the

  disadvantages linked to the diversity of the above situations

  ted, suffered by users of the INTERNET network, in order to allow, in particular, visual or audio reproduction

  HTML documents in their original language and pre-

initial feeling.

  Another object of the present invention is also

  ment the implementation of a process of self-discrimination

  subject of an alphabet for viewing a document

  HTML to allow, for current and future terminals

  turs accepting the UNICODE standard, the conversion of these

  HTML documents with the aforementioned UNICODE display standard.

  Another object of the present invention is, also

  The implementation of a discrimination process

  automatic alphabet for viewing a document

  HTML, by conversion for viewing all or part of the HTML document considered in the form of cells

images in a specific format.

  Another object of the present invention is, also

  The implementation of a discrimination process

  automatic alphabet for viewing a document

  HTML, by conversion for viewing all or part of the HTML document considered in the form of fonts

  characters resident and / or downloadable in the ter-

mineral.

  Another object of the present invention is, also

  The implementation of a discrimination process

  automatic of an alphabet, and more precisely of a lan-

  gue, to perform voice playback of all or part of the

HTML document considered.

  Another object of the present invention is, also

  The implementation of a discrimination process

  automatic alphabet for viewing a document

  HTML ment, by conversion for visualization of all or part of the HTML document considered in the form of DRCS characters, for Dynamical Redefinition Character Set in English language, directly usable by the terminals of

VIDEOTEX type.

  Another object of the present invention is, also

  The implementation of a discrimination process

  automatic of an alphabet, and more precisely of a lan-

  gue, to carry out the restitution of all or part of a

  HTML document in the form of an AUDIOTEL type service.

  The process of automatic discrimination of an al-

  phabet for visual or audio reproduction of a document

  HTML in its original language and its initial presentation, this document being delivered by a WEB site following the selection, for example, of a determined URL address,

  in accordance with the object of the present invention, is remarkable

  ble in that it comprises at least, for each current HTML document, the steps consisting in discriminating in the URL addresses the constitutive suffix of this URL address, and following the discrimination or nondiscrimination of

  this suffix, to order the receipt of this document

  rant, and to determine, relative to the current document, at least one reference alphabet as a function of header data of documents descriptive of the alphabet used,

  and on positive response to this step consisting in determining

  mine a reference alphabet, to assign to this document

  running the aforementioned determined reference alphabet.

  The present invention also relates to the

  implementation, in an interfacing and adaptation system

  between an access terminal to an interconnected server

  to this server through a network or in an interface system

  gateway type connection between a VIDEOTEX type terminal

  and a server interconnected to this pass-type system

  through a network subject to a proto

  IP type intercommunication and HTML document communication school, automatic management system

  of the alphabet chosen for visual or social restitution

of these HTML documents.

  It will be better understood on reading the

  description and observation of the drawings below in the

  which: - the figure represents, by way of illustration, a

  flowchart relating to the process which is the subject of this information

  vention allowing a recognition operation of

  the alphabet used by a current HTML document and allo-

  cation to this document of a recognized alphabet, the alphabet re-

  known being explicitly defined, other than a default alphabet, or corresponding to UNICODE coding; - Figure lb shows, in conjunction with Figure

  the, by way of illustration, an organization chart relating to the process

  object of the present invention allowing an operation of allocating a default alphabet respectively an estimated defined alphabet, called reference alphabet, when no alphabet used by the current HTML document is recognized, the estimated defined alphabet being determined by a specific estimation procedure; - Figure lc illustrates, in an illustrative way, the principle of a masking process applied to recognized words of different languages, on a graduated diagram

  in number of words for a plurality of languages considered

  rees; - Figure 2a shows a block diagram of an interface and adaptation system between a terminal

  access to a server interconnected with this server by a

  INTERNET type bucket equipped with an HTML document analysis and transformation module, in which is installed an automatic alphabet management device chosen for the visual or audio reproduction of these documents

  HTML, in accordance with the process which is the subject of this invention

tion;

  - Figure 2b shows a general flowchart

  relative to a procedure given by way of nonlimited example

  mitative of the automatic management system implanted in

  the analysis and transformation module of the information system

  facing and adaptation shown in Figure 2a; FIG. 3a represents a functional diagram of a gateway type interface system between a VIDEOTEX type terminal and a server interconnected to this gateway type system by means of a subject network.

  to an IP type intercom and communication protocol

  munication of HTML documents, in which is installed an automatic management system of the alphabet chosen for the visual or audio reproduction of these documents

  HTML, in accordance with the process which is the subject of this invention

tion;

  - Figure 3b shows a general flowchart

  relative to a procedure given by way of nonlimited example

  mitative of the automatic management system implanted in

  the gateway to the interfacing and adaptation system

  shown in Figure 3a; FIG. 4a represents the visualization, in France, of an HTML document delivered by a WEB site located in Greece in initial presentation established from the Greek alphabet, in the absence of any implementation of the process object of the present invention; FIG. 4b represents the visualization, in France, of the HTML document represented in FIG. 4a, during

  the implementation of the process which is the subject of this invention

  tion, the characters of the original Greek alphabet being restored; - Figure 5a shows the display, in France, of an HTML document issued by a WEB site located

  in Japan in initial presentation based on ideo-

  Japanese grams, in the absence of any implementation of the process which is the subject of the present invention; FIG. 5b represents the visualization, in France, of the HTML document represented in FIG. 5a, during

  the implementation of the process which is the subject of this invention

  tion, the original ideograms being restored in the form of image cells displayed on a terminal of the type

VIDEOTEX.

  A more detailed description of the discrimination process

  automatic mination of an alphabet or a language for the visual or audio reproduction of an HTML document in accordance with the object of the present invention will now be given in conjunction with FIGS. In general, we recall that a document

  HTML is transmitted in its original language and its presentation

  initial tion from a WEB site delivering the latter from a determined URL address. The selection of a

  URL address thus makes it possible to obtain the HTML documents available

  available at the aforementioned WEB site.

  As shown in FIG. 1a, when the user on his terminal selects a URL address, the method which is the subject of the present invention, with reference to FIG. 1a, for this current document delivered by the requested URL address, consists to implement a standard alphabet function consisting of at least one step (a) in discriminating in the aforementioned URL address the constituent suffix of this URL address. It should be remembered that URL addresses usually include a character string whose extension symbolized by a period and a suffix is significant for the country in which the WEB site

  accessed and interrogated is located and therefore of the original language

  gine in which the corresponding HTML document was normally drawn up. Thus we recall that the current extensions are .gr, for Greece, .ru, for Russia, fr, for France, .ch for China, .jp, for Japan

And so on.

  This discrimination step is followed for example by a step of memorizing a logical value of suf-

  fixed denoted VS to which the value 1 is assigned when

  that a suffix has been recognized and to which is on the contrary

  assigned the value 0 when no suffix has been recognized

  bare. In any event, the logical value assigned to

  the variable VS is stored for later use.

  The process which is the subject of the present invention consists, following discrimination or non-discrimination of the

  above suffix and the allocation of the value 0 or the value

  their 1 to the logic variable VS to order in a step (b) the actual reception of the current document

  i.e. admission for visual or visual restitution

  nore at the terminal of the latter, the instruction of

  however, visual or audio reproduction is not called

  for the visual or audio reproduction of the document

current.

  Step (b) to order receipt of this document

  current is then followed by a step (c) consisting of

  complete for this current document at least one reference alphabet that can be allocated to the current document

considered.

  In a particularly advantageous manner, it is indicated

  that the reference alphabet can consist of alpha-

  bet explicitly defined in the header and / or in the information document relating to the current HTML document and / or where appropriate in recognition, for the document

  standard, of a UNICODE type coding, as well as

  tères and / or sequences private or corresponding to ISO alphabets adopted as standard in the country of origin, from which the current HTML document considered emanates. We recall

  that the acronym ISO indicates the Institute for Standard Organi-

  zation in Anglo-Saxon language. By way of nonlimiting example, it is indicated that in HTML documents established in Japan or at least "Japanese" the code sequence lB 24 42 indicates that the characters which follow this sequence relate to ideograms defined in the JIS game while the sequence

  of code lB 28 42 introduces the actual HTML coding.

  Upon a positive response to the aforementioned step (c), i.e.

  say when recognizing one of the pre-

  cited as a reference alphabet, the method which is the subject of the present invention then consists in assigning, in step (d), to the current document the aforementioned reference alphabet

  that is to say one of the recognized alphabets, particular alphabet

  link or default. In step (d) above, the allocation to the current document of the reference alphabet is noted

VAL PC = VALR.

  In addition, without limitation, at the same stage

  (d), and for a notion of completeness of the iterative character

  tif of the process which is the subject of the present invention relating to

  any successive HTML document, to a representative variable

  tive from the alphabet of the previous HTML document noted

  VALPP, which is of course none other than the document

  rant current, is also allocated the aforementioned reference alphabet this operation being noted in step (d) of the figure the VAL PP = VAL R. The step (d) as such is then followed by a

  end of procedure stage since for the restitution vi-

  soot or sound of the aforementioned current document and the case

  any subsequent document from that same document.

  current cument constituting the previous document of the latter has been assigned a reference alphabet consisting of one of the alphabets, recognized in step (c). In general, it is indicated that the end of procedure step can consist of a call to a document processing function to ensure a layout of the current document.

  above and finally ensure a display and / or a read-

voice ture of the latter.

  The process which is the subject of the present invention such as

  previously written in connection with FIG. 1a allows a satisfactory visual reproduction for the majority of modern terminals and browsers meeting the standard

HTML 4.x.

  However, when an alphabet in the ISO sense of the term is common to several languages, such as the default alphabet ISO LATIN 1 (ISO0-8859-1), an advantageous but non-limiting way of allowing more definition specifies the language to be used, in particular for the sound reproduction of the document, will consist, upon a positive response in step (c), among others: during a step (g '), to be identified in the data to be restored a set of distinctive words vis-à-vis a plurality of reduced dictionaries of preset words not common to several original languages and memorize this comparison and identification operation in the form of lists of words classified by number of words detected by original language and by total number of words, to establish a

  plurality of classification variables for the number of words

  tect by original language and by total number of words;

  during a step (h '), to submit the data to res-

  to replace the current document with a test for the existence or non-existence of an end of document of the current document, and,

  on positive response to this test of existence or nonexistence

  tence of an end of document, (i ') allocate to the current document a reference language chosen as a function of the value of at least one classification variable of the number of words detected by original language and by total number of words. However, and due to the diversity of materials

  connected to this network previously mentioned in the

  cription, when with reference to step (c) of determining

  tion of at least one reference alphabet no alphabet defined explicitly in the header and / or in the information document relating to the HTML document is detected or when the UNICODE encoding of the document is not detected or that private sequences or corresponding to an alphabet

  ISO other than the default, are also not de-

  the response to step (c) of determination is

then negative.

  In such a case and in accordance with an embodiment

  particular non-limiting indication of the process which is the subject of the pre-

  sente invention, this method further consists in a step (e) to allocate to the current document an alphabet by

  default. This operation is noted VALPC = AD, AD desi-

  gaining the default alphabet in figure la in step (e). Of course, the allocation of a default alphabet in step (e) is not an end in itself but constitutes, in fact, a minimal display solution in order to ensure

  display and / or voice reading of the HTML document

rant corresponding.

  Consequently, as shown in the figure la

  step (e) can then be followed by step (f) consisting of

  both to submit the data to be returned from the HTML document

  current to a consistent filtering process, e.g.

  ple, to discriminate the percentage of code values of these data to be restored, this percentage of code values being noted N. code values noted CODE comprised between a first and a second limit value to establish

  a descriptive percentage code value of these data

  to be returned for the current document considered. We

  of course understands that depending on the value of percent

  tage N above the process which is the subject of the present invention makes it possible to establish and discriminate an alphabet other than

  the aforementioned default alphabet, but likely to present

  ter a better adaptation to the representation and the af-

  file and / or voice reading of the current HTML document

  in its original language version.

  In a preferred, non-limiting embodiment,

  tif we indicate that the first and second values li-

  mites can be taken equal to the value A0 and the value FF respectively, these values being expressed in this case in hexadecimal code. The percentage of the values of

  code between the aforementioned limit values is no-

tee N, N being expressed in%.

  The aforementioned step (f) is then itself followed by a

  step (g) of finding an alphabet or a language by

  particular in the current HTML document, this step of re-

  looking for a particular alphabet language or

  consisting in identifying and locating in the data to be

* titrate above a set of distinctive words, by

  comparison and identification of these distinctive words vis-à-vis

  vis a plurality of reduced dictionaries of pre-

  established, not common to several languages of origin.

  The operation carried out in the aforementioned step (g) can be carried out so as to advantageously obtain a classification by number of different words detected by language and by

total number of words.

  It is understood in particular that the above-mentioned operating mode in step (g) represented in the figure then allows it

  to establish a plurality of name classification variables-

  number of words detected by original language, i.e.

  actually by alphabet allowing to express such a lan-

  original word, and of course by total number of words

  in the original language considered.

  In general, we indicate that the variables

  can be organized into successive files

  as shown schematically in the table below: Alphabet Ai Alphabet A2 Alphabet AN-1 Alphabet AN NBMotdifferent NB Different word NB_Mot different NBMotdifferent Alphali Alpha2 AlphaN-1 AlphaN TotalMot_ 1 A1 TotalMot_ 1 A2 Total Mot_ 1 AN-1 TotalMot 1 AN Total Word 2 Ai Total Word 2 A2 ITotal Word 2 AN-1 Total Word 2 AN Total_Mot 3 A1 Total_Mot_3 A2 TotalMot_3_AN-1 Total_Mot 3 AN TotalMotN-lA1 TotalMotN-lA2 TotalMotN-1iAN-1 TotalMotN-lAN Total_Mot NA1 [TotalMotN A2 Total Word N AN-1 Total Word N AN

  In the above table, it is indicated that each dic-

  reduced tionnaire can include for example between five and ten words for each language or support alphabet of each

  language considered. By way of nonlimiting example, we in-

  states that the words in the reduced dictionary of each language are words with a high frequency of repetition, specific to each language considered, such as articles, personal pronouns, possessive adjectives, either for French le, la, les, je, tu, il, mon, ton, son, for English the, my, you etc. In the above table the different alphabets are

  noted alphabet A1 to alphabet An the number of different words

  rents are noted Total word 1 A1 for the first word of

  the alphabet A1, Totalmot_1 A2 for the first word of the al-

  phabet A2 and so on. The whole is then classified

  as shown in the above table.

  We understand of course that step (g) of searching for a particular alphabet or language by analysis and

  comparison of words against words in dictionaries

  reduced res is performed for each word of the document

  rant considered. For this reason, after each word of the latter the data to be restored are subjected to a step (h) consisting in subjecting the data to be restored of this current document to a test of existence or non-existence of an end of document. of the last. This test can be simply carried out by detecting an end-of-document code for the HTML document considered, or by detecting the complete reception of the document, or upon reception

  of part of the document deemed sufficient to terminate

  from the process of the invention, or by detection

an end of file code.

  On a negative response to the end-of-document existence test of step (h), the process is continued for the next word by returning to step (b) receiving the document for example. It is in fact understood that the above-mentioned document reception step (b) can be carried out relative to the

  successive word constituting the current HTML document considered

  déré. On the contrary, on a positive response to the test of step (h) previously mentioned, the method which is the subject of the present invention consists in allocating to the current document a

  reference alphabet chosen according to the value of-

  code percentage code of this data to resti-

  kill, the aforementioned value N, for the current document

  considered and the value of at least one class variable

  the number of words detected by original language and by total number of words. We conceive for example as

  non-limiting example that a procedure allowing

  rent a corresponding reference alphabet from the current document

  corresponding to that chosen in step (g) can consist in taking for example as a reference alphabet the one which

  is the sum of the total number of words in the diction-

  corresponding reduced naire with the highest value. This simple embodiment is not necessarily the most suitable and it is preferable, in particular, to subject the aforementioned process, as shown in FIG. 1a in a step (i), to an additional test consisting of a comparison of the descriptive value of

  percentage of data code to be restored, for the current document considered, ie the aforementioned value N, to a threshold value.

  This comparison test, noted under these conditions in FIG. 1a, can for example consist in comparing the value N of the aforementioned percentage with a value

  of reference threshold which can be taken equal to 50%.

  The test carried out in step (i) and in particular (it) can then be advantageously followed by a step of memorizing a logic variable C2 representing the

  condition 2 above, with this logical variable being at-

  tributed the value 1 on positive response to the test it precedes

  Also mentioned when a negative response to the above test is assigned a value 0 to the logic variable C2. The previously mentioned memorization steps can then be followed either by the choice of the sum of the number of majority words previously mentioned, or by a much more elaborate process which will be described in

  link with Figure lb below.

  With reference to FIG. 1b previously cited, the test (i) of FIG. 1a, following the storage of the logic variable C2, can understand and be followed, as shown in FIG. 1b of at least one test (j ) recognition of discrimination or non-discrimination of the aforementioned suffix, this test can then be simply carried out by calling the logic variable VS

  previously mentioned in description and comparison

  from the value of this logic variable to the value 1. The

  suffix having been memorized previously, it is then pos-

  likely to continue the above step (i) in the mode of

  preferential realization below.

  On positive response to the test (j) previously men-

  tioned, a test (k) is planned consisting in carrying out a

  first criterion for choosing a specific alphabet respec-

  of the default alphabet allocated in step (e) above. This first criterion of choice may consist of

  example in a comparison of the logical variable C2 related

  tive at condition 2 to the value 1. On positive response to the above test (k), and therefore to the first selection criterion,

  a step (k2) is introduced consisting in allocating to the do-

  Current cument a specific alphabet as a reference alphabet depending on the type of suffix recognized. This operation is noted VAL_PC, variable relating to the alphabet of the current document, = VALR the recognized alphabet being taken as reference alphabet. In the same way as in step (d), this specific alphabet in one step (kl) is also

  allocated to the document preceding the current document

  laughing, that is to say the current document. This operation

  ration is noted in the same way as in the case of

step (d) VALPP = VAL_R.

  On the contrary, when the test (k) on the variable lo-

  gic C2 is not satisfied, and therefore on a negative response to the first choice criterion 1, the current document is

  then allocated the default alphabet as defined above

  Use step (e) as the reference alphabet.

  Step (1) is noted default display in Figure lb. In the same way as with step (kl) previously described, this default alphabet can then

  be allocated to the document preceding the current document ul-

  térieur. On negative response to the above test (j), the variable

  VS logic being different from the value 1 with the con-

  previously defined, a second criterion of

  choice of a specific alphabet respectively of alpha-

  default bet is then introduced. This second choice criterion carries the reference m and consists of a test of the logical variable C2 relating to condition 2 previously defined in step (ii) at the value 1. On a negative response to the second choice criterion m above, the method object of the present invention is to perform a display

  and / or a default voice reading, i.e. the display

  chage and / or default voice playback as defined

  nor (s) in step (1) previously described in the

  description. Thus in the above-mentioned step (1), in the document

  current is allocated as the reference alphabet, the alphabet

  by default. It is indicated that the step (1) of display by

  is then followed by a semi-procedural end step

  similar to that described in connection with step

  (d) previously mentioned in the description. It is

  similarly for steps (kl) and (k2) previously described

your.

  In the same way as on a negative response to the first

  mier choice criterion k, the negative response to the second choice criterion m implies the call of step (1) and in

  particular the allocation of the alphabet by default to the docu-

  preceding the subsequent current document in con-

similar editions.

  On a positive response to the second selection criterion m, the method which is the subject of the present invention then consists in performing a third test criterion (n) for allocation to the document preceding the current document with a reference alphabet, that is to say -to say of an alphabet other than the alphabet

by default.

  On positive response to the third test criterion n above, an operation to allocate to the current document the reference alphabet allocated to the document preceding the

  current document is performed in step (o), this operation

  tion being noted accordingly VAL PC = VALPP. The stage

  (o) above is then followed by the end of process step.

  duration previously mentioned in the description.

  On the contrary, on a negative response to step (n) of the third test criterion, a fourth test criterion p is then carried out, this fourth test criterion consisting

  actually in a recognition test of at least one alpha-

  bet to determine a recognized alphabet. By way of nonlimiting example, it is indicated that the fourth test criterion p may consist in carrying out a comparison pl of words characteristic of one or more alphabets with a predefined decision mask such as a

OR logical for example.

  With reference to FIG. 1c, it is indicated that the function

  masking can be introduced when there is ambiguity in the number of words recognized in relation to two languages, languages L1 and L4 for example, as well as

  relative to their distance from other lan-

  gues such as languages L2 and L3 when these distances

  are not considered significant.

  The step (pl) of comparing words can then be followed by a step (p2) consisting in defining the existence or non-existence of an alphabet. This criterion consists simply in applying to the logical variables obtained by applying the logical OR of masking a comparison

  at the value 1 of this logic variable.

  Upon a positive response to the criterion of existence of an al-

  phabet p2 and therefore on positive response to the fourth cry-

  test mother p the process which is the subject of the present invention

  then consists of a step (q) to allocate to the document

  rant the recognized alphabet, this recognized alphabet being taken as the reference alphabet, this operation is denoted VAL PC = alphabet recognized in the aforementioned test q. In the same way as in the case of operations k2 and kl or in step (d) this recognized alphabet is also allocated to the

  document preceding the subsequent current document, i.e.

  tell the current document. This operation is noted in the same way VALPP = recognized alphabet. Step (q) is of course followed by the step at the end of the previous procedure.

  as mentioned in the description.

  On the contrary, on a negative response to step (p2) of the alphabet criterion, that is to say, on a negative response to the fourth test criterion p of recognition of at least

  an alphabet, no alphabet having been recognized

  formally to the implementation of the process object of the pre-

  invention, step (r) is then carried out

  consisting in allowing a removal of doubt as to the

  knowledge of the alphabet by the user. This step then consists in making a manual choice of an alphabet

  by the latter. This alphabet constitutes a preci-

  sé, which is then allocated to the current document, and, in the same way as with steps (q), (kl), (k2)

  and (d) the document preceding the subsequent current document.

  laughing. Step (r) is then followed by an end step of

  procedure as described previously in the description

  tion. Regarding the initialization of the object process

  of the present invention, that is to say when receiving

  tion of the first HTML document, this initialization step

  tion can consist in choosing as alphabet by default an alphabet such as the ISO Latin alphabet for example, this type

  of alphabet corresponding to the American ASCII alphabet above

  which European accented characters are added.

  In addition, it is indicated that according to a non-limiting implementation variant of the method which is the subject of the present invention, step a) consisting in discriminating in the URL address the suffix constituting the latter may not be executed prior to the 'step b). In

  Indeed, the aforementioned step a) can, for example, be executed.

  ted before step j) shown in figure lb, following the execution of step i) shown in figure

  without departing from the scope of the subject of this invention

  tion, the URL address having been stored.

  A specific embodiment of a

  automatic management of the alphabet chosen for the restitution

  visual or audio tion of HTML documents implemented in an interface and adaptation system between an access terminal to a server interconnected by a network subject to an IP type intercom protocol for INTERNET PROTOCOL and HTML communication of documents

  mentioned above, will now be described in connection with the fig-

res 2a and 2b.

  Generally speaking, it is recalled that this type of interfacing and adaptation system includes an HTML analysis and transformation module denoted Ml, an anticipation module denoted M2, an online conversion module

  of documents noted M3 and a cache memory module of

  converted cuments noted M4. All of these elements pre-

  cited is connected via a system called

  "proxy" this proxy system, on INTERNET networks, serves

  front of adaptation interface between the terminal and a service

  veur. The terminal and the server are represented and designated as such in FIG. 2a. the set made up of the aforementioned elements, namely the HTML analysis and transformation module Ml, the anticipation mechanism M2,

  the M3 document conversion module and the mo-

  cache of converted M4 documents, associated

  to the proxy system is called a conjugate. A conjugate per-

  puts, in fact, to ensure the transfer, with or without conver-

  sion, documents delivered by the server to the terminal in HTML format. As shown diagrammatically in FIG. 2a, the device for automatic management of the alphabet chosen for visual or audio reproduction

  of these HTML documents, is implemented in the analysis module

  lysis 1 and HTML transformation and for that reason

  carries the reference G in this module.

  The management device G naturally includes modules for implementing the method as previously described in connection with FIGS. 1a and 1b allowing,

  under the conditions described above in the description

  tion, to assign to each HTML document running an alpha-

  reference bet to allow the visual or audible reproduction of this current document in its original language

  and its initial version. This program implementation module

  previously described gives the reference 1 on the figure

gure 2b.

  In addition, the management device G which is the subject of the pre-

  sente invention includes a module 2 for discriminating the presence or absence of coding

  of the current document according to a UNICODE alphabet.

  As shown in FIG. 2b above, the discrimination module 2 can comprise a module 20 for discriminating the presence of characters other than those of the default alphabet, the alphabet ISO-8859-1 by

  example of the UNICODE alphabet or another alphabet as well

  if a module 21 activated on positive response to the

  dule 20, the module 21 allowing to discriminate the pre-

  sence of a UNICODE coding of the current document or character

  ters coded in UNICODE mode, other than those in the default alphabet with reference to Figure 2b. It is indicated that the module 2 of discrimination of the presence respectively

  the lack of coding of the current document according to alpha-

  bet UNICODE corresponds to a negative response to module 20 while the presence of coding of the current document according to

  the UNICODE alphabet corresponds to the positive response of the

  dule 20 and the positive response from module 21, presence of

  UNICODE coding, an intermediate case corresponding all-

  times to the absence of UNICODE coding of the document although certain UNICODE characters are present in response to the

module 20.

  In addition, a recognition module 3 for the compatibility

  terminal flexibility with UNICODE coding is expected, this

  module responding only to the positive response respec-

  negative of module 21 previously mentioned. Terminal compatibility recognition module 3

  with UNICODE coding can include a module 30 of allo-

  cation, upon negative response from module 21, of a recognized alphabet to the current document, this recognized alphabet being distinct from the default alphabet, the current document

  may however contain in addition to characters of the al-

  default phabet. Compass recognition module 3

  the UNICODE coding capability of the terminal also includes a terminal recognition module itself

UNICODE 31, 32.

  Module 3 for recognizing the compatibility of

  0 terminal with UNICODE coding allows, as well as

  felt in FIG. 2b, the emission and the restitution vi-

  audible or audible from the current document to the terminal at

  step A when the reference alphabet allocated to this do-

  current cument is the default alphabet or that this alpha-

  reference bet is the UNICODE alphabet and that the compatibility character of the terminal with coding

  UNICODE is recognized. Indeed, in such a case, a termi-

  nal UNICODE is able to display the current HTML document without change or intervention. Step A is followed

  of course an end of procedure step.

  Module 3, associated with module 2 previously described, also allows the transformation of the current document into image cell files in step B when the reference alphabet is the UNICODE alphabet on positive response from module 21 but the character of compatibility of the terminal with UNICODE coding is not recognized, negative response from module 31, or when the alphabet of the current document is not the UNICODE alphabet on negative response

  in module 21 and that the compatibility character of the

  minimum with UNICODE encoding is not recognized, on re-

  negative response from module 32. Step B in fact constitutes a step of transforming the initial HTML document into an HTML document in which the alphabetic characters

  and / or the ideograms are replaced by ima-

  ges which then make it possible to represent these. Of

  advantageously these image cells are then delimited

  in JPEG, GIF or other formats for example. We understand

  takes in particular that these image cells can be combined or grouped to represent one or more

ideograms in a single image.

  In addition, module 3 associated with the aforementioned module 2 per-

  still puts the transformation of the current document by co-

  dage using the UNICODE alphabet in step C when the reference alphabet is not the UNICODE alphabet on a negative response from module 21 but the character of

  compatibility of the terminal with UNICODE coding is re-

  known, that is to say on a positive response from module 32. Steps B and C are then followed by step D allowing the transmission of the transformed document to the terminal as

  than a MINITEL-INTERNET, a microcomputer or other.

  Step D is itself followed by an end of pro-

cedure.

  In the same way an automatic management device

  that the alphabet chosen for the visual or audio reproduction of HTML documents can be implemented in an interface system of the gateway type between a terminal

  VIDEOTEX type interconnected by means of the

  PSTN switched telephone bucket and a dedicated link LT of a PAVI access system and an X.25 type line

  connected to this gateway, the gateway being itself

  even interconnected by a network such as the INTERNET network to different servers. The network is of course subject to an intercom protocol of the IP type and communication of HTML documents. In such a case, the management device object of the present invention is

  installed in the gateway constituting an adapta-

  INTERNET / TELETEL type for the VIDEOTEX terminal

for example.

  The management system that is the subject of this invention

  tion is then integrated and installed in the gateway and

  therefore bears the reference G 'in FIG. 3a.

  With reference to FIG. 3b, it comprises, in the same way as in relation to FIG. 2b, a module 1 for implementing the method which is the subject of the present invention allowing

  to assign to the current HTML document a reference alphabet

  to allow the display of this current document in its original language and its initial version. In addition, the management device comprises a module 4 for discriminating

  nation of the presence respectively of the absence of co-

  Dage of the current document according to a reference alphabet corresponding to a default alphabet, such as the ISO-8859-1 alphabet, as well as a 4o module allowing, upon response

  positive from module 4 to the presence of the alphabet by de-

  above, the transformation of the current document into service in VIDEOTEX format from the basic character sets of the above-mentioned VIDEOTEX terminal. The 4o module of

  transformation of the current document then allows the affi-

* chage the HTML document thus transformed into a service

  VIDEOTEX by returning to an end of procedure step.

  In addition, a module 5 on discrimination, in the docu-

  common, the presence of ideograms and / or characters

  ters coded according to the UNICODE code corresponding to

  ideograms is expected. This module 5 is actuated on re-

  negative response from module 4 previously mentioned.

  A 5o module is planned to ensure the transformation

  of the current document in service in VIDEOTEX format on re-

  negative response from module 5. This transformation is ef-

  made from basic terminal character sets

  VIDEOTEX and dynamic redefinable character sets

  only DRCS according to the reference alphabet al-

  rented, this allows a display limited in number

of ideograms of the current document.

  A module 6 is also provided, which makes it possible to

  discriminating the capacity of the VIDEOTEX terminal

  aforementioned to display image files. On posi-

  tive of module 5 of ideogram presence discrimination-

  mes and / or characters coded according to the UNICODE code corresponding to ideograms, and on positive response to the aforementioned module 6, of discrimination of the terminal's ability to view and display image files a

  transformation module 60 ensures the transformation

  mation of the current document in VIDEOTEX service from basic character sets of the VIDEOTEX terminal and sets

  dynamically redefinable characters DRCS in function

  tion of the allocated reference alphabet, as well as by

  the display of ideograms recognized in the form of cells-

  images. The 6o module uses the basic character sets of VIDEOTEX terminals, such as MINITEL,

  DRCS character sets downloaded from a number of ideo-

  recognized grams limited in number to 47 per document. All or part of the alphabetic characters and / or ideograms

  can be replaced by image cells.

  feeling in JPEG and / or GIF format for example.

  Finally, on negative response to module 6 a module 61

  allows to transform the HTML document

  rant in VIDEOTEX service using character sets

  of MINITEL, the DRCS character sets previously

  ment downloaded in the MINITEL according to the recognized alphabet as well as in the case of the 5o module. However, a display in a limited number of ideograms, 47 per document,

  is also planned and implemented by module 61.

  Figure 4a shows the example of an HTML document

  displayed on the monitor of a terminal in France, this docu-

  ment HTML being delivered by a WEB site located in Greece for example. FIG. 4a represents such an HTML document in the absence of application of the method which is the subject of the present invention. We see that all of the text displayed

  consists of a succession of characters substantially

  intelligible due to the type of alphabet used to display the HTML document considered. In fact,

  no reason why the codes relating to each card

  the alphabet of the Greek alphabet make it possible to obtain a coherent code from the corresponding codes used in France. On the contrary in FIG. 4b there is shown an HTML document delivered by the same WEB site located in Greece and in which, thanks to the implementation of the process which is the subject of the present invention, all the characters displayed have been reconstructed from the original Greek alphabet, the visualization of the document being restored in its original version.

  The same is true with regard to FIG. 5a.

  later to an HTML document delivered by a WEB site if-

  killed in Japan. The display of such a document from an alphabet used in France also delivers a series of characters totally unintelligible in the absence of the

  implementation of the process which is the subject of the present invention.

  On the contrary in FIG. 5b, an HTML document has been represented from the same WEB site located in Japan in which, however, thanks to the implementation of the process which is the subject of the present invention, the corresponding codes have been

  transformed to represent image files

  presentation of significant ideograms for any person

  sounds having knowledge of ideograms

cited above.

  A method of self-discrimination has thus been described.

  matic of an alphabet for visual or social restitution

  of a particularly efficient HTML document in

  measure o due to the specific implantation of mo-

  dules for implementing the method in the form of corresponding management modules according to the terminals used and in particular the mode of connection of these terminals via the INTERNET network to different servers, it is thus possible to restore any HTML document in its original version and in particular in its original language.

  The method of automatic discrimination of an alpha-

  bet for the visual or audio reproduction of a document

  HTML can be implemented, as examples not limited to

  tifs, at the file server level, in a pass

  between this server and the consultation terminal, or

in this terminal.

Claims (8)

  1. Method of automatic discrimination of an al-   phabet, in particular for visual or sound reproduction   an HTML document in its original language and its presentation   initial tation, this document being issued by a WEB site   obtained from a determined URL address, process   bearing at least, for this current HTML document, the pes consisting of:   a) discriminate in said URL address the suffix   guardian of this URL address; and following the discrimina-   tion or non-discrimination of said suffix, b) order the reception of said current document; and c) determining, with respect to said current document, at least one reference alphabet as a function of data   of descriptive documents from the alphabet used   read; and, upon a positive response from this step,   both to determine, relative to the current document, a reference alphabet, d) assign to said current document the reference alphabet   determined in step c) above.   2. Method according to claim 1, characterized   in that, on a negative response to said step c) consis-   as long as to determine a reference alphabet relative to said current document, said method also consists in: e) allocating a default alphabet to said current document; and   f) submit the data to be returned from said document   rant to a filtering process consisting in discriminating   ner the percentage of the code values of these data to be restored comprised between a first and a second limit values, in order to establish a descriptive percentage value of code of these data to be restored for the current document considered;   g) identify in said data to be returned a set   ble of distinctive words, by comparison and identification   tion of these distinctive words with respect to a plurality of reduced dictionaries of non-common preset words   to several original languages and memorize this operation-   comparison and identification in the form of lists of words classified by number of words detected by   original language and by total number of words, for   blur a plurality of number ranking variables   of words detected by original language and by total number tal of words;   h) submit the data to be returned from said document   rant to a test of existence or non-existence of an end of document of said current document, and, on positive response to this test of existence or non existence of an end of document, i) allocate to said current document a reference alphabet   chosen according to the descriptive value of percent   tage of code of these data to be returned for this current document and of the value of at least one variable of   classification of the number of words detected by original language gine and by total number of words.   3. Method according to claim 1, characterized   in that, upon a positive response to said step c) consis-   as long as a reference alphabet is to be determined with respect to said current document, said method also consists in: g ') identifying in said data to restore a   set of distinctive words, by comparison and iden-   tification of these distinctive words vis-à-vis a   reality of reduced dictionaries of pre-established words not common to several languages of origin and memorize this comparison and identification operation under   form of word lists classified by number of words   tect by original language and by total number of   words, to establish a plurality of class variables-   the number of words detected by original language and by total number of words; h ') submit the data to be returned from said current document to a test for the existence or non-existence of an end of document of said current document, and, on positive response to this test for existence or non-existence of an end of document,   i ') allocate to said current document a reference language   rence chosen according to the value of at least one classification variable of the number of words detected by   original language and by total number of words.   4. Method according to claim 2, characterized   in that said step i), consisting in allocating an alpha-   reference bet to said current document, includes at least   a descriptive value comparison test of pros   centering of data code to restore for the document   current considered at a threshold value.   5. Method according to claim 2 or 4, character-   laughed at in that step i) consisting in allocating said docu-   a current reference alphabet comprises at least: j) a test for recognizing discrimination or non-discrimination of said suffix, and, on a positive response to said test: k) a first criterion for choosing a specific alphabet,   respectively the default alphabet, and, on re-   positive response to the first choice criterion, k2) al-   rental to the current document of a specific alphabet   as a reference alphabet depending on the type of suf-   recognized fixed, allocation k1) of this alphabet specific to the document preceding said subsequent current document, and, on negative response to said first selection criterion, 1) allocation to the current document of the alphabet by default as the reference alphabet, and allocation of this default alphabet for said document preceding the subsequent current document, and, on negative response to said test j, m) a second criterion for choosing a specific alphabet respectively of the default alphabet and, on negative response for said second choice criterion, 1) allowance   to the current default alphabet document as al-   reference phabet, and allocation of this alphabet by   defect in the document preceding the current document   and, on positive response to said second cry- mother of choice,   n) a third test criterion for allocation to said document   preceding the current document with a alphabet of re-   reference, and, on positive response to said third test criterion, o) allocation to said current document   of the reference alphabet allocated to said previous document   in the current document, and, if the answer is negative,   said third test criterion, p) a fourth recognition test criterion of at least one alphabet to determine a recognized alphabet,   by comparison of characteristic words of one or more   several alphabets with a decision mask, and, on   positive response to the fourth test criterion, q) alloca-   tion to said current document of the recognized alphabet, and   allocation of this alphabet recognized in said pre-   assigning the subsequent current document, and, in the event of a negative response to the fourth test criterion, in the absence of a recognized alphabet, r) manual allocation by a user of an alphabet specified in said current document, and allocation of this specified alphabet audit document   preceding the subsequent current document.   6. Method of automatic discrimination of an al-   phabet, in particular for visual or sound reproduction   an HTML document in its original language and its presentation   initial tation, this document being issued by a WEB site   obtained from a determined URL address, process   bearing at least, for this current HTML document, the   pes consisting in: b) controlling the reception of said current document; and c) determining, with respect to said current document, at least one reference alphabet as a function of data   of descriptive documents from the alphabet used   read; and, upon a positive response from this step,   both to determine, relative to the current document, a reference alphabet, d) assign to said current document the reference alphabet   determined in step c) above, and on negative response   tive to said step c) of determining relative   vement to said current document a reference alphabet: e) allocate to said current document a default alphabet; and   f) submit the data to be returned from said document   rant to a filtering process consisting in discriminating   ner the percentage of the code values of these data to be restored comprised between a first and a second limit values, in order to establish a descriptive percentage value of code of these data to be restored for the current document considered;   g) identify in said data to be returned a set   ble of distinctive words, by comparison and identification   tion of these distinctive words with respect to a plurality of reduced dictionaries of non-common preset words   to several original languages and memorize this operation-   comparison and identification in the form of lists of words classified by number of words detected by   original language and by total number of words, for   blur a plurality of number ranking variables   of words detected by original language and by total number tal of words;   h) submit the data to be returned from said document   rant to a test of existence or non-existence of an end of document of said current document, and, on positive response to this test of existence or non existence of an end of document, i) allocate to said current document a reference alphabet   chosen according to the descriptive value of percent   tage of code of these data to be returned for this current document and of the value of at least one variable of   classification of the number of words detected by original language gine and by total number of words.   7. In an interfacing and adaptation system between an access terminal to a server interconnected by a network, subject to an intercom protocol of the IP type and communication of HTML documents, this system   interfacing and adaptation including an analysis module   HTML lysis and transformation, a anticipation module, an online document conversion module and a cache module for converted documents, an automatic alphabet management device chosen for the visual or audio reproduction of these current HTML documents comprising at least, in said HTML analysis and transformation module: - means for implementing the method according to   one of claims 1 to 6, said means allowing   to assign to the current document a reference alphabet,   allowing the visual or audio reproduction of this document   fluent in its original language and original version   tial; - means of discrimination of presence,   respectively the absence of coding of said document   rant according to a UNICODE alphabet; means for recognizing the compatibility of said terminal with UNICODE coding; said means of recognizing the compatibility of this terminal with UNICODE coding allowing: ò either the transmission and the visual or audio reproduction of said current document when the reference alphabet allocated to this current document is the default alphabet or when this reference alphabet is the UNICODE alphabet and the compatibility of the terminal with UNICODE coding is recognized;   * or the transformation of said current document into fi-   image cell les when this reference alphabet   rence is the UNICODE alphabet but the compatibility character of the terminal with UNICODE coding is not   not recognized or when this alphabet is not the alpha-   bet UNICODE of reference and that the character of compa-   tibility of the terminal with UNICODE coding is not recognized, then transmission and display of the current document transformed; or transformation of the current document by coding   using the UNICODE alphabet then emission and resti-   visual or audio tution of the current document transformed   even when this reference alphabet is not   the UNICODE alphabet but that the character of compatibi-   The unity of the terminal with UNICODE coding is recognized.   8. In a pass-type interface system   between a VIDEOTEX type terminal interconnected via the switched telephone network and a dedicated link to this gateway type system and a server interconnected to this gateway type system by   through a network subject to an inter-   IP type communication and HTML document communication, an automatic alphabet management device   chosen for the visual or audio reproduction of these documents   common HTML elements including at least, in said step   serelle: - means of implementing the method according to   one of claims 1 to 6, said means allowing   to assign to the current document a reference alphabet,   allowing the visual or audio reproduction of this document   fluent in its original language and original version   tial; - means of discrimination of presence,   respectively the absence of coding of said document   rant according to a reference alphabet corresponding to an al-   phabet by default, and, on positive response to the presence of this alphabet by default,   u transformation of said current document into one or more   several services in VIDEOTEX format from basic character sets of said VIDEOTEX terminal; and on re-   negative answer to the presence of this alphabet by de-   is necessary, - means of discrimination in said current document of the presence of ideograms and / or characters   coded according to the UNICODE code corresponding to ideograms-   mes, and on negative response to said discrimination of the presence of ideograms and / or characters coded according to the UNICODE code corresponding to ideograms   transformation of said current document into one or more   several services in VIDEOTEX format from basic character sets of the VIDEOTEX terminal and dynamically redefinable character sets (DRCS), and visual or audio reproduction of the corresponding current transformed document; means of discriminating the capacity of said VIDEOTEX terminal to display image files, and   on positive response to said pre-discrimination test   sence of ideograms and / or characters coded according to the UNICODE code corresponding to ideograms;   * on negative response to said discrimination of capacity   of said VIDEOTEX terminal to display ima-   ges, transformation of said current document into one or more services in VIDEOTEX format from basic character sets of the VIDEOTEX terminal and dynamically redefinable character sets (DRCS),   function of the reference alphabet allocated and returned   visual or audible tion of the document thus transformed,   visual restitution including also the   chage of ideograms of the current document, and   upon a positive response to said discrimination in capacity   t of said VIDEOTEX terminal to display image files, transformation of said current document into one or more services in VIDEOTEX format from basic character sets of the VIDEOTEX terminal and dynamically redefinable character sets (DRCS), into   function of the reference alphabet allocated and returned   visual or audible tion of the document thus transformed,   visual restitution including also the   chage of ideograms recognized in the form of image cells.