Automatic topic identification and switch for natural language search of textual document collections
First Claim
1. A method for automatic topic identification and switch for natural language search of textual document collections comprising the steps of:
- receiving a user'"'"'s textual free-form natural language input query;
maintaining a session history of previous input queries by the user;
interpreting the user'"'"'s input query in the context of the session history by the further steps of;
combining said input query with said previous input queries to create combined input;
creating combined search terms from the combined input;
using the combined search terms to search prior results (PR) and return new results (NR); and
detecting an implicit change in topic based on the session history, said prior results (PR) and said new results (NR) in response to said input query, wherein the step of detecting an implicit change in topic is performed by determining if the set of new search results (NR) based on the combined search terms is empty; and
iteratively drilling-down on successive queries in the context of previous queries on a topic by searching within a previously determined set of previous responses to narrow the documents within the search if there is no implicit change in topic.
1 Assignment
0 Petitions
Accused Products
Abstract
A method for iteratively drilling-down on a user'"'"'s textual free-form natural language query uses a session history to interpret successive queries in the context of previous queries on a topic or topics and to detect an implicit switch in topic. By maintaining a session history of the user'"'"'s free-form natural language input and by automatically determining whether there is a topic or context switch, the search process is substantially simplified and is more effective; that is, more accurate answers to a user'"'"'s queries are found faster. In addition, as the system operates on free-form natural language input, automatically constructing the actual search expressions, the complexity of constructing successive search expressions is obviated. If the system determines the user is, according to its session history and tests, asking successive questions within a given topic or context, the system keeps searching within a previously determined given set of previous responses on that context or topic. This effectively narrows the documents found allowing the user to quickly and accurately find just the documents of interest. If the system determines the user has implicitly changed context or topic, based on its session history and tests, it searches all the information at its disposal; i.e., all of the collections of documents.
173 Citations
11 Claims
-
1. A method for automatic topic identification and switch for natural language search of textual document collections comprising the steps of:
-
receiving a user'"'"'s textual free-form natural language input query;
maintaining a session history of previous input queries by the user;
interpreting the user'"'"'s input query in the context of the session history by the further steps of;
combining said input query with said previous input queries to create combined input;
creating combined search terms from the combined input;
using the combined search terms to search prior results (PR) and return new results (NR); and
detecting an implicit change in topic based on the session history, said prior results (PR) and said new results (NR) in response to said input query, wherein the step of detecting an implicit change in topic is performed by determining if the set of new search results (NR) based on the combined search terms is empty; and
iteratively drilling-down on successive queries in the context of previous queries on a topic by searching within a previously determined set of previous responses to narrow the documents within the search if there is no implicit change in topic. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
determining if there was a prior user input upon receiving the user'"'"'s textual free-form natural language input query;
if not, creating new search terms from the user'"'"'s textual free-form natural language input query.
-
-
3. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 2, wherein said interpreting step further comprises the steps of:
-
categorizing the user'"'"'s textual free-form natural language input query;
creating final new search terms by adding terms and/or weight terms based on categories;
categorizing the combined user'"'"'s textual free-form natural language input query with prior user input from the session history; and
creating final combined search terms by adding terms and/or weight terms based on categories.
-
-
4. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 3, further comprising the steps of:
-
searching a textual document collection with the new search terms;
returning and processing new search results based on the new search terms; and
updating the session history with the new search results based on the new search terms.
-
-
5. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 4, wherein said interpreting step further comprises the steps of:
-
searching prior results with combined search terms;
returning and processing new search results based on the combined search terms; and
if an implicit change in topic is not detected, updating the session history with the new search results based on the combined search terms.
-
-
6. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 5, further comprising the steps of:
-
creating new search terms from the user'"'"'s textual free-form natural language input query;
searching a textual document collection with the new search terms;
returning and processing new search results based on the new search terms; and
updating the session history with the new search results based on the new search terms.
-
-
7. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 6, further comprising the step of displaying the new search results.
-
8. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 2, further comprising the steps of:
-
searching a textual document collection with the new search terms;
returning and processing new search results based on the new search terms; and
updating the session history with the new search results based on the new search terms.
-
-
9. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 8, wherein said interpreting step further comprises the steps of:
-
searching prior results with combined search terms;
returning and processing new search results based on the combined search terms; and
if an implicit change in topic is not detected, updating the session history with the new search results based on the combined search terms.
-
-
10. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 9, further comprising the steps of:
-
creating new search terms from the user'"'"'s textual free-form natural language input query;
searching a textual document collection with the new search terms;
returning and processing new search results based on the new search terms; and
updating the session history with the new search results based on the new search terms.
-
-
11. The method for automatic topic identification and switch for natural language search of textual document collections recited in claim 10, further comprising the step of displaying the new search results.
Specification