Semantic Interpretation





 Semantics  Word  The

& Logical Form

Senses and Ambiguity

Basic Logical Form Language



Semantics & Logical Form

is traditionally defined as the

study of meaning in language



I was at an airport recently and while I was walking towards my departure gate. “Do you know what gate you are going to?” 11/22/2012 4 .Introduction  Defining the notions of semantics & meaning is surprisingly difficult because the terms are used for several different purposes in Natural & technical usage  For example. a guard at the entrance asked.

I answered “yes”. as he then asked. But this response was based on a misunderstanding of what the guard meant. but may be used for different purposes 11/22/2012 5 . This is a complex issue.Cont…. "Which gate is it?" He clearly had wanted me to tell him the gate number. because "Do you know what gate you are going to?" has a single meaning.

Cont…. We will use the term "meaning" in this context-independent sense. The process of mapping a sentence to its logical form is called semantic interpretation 11/22/2012 6 . and we will use the term "usage" for the context-dependent aspects. The representation of contextindependent meaning is called the logical form.

The process of mapping the logical form to the final knowledge representation (KR) language is called contextual interpretation 11/22/2012 7 ..Cont.

11/22/2012 NILESH.SHEWALE 8 .A.

Logical Form? Logical form is defined as the literal meaning of the utterance. 11/22/2012 9 . and the logical form language is the same as the final knowledge representation language.

and include the relationship that the person owns the ball. a ball B0005 and a person P86. 11/22/2012 10 . (OWNS P86 B0005)}. (PERSON P86). Let us encode this situation as the set {(BALL B0005).A situation is a set of objects and relations between those objects. A very simple situation might consist of two objects.

(RED B0005)}. (PERSON P86). (OWNS P86 B0005). 11/22/2012 11 .If a new sentence is interpreted with respect to this situation then it produces a new situation The utterance "The ball is red" might produce a new situation that consists of the old situation plus the new fact that B0005 has the property {RED: ((BALL B0005).

2 shows this view of the interpretation process. Figure 8. the logical form is a function that maps the discourse situation in which the utterance was made to a new discourse situation that results from the occurrence of the utterance.In this. treating the logical form as a function between Situations 11/22/2012 12 .

A.SHEWALE 13 .11/22/2012 NILESH.

Overview  Semantics  Word  The & Logical Form Senses and Ambiguity Basic Logical Form Language 11/22/2012 14 .

What is word sense?  Word sense is one of the meanings of a word .  “Words” are having different meanings based on the context of the word usage in a sentence. Example:-We went to see a play at the theater 11/22/2012 -The children went out to play in the 15 .

we need a much broader ontology than commonly found in work on formal logic 11/22/2012 16 .Every word has one or more senses senses can be organized into a set of Different broad classes of objects by which we classify the world The set of different classes of objects in a representation is called its ontology To handle a natural language.

state. position. quality (such as bright red). action. quantity (such as numbers).C.relation.) The major classes that Aristotle suggested were substance (physical objects). place.classifications of objects have been of interest for a very long time and arise in the writings of Aristotle (384—322 B. Two of the most influential classes are actions and events 11/22/2012 17 . and affection. time.

It was hard work. 18 . the pronoun "it" refers to the action of lifting the box  Another very influential category is the situation "We laughed and sang at the football game"  It describes a set of activities performed at a particular time and location.Example.  Here. described as the situation "the football11/22/2012 game". We lifted the box.

"Happy cats and dogs live on the farm“ ambiguous between whether the dogs are also happy or not 11/22/2012 19 . the word "kid" seems to be ambiguous between a baby goat and a human child For instance.Ambiguity is a serious problem during semantic interpretation We can define a word as being semantically ambiguous if it maps to more than one sense.

does the sentence "Every boy loves a dog" mean that there is a single dog that all boys love. A very common example involves quantifier scoping. structural ambiguity are truly semantic and arise from a single syntactic structure. but the difference lies in how the 11/22/2012 quantifiers are scoped. 20 .  For instance. or that each boy might love a different dog?  The syntactic structure is the same in each case.

Overview  Semantics  Word  The & Logical Form Senses and Ambiguity Basic Logical Form Language 11/22/2012 21 .

 This language will resemble FOPC. such as network-based representations  The word senses will serve as the atoms or constants of the representation 11/22/2012 22 . It defines a language in which you can combine elements to form meanings for more complex expressions. although there are many equivalent forms of representation.

Constants that describe relations and properties are called predicates "Fido is a dog“ The proposition corresponding to the above 23 sentence would be constructed from the term . including abstract objects such as events and situations. constants that describe objects in the world.These constants can be classified by the types of things they describe. are called terms. For instance.

“Sue loves Jack” written as (LOVES1 SUE1 JACK1) 11/22/2012 24 .cont……. “(DOG1 FIDO1)” Predicates that take a single argument are called unary predicates or properties Predicates that take two arguments are called binary predicates Predicates that take n arguments are called n-ary predicates For instance.

11/22/2012 25 .

the operator NOT allows you to construct a proposition that says that some proposition is not true. “Sue does not love Jack” would be (NOT (LOVES1 SUE1 JACK1)) 11/22/2012 26 .More complex propositions are constructed using a new class of constants called logical operators For example.

English also contains operators that combine two or more propositions to form a complex proposition The general form for such a proposition is (connective proposition proposition) "Jack loves Sue or Jack loves Mary“ (OR1 (LOVES1 JACK1 SUE1) (LOVES1 JACK1 MARY1)) Natural language quantifiers have restricted ranges and thus are more complex than those found in FOPC 11/22/2012 27 .

Note that this has a very different meaning from the formula 11/22/2012 28 ."all dogs bark" or "most people laughed". the sentence "Most dogs bark" would have the logical form (MOST1 d1 (DOG1 d1) (BARKS1 d1)) This means that most of the objects dl that satisfy (DOG1 d1) also satisfy (BARKS1 d1). which require constructs that are often called generalized quantifiers These quantifiers are used in statements of the general form (quantifier variable : restriction-proposition body-proposition) For instance.

The sentence "The dog barks" would have a logical form (THE x: (DOG1 x) (BARKS1 x))  Another construct needs to be introduced to handle plural forms. as in a phrase such as “the dogs bark” (THE x: ((PLUR DOG1) x) (BARKS1 x)) 11/22/2012 29 . A very important class of generalized quantifiers corresponds to the articles "the" and "a".

PRES. and FUT  With these new operators. we can represent the difference in meaning between => John sees Fido => John saw Fido => John will see Fido 11/22/2012 30 . PAST. An important class of modal operators for natural language are the tense operators.

(PRES (SEES1 JOHN1 FIDO1)) (PAST (FUT (SEES1 JOHN1 FIDO1)) (SEES1 JOHN1 FIDO1)) 11/22/2012 31 .Cont……….

THANK YOU 11/22/2012 32 .

Sign up to vote on this title
UsefulNot useful