Bidirectional Natural Language Parsing using Streams and Counterstreams
نویسندگان
چکیده
This thesis investigates the bidirectional exchange of information between linguistic and non-linguistic semantic inputs containing ambiguities. Such exchange is critical to Cognitively Complete Systems, in which collections of related representations and processes cooperate for their mutual problem-solving benefit. The exchange paradigm of reconciliation is defined, in which ambiguities and gaps in multiple input domains are simultaneously resolved. A complete architecture for implementing reconciliation between the linguistic and non-linguistic semantic domains is described, employing the Streams and Counterstreams bidirectional search technique, Combinatory Categorial Grammar, and Lexical Conceptual Semantics. This architecture has been implemented, resulting in a system that can act as an ambiguity resolving constraint between linguistic and non-linguistic inputs, with the side effect of also producing a language parsing and generating system. For example, if the system is presented with the phrase “The orange rolls,” and a non-linguistic input (perhaps from a vision subsystem) describing the presence of bread (rolls) in the scene, the system will automatically select the interpretation of the the linguistic input describing pieces of bread of orange color. Thus, ambiguity is resolved in the linguistic domain, and the statement that the rolls have the property orange can be propagated to the non-linguistic semantic domain. Thesis Supervisor: Patrick H. Winston Title: Ford Professor of Artificial Intelligence and Computer Science
منابع مشابه
Bidirectional Natural Language Parsing using Streams and Counterstreams by Keith A . Bonawitz
This thesis investigates the bidirectional exchange of information between linguistic and non-linguistic semantic inputs containing ambiguities. Such exchange is critical to Cognitively Complete Systems, in which collections of related representations and processes cooperate for their mutual problem-solving benefit. The exchange paradigm of reconciliation is defined, in which ambiguities and ga...
متن کاملAn improved joint model: POS tagging and dependency parsing
Dependency parsing is a way of syntactic parsing and a natural language that automatically analyzes the dependency structure of sentences, and the input for each sentence creates a dependency graph. Part-Of-Speech (POS) tagging is a prerequisite for dependency parsing. Generally, dependency parsers do the POS tagging task along with dependency parsing in a pipeline mode. Unfortunately, in pipel...
متن کاملSystem Demonstration Natural Language Generation With Abstract Machine
We present a system for Natural Language Generation based on an Abstract Machine approach. Our abstract machine operates on grammars encoded in a uni cation-based Typed Feature Structure formalism, and is capable of both generation and parsing. For e cient generation, grammars are rst inverted to a suitable form, and then compiled into abstract machine instructions. A dual compiler translates t...
متن کاملReversible delayed lexical choice in a bidirectional framework
We describe a bidirectional framework for natural language parsing and generation , using a typed feature formalism and an HPSG-based grammar with a parser and generator derived from parallel processing algorithms. We present an approach to delayed lexical choice in generation , based on subsumption within the sort hierarchy, using a lexicon of nnder-instantiated signs which at'(; derived fi'om...
متن کامل