Back to results
US 7,624,099 B2Granted
Client-server word-breaking framework
AssigneeMicrosoft Corporation
InventorsSanjeev Katariya, William D. Ramsey
Life of this patent
Estimating the term…
Abstract
Word-breaking of a query from a client machine in a client-server environment includes determining whether to use a first word breaking module operable with a client machine in the client-server environment and/or a second word breaking module operable with a server in the client-server environment.
Claims16
A computer-implemented method of performing word-breaking in a client-server environment, the method comprising:
- providing a first word breaking module with first word breaking boundaries operable with a client machine in the client-server environment and a second word breaking module with second word breaking boundaries operable with a server in the client-server environment;
- determining whether to use the first word breaking module and/or the second word breaking module to perform word breaking of a query received by the client machine depending upon whether the first and second word breaking boundaries would provide the same word breaking of the query;
- performing word breaking of the query to obtain a set of tokens using at least one of the first word breaking module and the second word breaking module;
- and
- wherein the step of providing includes providing a named entity recognizer operable with the client machine, and the method further comprises performing named entity recognition on the client machine using the named entity recognizer to identify named entities known to the client machine, but not known to the server.
2 more elements
A computer-implemented method of performing word-breaking and named entity recognition in a client-server environment, the method, comprising:
- receiving a query at a client machine;
- providing the query to a server;
- performing word-breaking of the query on the server to obtain a first set of tokens comprising words of the query;
- performing word-breaking of the query on the client machine to obtain a second set of tokens comprising words of the query;
- determining whether to utilize the first or second set of tokens depending upon whether there are differences in the first and second set of tokens;
- selecting the first set of tokens if there are differences in the first and second set of tokens;
- selecting the second set of tokens if there are no differences in the first and second set of tokens;
- performing named entity recognition on the client machine to identify one or more named entities in the selected set of tokens and data accessible to the client machine;
- providing at least information related to the identified named entity to the server in order to complete a task.
6 more elements
A computer-implemented method of performing word-breaking and named entity recognition in a client-server environment, the method, comprising:
- providing a first client machine, a second client machine and a server in communication over a network with both the first client machine and the second client machine;
- providing a first word breaking module operable with the first client machine in the client-server environment and a third word breaking module operable with the server in the client-server environment;
- providing a second word breaking module operable with the second client machine in the client-server environment and the second word breaking module operable with the server in the client-server environment,;
- receiving a first query at the first client machine requiring word breaking;
- receiving a second query at the second client machine requiring word breaking;
- performing word breaking on the first query utilizing the first client machine to generate a first set of tokens accessible to the first client machine by using the first word breaking module on the first client machine to identify words known to the first client machine;
- performing word breaking on the second query utilizing the second client machine to generate a second set of tokens accessible to the second client machine by using the second word breaking module on the second client machine to identify words known to the second client machine;
- performing word breaking on the first query utilizing the server to generate a third set of tokens and data accessible to the server by using the third word breaking module on the server to identify words known to the server;
- performing word breaking on the second query utilizing the server to generate a fourth set of tokens and data accessible to the server by using the third word breaking module on the server to identify words known to the server;
- comparing the first set of tokens to the third set of tokens and if different, transmitting the third set of tokens from the server to the first client machine to perform named entity recognition on the third set of tokens;
- comparing the second set of tokens to the fourth set of tokens and if different, transmitting the fourth set of tokens from the server to the first client machine to perform named entity recognition on the fourth set of tokens;
- completing a task on the first client machine using at least information related to the identified words from the first set of tokens if the first set of tokens and the third set of tokens are identical, or from the third set of tokens if the first set of tokens and the third set of tokens are different;
- and
- completing a task on the second client machine using at least information related to the identified words identified from the second set of tokens if the second set of tokens and the fourth set of tokens are identical, or from the fourth set of tokens if the second set of tokens and the fourth set of tokens are different wherein the task completed on the first client machine is different from the task performed on the second client machine.
11 more elements
Dependencies are inferred from the claim text.