Abstract | ||
---|---|---|
The paper describes refinements that are currently being investigated in a model for part-of-speech assignment to words in unrestricted text. The model has the advantage that a pre-tagged training corpus is not required. Words are represented by equivalence classes to reduce the number of parameters required and provide an essentially vocabulary-independent model. State chains are used to model selective higher-order conditioning in the model, which obviates the proliferation of parameters attendant in uniformly higher-order models. The structure of the state chains is based on both an analysis of errors and linguistic knowledge. Examples show how word dependency across phrases can be modeled. |
Year | DOI | Venue |
---|---|---|
1989 | 10.3115/1075434.1075451 | HLT |
Keywords | Field | DocType |
higher-order model,part-of-speech assignment,equivalence class,vocabulary-independent model,hidden markov model,state chain,parameters attendant,linguistic knowledge,selective higher-order conditioning,pre-tagged training corpus,phrase-dependent word tagging,unrestricted text,higher order,model selection,part of speech | Computer science,Markov model,Phrase,Speech recognition,Artificial intelligence,Natural language processing,Equivalence class,Hidden Markov model,Hidden semi-Markov model | Conference |
ISBN | Citations | PageRank |
1-55860-112-0 | 13 | 72.05 |
References | Authors | |
3 | 1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Julian Kupiec | 1 | 1061 | 381.10 |