Natural language dialogue method and natural language dialogue system
First Claim
Patent Images
1. A natural language dialogue system comprising:
- a properties database storing a plurality of records, wherein each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores value data;
a speech sampling module receiving a first speech input, wherein the first speech input is a sentence; and
a natural language comprehension system coupled to the properties database and the speech sampling module, the natural language comprehension system capturing a noun of the sentence as a first keyword, parsing the first keyword included in the first speech input and executing a full-text search in a structured database according to the first keyword to obtain at least one report answer to form a candidate list, the natural language comprehension system selecting the at least one report answer from the candidate list according to the properties database and outputting a first speech response that is a speech converted by the at least one report answer, each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, the speech sampling module receiving a plurality of speech inputs other than the first speech input, the natural language comprehension system capturing a user'"'"'s preference data from the speech inputs other than the first speech inputs and storing the user'"'"'s preference data into the properties database,when the speech inputs other than the first speech input present the at least one report answer is incorrect, the at least one report answer is removed from the candidate list,wherein the first speech response is a sentence outputted through a speech output interface,wherein the full-text search compares the value data of each of the records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records.
1 Assignment
0 Petitions
Accused Products
Abstract
A natural language dialog method and a natural language dialog system are provided. In the method, a first speech input is received and parsed to generate at least one keyword included in the first speech input, so that a candidate list including at least one report answer is obtained. According to a properties database, one report answer is selected from the candidate list, and a first speech response is output according to the report answer. Other speech inputs are received, and a user'"'"'s preference data is captured from the speech inputs. The user'"'"'s preference data is stored in the properties database.
24 Citations
60 Claims
-
1. A natural language dialogue system comprising:
-
a properties database storing a plurality of records, wherein each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores value data; a speech sampling module receiving a first speech input, wherein the first speech input is a sentence; and a natural language comprehension system coupled to the properties database and the speech sampling module, the natural language comprehension system capturing a noun of the sentence as a first keyword, parsing the first keyword included in the first speech input and executing a full-text search in a structured database according to the first keyword to obtain at least one report answer to form a candidate list, the natural language comprehension system selecting the at least one report answer from the candidate list according to the properties database and outputting a first speech response that is a speech converted by the at least one report answer, each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, the speech sampling module receiving a plurality of speech inputs other than the first speech input, the natural language comprehension system capturing a user'"'"'s preference data from the speech inputs other than the first speech inputs and storing the user'"'"'s preference data into the properties database, when the speech inputs other than the first speech input present the at least one report answer is incorrect, the at least one report answer is removed from the candidate list, wherein the first speech response is a sentence outputted through a speech output interface, wherein the full-text search compares the value data of each of the records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. A natural language dialogue method comprising:
-
receiving a first speech input, wherein the first speech input is a sentence; capturing a noun of the sentence as a first keyword from the first speech input; parsing the first keyword included in the first speech input, and executing a full-text search in a structured database according to the first keyword to obtain at least one report answer to form a candidate list; selecting the at least one report answer from the candidate list according to a properties database, wherein each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, wherein the properties database stores a plurality of records, each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores value data; outputting a first speech response that is converted by the at least one report answer, wherein the first speech response is a sentence outputted through a speech output interface; receiving a plurality of speech inputs other than the first speech input; capturing a user'"'"'s preference data from the speech inputs other than the first speech input; and storing the user'"'"'s preference data into the properties database, wherein when the speech inputs other than the first speech input present the at least one report answer is incorrect, the at least one report answer is removed from the candidate list, wherein the full-text search compares the value data of each of the records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records. - View Dependent Claims (10, 11, 12, 13, 14, 15, 16)
-
-
17. A language dialogue system comprising:
-
a properties database for storing a user'"'"'s preference data and a plurality of records, wherein each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores value data; and a language comprehension system capturing a noun of the sentence as a first keyword from a first speech input comprising a sentence, parsing a first keyword included in the first speech input and executing a full-text search in a structured database according to the first keyword to obtain at least one report answer to form a candidate list, the language comprehension system selects one of the at least one first report answer from the candidate list according to the user'"'"'s preference data stored in the properties database and outputs a first speech response that is converted by the selected one of the at least one first report answer, each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, wherein the language comprehension system obtains a first user'"'"'s preference data according to the first speech input and revises the user'"'"'s preference data stored into the properties database according to the first user'"'"'s preference data, when the speech inputs other than the first speech input present the at least one report answer is incorrect, the at least one report answer is removed from the candidate list, wherein the first speech response is a sentence outputted through a speech output interface, wherein the full-text search compares the value data of each of the records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records. - View Dependent Claims (18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36)
-
-
37. A language dialogue method comprising:
-
receiving a first speech input, wherein the first speech input is a sentence; capturing a noun of the sentence as a first keyword from the first speech input; parsing the first keyword included in the first speech input and executing a full-text search in a structured database according to the first keyword to obtain at least one report answer to form a candidate list; selecting the at least one first report answer from the candidate list according to a user'"'"'s preference data stored in a properties database, wherein each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, wherein the properties database stores a plurality of records, each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores value data; outputting a first speech response that is a speech converted by the at least one first report answer; obtaining a first user'"'"'s preference data according to the first speech response; and revising the user'"'"'s preference data stored in the properties database according to the first user'"'"'s preference data, when the speech inputs other than the first speech input present the at least one report answer is incorrect, the at least one report answer is removed from the candidate list, wherein the first speech response is a sentence outputted through a speech output interface, wherein the full-text search compares the value data of each of the records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records. - View Dependent Claims (38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56)
-
-
57. A method of providing a corresponding report answer corresponding to user'"'"'s speech inputs in response to a user'"'"'s habit, the method being suitable for a mobile communication apparatus and comprising:
-
receiving a first speech input from a user, wherein the first speech input is a sentence comprising a noun of the sentence as a keyword and a preference indication; and outputting a speech response that is a speech converted by the corresponding report answer in response to a user'"'"'s preference data stored by the user, the user'"'"'s preference corresponding to the keyword, wherein the speech response is output according to a most popular user'"'"'s preference of the user'"'"'s preference, wherein the first speech response is a sentence outputted through a speech output interface, wherein each of the report answers has a priority, the priorities of the report answers of the candidate list are determined according to usages of the report answers, the first report answer is one of the report answers with the highest priority, wherein when a plurality of speech inputs other than the first speech input present the at least one report answer is incorrect, the corresponding report answer is removed from a candidate list which is obtained by parsing the keyword and executing a full-text search in a structured database according to the keyword, wherein structured database comprises a plurality of records, each of the records comprises a title field, the title field comprises at least one sub-field, each of the at least one sub-field comprises an indication field and a value field, the indication field stores indication data, and the value field stores the value data, wherein the full-text search compares the value data of each of records and the first keyword to determine whether each of the records is matched and selects the report answers by the matched records, wherein the user'"'"'s preference data is revised according to the preference indication included in the first speech input. - View Dependent Claims (58, 59, 60)
-
Specification