Name card information Chinese to English automatic translation method based on domain ontology

Name card information Chinese to English automatic translation method based on domain ontology

  • CN 101,216,819 A
  • Filed: 12/28/2007
  • Published: 07/09/2008
  • Est. Priority Date: 12/28/2007
  • Status: Active Application
First Claim
Patent Images

1. translate the English automatic translating method in the card information based on domain body, it is characterized in that, with speech material storehouse, corpus is the basis, for included information in the business card is set up field ontology library, at least comprise the address domain class in the described field ontology library, the name in an account book domain class, the organization class, wherein, described contact address class comprises at least:

  • address separator, prefix vocabulary, address trifle and address, be address separator, prefix vocabulary, address trifle and address setting attribute and property value are described address separator simultaneously, prefix vocabulary, example is set up in address trifle and address;

    Described name in an account book domain class comprises at least;

    address designation, the self-defined appellation in address, name in an account book sign and sub-name in an account book, name in an account book, for described address designation, the self-defined appellation in address, name in an account book sign and sub-name in an account book set a property and property value, set up example for described address designation, the self-defined appellation in address, name in an account book sign and sub-name in an account book simultaneously;

    The semanteme that described attribute and property value obtain according to the training of institute predicate material storehouse, corpus is provided with;

    The organization class is to collect organization and the semantic relation between setting up;

    It is the relational database that field ontology library is based on semantic association;

    Set up the grammer translation instance of word, words and phrases with English-Chinese dictionary, Chinese-English Dictionary and Chinese-English syntax rule, set up each element in translation instance and the field ontology library simultaneously and set up corresponding relation between the example;

    This method is further comprising the steps of;

    A, provide the card information inputting interface, receive contact address, postcode, contact name, position, organization, the telephone number of user'"'"'s business card, the information input of fax to the user;

    B, obtain the parameter value of contact address, postcode, contact name, position, organization, telephone number, fax, for the information of postcode, telephone number, fax, literal translate, the digital parameters of user'"'"'s input is not translated;

    For contact name, judge whether to be Chinese, if carry out cutting, search the name whether described translation ontology library has coupling by first name and last name, then be translated into corresponding English if having, otherwise be translated into phonetic, then do not translate if not Chinese;

    For position, then be directly inputted to the translation ontology library and mate, if coupling then with the translator of English of correspondence as output, then be translated as the Chinese phonetic alphabet when not matching;

    For the contact address, then the parameter value with user'"'"'s input is a querying condition, mate at the address of field ontology library domain class, name in an account book domain class, if there is occurrence, then speech is cut in contact address and organization by the speech mode of cutting in the field ontology library, search the translation instance that these cut the speech correspondence again, translate according to the English grammer translation rule of training in the example again;

    If there is not occurrence, then directly cut speech, each is cut speech be input to field ontology library and mate, if coupling then obtains the translation result that each cuts speech, then be not translated as the Chinese phonetic alphabet if match, translate according to the English grammer translation rule of training in the example;

    Then be input in the organization class for organization and mate, the match is successful then translates by the rule of setting, otherwise with the Chinese phonetic alphabet as translation result;

    The information output of C, the contact address after will translating, postcode, contact name, position, organization, telephone number, fax, and receive user'"'"'s correction, correction result is exported as translation result;

    Then the words and phrases result that the user is revised is saved in the corresponding instance of translating in the ontology library, and upgrades the translation result that should revise words and phrases in the example.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×