Page 241 - Ai_C10_Flipbook
P. 241
In the above example the lengthy text is broken down into individual words.
• Syntactic Analysis: In this stage, the converted
words and sentences are arranged according to The agile deer over leaped the sluggish bear
the grammar of the language. This arrangement
highlights the relationships between words. For
WRONG
example, if you are using the English language,
the syntactic analyser will reject unorganised
sentences like “deer over leaped the sluggish The agile deer leaped over the sluggish bear
bear”.
RIGHT
In the above example, the grammar is checked.
• Semantic Analysis: In this stage, the semantic
The agile deer over leaped the sluggish bear
analyser checks the meaningfulness of the text
and extracts logical meaning from it. This task is
WRONG
done by mapping syntactic structures and
objects in the task domain. In the example the
meaning of the sentence is checked to make The agile deer leaped over the sluggish bear
sure it actually makes sense. For example, it will
reject a sentence that contains “cruel kindness” RIGHT
in it.
• Discourse Integration: In this stage, the
meaning of the sentence is drawn according to The agile deer leaped over the sluggish bear. Then it went
into the hedges.
the meanings of the preceding and succeeding
sentences. Here "it" means "the deer"
In the example, the meaningfulness of the The agile deer leaped over it. Then it went into the hedges.
preceding statement is dependent on the
Here "it" is unknown
meaningfulness of the previous statement.
• Pragmatic Analysis: In this stage, deriving
those aspects of language which require real
Hang on, I will see you in a while.
world knowledge. The actual meaning of the
sentence is rechecked in this process.
Means "Wait a moment, I will be back
to see you later"
In the above example the literal meaning is
not taken from each word, but the intended
It does not mean that one needs to
meaning of the word is taken in that sentence.
hang on to something
Natural Language Processing 239

