Loading...
Please wait, while we are loading the content...
Similar Documents
Grammar-based Natural Language Understanding
| Content Provider | Semantic Scholar |
|---|---|
| Author | Noord, Gertjan Van |
| Copyright Year | 2003 |
| Abstract | This chapter describes the grammar-based natural language understanding (NLU) component of the OVIS system. The NLU component receives its input from the speech recognizer and passes the result of linguistic analysis on to the pragmatic interpretation component. The output of the speech recognizer is a word graph, which represents all different hypotheses for the spoken input from the user. Linguistic analysis of such a structure is more complicated than analysis for a single input sentence (cf. linguistic analysis for written input), as we will show later. The interface between the NLU component and the Dialogue Manager is defined by a special interface language, called the Update Language, described elsewhere. In the design of the NLU component, the following problems must be addressed. Firstly, a NLU component will be faced with ambiguity. The combination of lexical and structural ambiguities often leads to an enormous amount of possible readings for an input utterance. In ovis this problem is even more acute because of the use of word graphs. Techniques must be available to be able to deal with large numbers of analyses, and to be able to choose the most appropriate reading from such a set of candidate analyses. While the system must somehow deal with large numbers of analyses, many of which may eventually turn out to be useless, another requirement is that the system should be robust, i.e. it should do something useful even if something happens that the system didn’t expect, or if something goes wrong. For example, the input utterance may contain disfluencies or ungrammaticalities. Or the speech recognition component may have failed to recognize part of the user utterance. In such cases, a linguistic analysis component might proceed by computing a partial analysis of the input. NLU needs to be robust for three reasons. Firstly, it is quite difficult to anticipate in the grammar all linguistic constructions that might occur. This is one of the traditional problems for grammar-based NLU. Secondly, spoken language is full of hesitations, corrections, false starts etc. which are not always easy to detect. The third reason is that the utterance that was actually spoken is not guaranteed to be a path in the word graph, due to limitations of state-ofthe-art speech recognition, as well as external factors such as background noise, etc. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://odur.let.rug.nl/~vannoord/papers/ovisfinal.ps |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
