06.03.2013 Views

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

Artificial Intelligence and Soft Computing: Behavioral ... - Arteimi.info

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Thus the elements of communicative plans can be formalized <strong>and</strong> used<br />

for executing a task.<br />

18.6 Applications of Natural Language<br />

Underst<strong>and</strong>ing<br />

Natural language underst<strong>and</strong>ing programs have wide usage starting from<br />

commercial software to high performance VLSI CAD tools design. In the<br />

long run it will be part <strong>and</strong> parcel of most of the software. It is mainly<br />

required as a front-end tool for communication with the users. One of the first<br />

successful applications of the natural language underst<strong>and</strong>ing program was in<br />

a ‘question answering’ system. In this system, a paragraph is first supplied to<br />

the machine, which constructs an internal representation of the paragraph <strong>and</strong><br />

then matches the question part with the internal representation to answer it.<br />

Winograd’s SHRDLU [9] system in this regard needs special mention. In<br />

database management systems, the natural language underst<strong>and</strong>ing program is<br />

equally useful for building front-end tools. The user may submit his query in<br />

plain English <strong>and</strong> the machine will answer him in English as well. In expert<br />

systems, natural language processing is very much useful for submitting many<br />

facts to the database /working memory. For augmenting knowledge in the<br />

expert systems, this will also play a vital role. In VLSI CAD tool design, the<br />

users are expert electronic or electrical engineers. They do not want to<br />

remember rigid format of the syntax of the end-users’ comm<strong>and</strong>. So, if a<br />

natural language interface is present as a front tool, they will be able to save<br />

their valuable time <strong>and</strong> consequently will improve their productivity. In<br />

banking, library management, post office <strong>and</strong> telephone diagnostic software<br />

the natural language underst<strong>and</strong>ing program will be employed as a front-end<br />

tool in the short run.<br />

Exercises<br />

1. Draw the parse trees for the following sentences:<br />

a) The boy smoked a cigarette.<br />

b) The cat ran after a rat.<br />

c) She used a fountain pen to write her biography.<br />

2. Parse the following sentence by a transition network parser: The man<br />

reacted sharply.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!