Database access using data field translations to find unique database entries

Claim 1

A database access apparatus comprising: (a) a database containing entries each comprising a plurality of fields which contain machine representations of items of information pertaining to the entry, the said representations forming a first vocabulary; (b) announcement means responsive to machine representations falling within a second vocabulary of such representations to generate audio signals representing spoken announcements; (c) input means operable to receive signals and to produce machine representations falling within a third vocabulary of such representations; (d) translation means defining a relationship between the first vocabulary and the second vocabulary and between the first vocabulary and the third vocabulary; and (e) control means operable (i) to generate, in accordance with the defined relationship, for each representation produced by the input means, one or more representations according to the first vocabulary; (ii) to identify database entries containing the generated representations; (iii) to examine each representation or combination of representations which is contained in a selected field or combination of fields of the identified entries to identify unique one(s) of those representations or combinations, a unique representation or combination being one which, when translated using the translation means into representations of the second vocabulary, differs from every other such unique representation or combination when similarly translated; and (iv) to control the announcement means to generate an announcement including at least one word or combination of words which correspond(s) to one of the unique representations or combinations.

Claim 11

A method of speech recognition comprising: (a) generating at least one announcement requiring a spoken response; (b) recognising the response(s); (c) identifying database entries containing fields matching the recognised responses; (d) in the event that the number of such entries exceeds a predetermined limit, generating an announcement containing at least one word corresponding to a selected field of an identified entry for a positive or negative response; (e) upon receipt of a positive response, identifying database entries which contain fields matching the recognised responses and whose selected fields match the said word; and (f) repeating steps (d) and (e) at least once.

Claim 12

A method of speech recognition comprising: (a) generating at least one announcement requiring a spoken response; (b) recognising the response(s); (c) identifying database entries containing fields matching the recognised responses; (d) in the event that the number of such entries exceeds a predetermined limit, generating an announcement containing at least one word corresponding to a selected field of an identified entry for a positive or negative response; (e) upon receipt of a positive response, identifying database entries which contain fields matching the recognised responses and whose selected fields match the said word; and (f) repeating steps (d) and (e) at least once.

Claim 13

A method of speech recognition comprising: (a) generating at least one announcement requiring a spoken response; (b) recognising the response(s); (c) identifying database entries containing fields matching the recognised responses; (d) in the event that the number of such entries does not exceed, or equals, a predetermined limit, generate an announcement presenting one or more of the records each containing one or more of the field(s) of the matching entries, and exit; (e) in the event that the number of such entries exceeds a predetermined limit, for a particular field or selection of fields, examine a unique representation or selection of representations, which is contained in the particular field or combination of fields of the identical entries, a unique representation or combination of representations being established according to a know criterion; (f) in the event that the number of such unique representation or combination of representations is above a predetermined limit, repeat (e), selecting another field or selection of fields, according to a pre-determined order, that has not already been considered until no such fields remain to be considered; and (g) generating at least one announcement requiring a spoken response that has not already been requested; (h) recognising the response(s); (i) repeat (c) at least once; (j) in the event that the number of such unique representation or combination of representations is below or equal to a predetermined limit, generate an announcement containing at least one word corresponding to the selected field(s) of an identified entry for a positive or negative response; (k) upon receipt of a positive response, identifying database entries which contain fields matching the recognised responses and whose selected fields match the said words or combination of words; and (l) repeat steps (d) and (e) at least once; (m) upon receipt of a negative response for all such word or words, exit.

Claim 14

A method of speech recognition comprising: (a) generating at least one announcement requiring a spoken response; (b) recognising the response(s); (c) identifying database entries containing fields matching the recognised responses; (d) in the event that the number of such entries does not exceed, or equals, a predetermined limit, generate an announcement presenting one or more of the records each containing one or more of the field(s) of the matching entries, and exit; (e) in the event that the number of such entries exceeds a predetermined limit, for a particular field or selection of fields, examine a unique representation or selection of representations, which is contained in the particular field or combination of fields of the identical entries, a unique representation or combination of representations being established according to a know criterion; (f) in the event that the number of such unique representation or combination of representations is above a predetermined limit, repeat (e), selecting another field or selection of fields, according to a pre-determined order, that has not already been considered until no such fields remain to be considered; and (g) generating at least one announcement requiring a spoken response that has not already been requested; (h) recognising the response(s); (i) repeat (c) at least once; (j) in the event that the number of such unique representation or combination of representations is below or equal to a predetermined limit, generate an announcement containing at least one word corresponding to the selected field(s) of an identified entry for a positive or negative response; (k) upon receipt of a positive response, identifying database entries which contain fields matching the recognised responses and whose selected fields match the said words or combination of words; and (l) repeat steps (d) and (e) at least once; (m) upon receipt of a negative response for all such word or words, exit.