Abstract: A method for hosting a programming environment and processing user input includes the step of receiving, during a first session between a user and an interaction environment, via one of a plurality of media gateways, a definition of an expression type, the definition specifying an expression format and a response type. The definition is stored. During a second session between a second user and the interaction environment, an expression having a semantic structure is received, from a user, via one of a plurality of media gateways. The semantic structure of the expression is evaluated. An expression format of the received expression is identified as the expression format specified by the definition of the expression type, responsive to the evaluation. A response to the expression is generated responsive to an execution of a computer program associated with the response type specified by the definition of the expression type.

Abstract: In a method and system for processing user input, a resource is received during a session between a user and an interaction environment, the resource including predefined markers and a first program executable by a browser, at least one of the predefined markers identifying a display format of at least one interface element in a user interface. A definition specifying a response type is received, the response type identifying a second program for execution upon an interaction with the at least one interface element. During a second session, the user provides an expression and interacts with the at least one interface element. A response to the expression is generated based on an execution of the second computer program identified by the response type. The response is displayed in a user interface including the at least one interface element formatted responsive to at least one of the plurality of predefined markers.

Abstract: In a method and system for hosting a programming environment and processing user input within an information retrieval system, a definition of an expression is received from a user, during a first session between the user and an information retrieval system, the definition specifying an expression format, a response type, and a display format. The definition is stored. During a second session between the user and the information retrieval system, an expression having a semantic structure is received. The semantic structure of the expression is evaluated. A response to the expression is displayed to the user in a user interface according to the display format specified by the definition of the expression type.

Abstract: System of correcting misspelled words in input text detects a misspelled word in the input text, determines a list of alternative words for the misspelled word, and ranks the list of alternative words based on a context of the input text. In certain embodiments, finite state machines (FSMs) are utilized in the spelling and grammar correction process, storing one or more lexicon FSMs, each of which represents a set of correctly spelled reference words. Storing the lexicon as one or more FSMs facilitates those embodiments of the invention employing a client-server architecture. The input text to be corrected may also be encoded as a FSM, which includes alternative word(s) for word(s) in need of correction along with associated weights. The invention adjusts the weights by taking into account the grammatical context in which the word appears in the input text.

Abstract: A method for processing user input includes the step of receiving, during a session, via one of a plurality of media gateways, from a user, an expression having a semantic structure. The semantic structure of the expression is evaluated. An expression type is identified, responsive to the evaluation of the semantic structure. Based on the expression type, a response to the expression is generated. A determination is made as to whether to store the received expression, the response, and an identification of the user.

Abstract: A method for processing user input includes the step of receiving, during a first session between a user and an interaction environment, via one of a plurality of media gateways, from the user, a resource including a plurality of predefined markers and at least one expression. An expression type is identified, responsive to an evaluation of the semantic structure and to one of the predefined markers. A determination is made as to whether to store at least one of: an identification of the user and the at least one expression. The resource is displayed in a viewable display format, responsive to a second predefined marker. A second expression is received. An expression type is identified, responsive to an evaluation of a semantic structure of the second expression. Based on the expression type and on the resource received during the first session, a response to the second expression is generated.

Abstract: System of correcting misspelled words in input text detects a misspelled word in the input text, determines a list of alternative words for the misspelled word, and ranks the list of alternative words based on a context of the input text. In certain embodiments, finite state machines (FSMs) are utilized in the spelling and grammar correction process, storing one or more lexicon FSMs, each of which represents a set of correctly spelled reference words. Storing the lexicon as one or more FSMs facilitates those embodiments of the invention employing a clinet-server architecture. The input text to be corrected may also be encoded as a FSM, which includes alternative word(s) for word(s) in need of correction along with associated weights. The invention adjusts the weights by taking into account the grammatical context in which the word appears in the input text.

Abstract: Described are techniques used in fulfilling an information need not formulated as a natural language question. A query is accepted that includes one or more words not formulated as a natural language question. The query is syntactically analyzed in order to extract a concept from the query. Documents are identified containing the one or more words. The analyzed query is matched to one or more question keys in a structures repository. A best match is determined between the analyzed query and a question key of the structures repository. The key of the at least one best key match is mapped to partially specified queries and the key of the best question key match is mapped to questions representing an information need corresponding to the query. The partially specified queries and questions are matched to potential answers contained in a collection of documents. One or more matching answers are returned based on a metric.

Abstract: The system initiates a search at a first network site for user-specified data in a remote database at a second network site and conducts the search at a third network site (e.g., at a host computer's site). To begin, the system receives, at the first network site, a provider identifier associated with the database from the second network site. Thereafter, the user-specified data is input at the first network site, following which the user-specified data and the provider identifier are output from the first network site to the third network site. The system then searches for the user-specified data in a database at the third network site using the provider identifier. This database at the third network site includes data that corresponds to data stored in the remote database at the second network site.

Abstract: The invention offers new approaches to fulfilling an information need, in particular to finding a result for a query based on a large body of information such as a collection of documents. The invention accepts a query containing an unspecified portion that expresses the information need. The invention locates matches for the query within a body of information and returns the matches or portions thereof in addition to or instead of identifiers for documents in which the matches are found. The invention ranks the matches in order to provide the most relevant information. One preferred method of ranking considers the number of instances of a match among a plurality of documents. The invention further defines a new type of index that includes contexts in which terms occur and provides methods of searching such indices to fulfill an information need.

Abstract: The present invention is a system for answering a natural language question. The system receives a question and transforms the question into one or more partially unspecified queries. The system then identifies matches for the queries in a body of information. The matches are optionally ranked, preferably based on the number of times each match is identified. The matches are provided as answers to the questions.

Abstract: The system initiates a search at a first network site for user-specified data in a remote database at a second network site and conducts the search at a third network site (e.g., at a host computer's site). To begin, the system receives, at the first network site, a provider identifier associated with the database from the second network site. Thereafter, the user-specified data is input at the first network site, following which the user-specified data and the provider identifier are output from the first network site to the third network site. The system then searches for the user-specified data in a database at the third network site using the provider identifier. This database at the third network site includes data that corresponds to data stored in the remote database at the second network site.

Abstract: A system builds a text fragment database for use in translating fragments of text from a source language into a target language. The system first stores a sentence database in memory, the sentence database comprising a plurality of sentence pairs, each sentence pair including a sentence in the source language and a corresponding sentence in the target language. The system then locates corresponding source and target text fragments in corresponding source and target language sentences, respectively, and stores the source text fragment together with the target text fragment in the text fragment database. The text fragment database can then be used to translate text from the source language into the target language. To this end, the system inputs text in the source language, extracts a text fragment from the input text, and locates the extracted text fragment in the text fragment database.

Abstract: A system of correcting misspelled words in input text detects a misspelled word in the input text, determines a list of alternative words for the misspelled word, and ranks the list of alternative words based on a context of the input text. The system then selects one of the alternative words from the list, and replaces the misspelled word in the text with the selected one of the alternative words. In certain embodiments of the invention finite state machines are utilized in the spelling and grammar correction process. Thus according to certain embodiments the invention stores one or more lexicon finite state machines (FSM), each of which represents a set of correctly spelled reference words. Storing the lexicon as one or more finite state machines facilitates those embodiments of the invention employing a client-server architecture. The input text to be corrected may also be encoded as a finite state machine, which includes alternative word(s) for word(s) in need of correction along with associated weights.