Some applications of NLP, such as grammar checking, depend only on the syntactic form of an input. Most applications, however, are dependent on what a sentence 'means'. For example, for information extraction, we want to find mentions of a type of an event, however expressed. For question answering, we need to connect the question to a body of knowledge which can provide the answer. In fact, as we shall see when we consider the processing of extended (multi-sentence) texts, we generally must integrate the sentence with a great deal of 'background knowledge' in order to understand what it means. How should we express the meaning of an utterance?
Clearly we can express the meaning in natural language. Why is this not satisfactory for analyzing the meaning? The components (the words) are ambiguous; even sentences in isolation are ambiguous. Resolving these ambiguities can be difficult. Even though syntactic analysis has made some relationships explicit, others are still implicit; for example, it does not indicate the quantificational structure of a sentence. Furthermore, the rules for inferring new facts from given facts in natural language may be very complicated.
Formal Languages for Meaning Representation [J&M 17.1]
In order to analyze and manipulate the meaning of sentences, we will transform the sentences into a meaning representation language. We want to transform the sentences into a language which
[J&M also mention the characteristic of having a canonical form for each meaning. This is an ultimate goal rather than an easily achieved criterion. However, we can see both syntactic analysis and semantic analysis as moving in this direction -- reducing paraphrase, i.e., reducing the variation in form for a given meaning.]
has simple rules of interpretation and inference, and in particular
has a logical structure determined by its form
is sufficiently expressive to capture the range of natural language meanings we shall require for our applications
These are the properties of the languages of logic. Actual systems may use different representations, but they are generally equivalent to the formal language (extensions of predicate calculus) we will use for presentation.
Predicate Logic [J&M 17.3]
The simplest form is propositional logic, but it is not powerful enough for our purposes. Predicate logic combines predicates and their arguments. Basic elements:
Predicate logic has simple rules of inference, such as modus ponens (from A and A==>B, infer B).
terms (constants, variables, and functions)
atomic formulas (predicate + arguments)
logical connectives (not, and, or, implies)
Representing events [J&M 17.4]
Predicate calculus is intended for representing “eternal truths” (like the facts of mathematics). We face several problems when we try to use it for representing events. First, how many arguments does an event have (consider J&M example of eating, p. 564)? In natural language, the same type of event may be described with many different sets of arguments and modifiers (time, location, speed, ...). We can use meaning postulates to relate these, but that requires many such postulates and may make commitments we do not intend. Second, we need to individuate events (say that two events are the same or different; count events).
We can address this problem by reification -- treating events as objects (J&M p. 566).
Other issues of expressiveness
There are many other issues which we may need to address in our meaning representation language:
Information extraction applications are generally concerned with identifying specific, individual events and relations between entities. They need to capture event modifiers such as time and location, but not quantification. Such applications typically use a frame or slot-filler representation. For each type of event (or set of event types taking the same arguments) we define a frame (template), with one slot containing a unique identifier of the event, and one slot for each possible argument/modifier. Similarly, a frame is defined for each type of entity. Slots may be filled with constants or the identifiers of other events or entities.
generalized quantifiers (for ‘some’, ‘most’, …)
tense and aspect
modality and belief (need to allow formulas as arguments: “John believes Fred likes Mary” = believe(John,like(Fred,Mary)) )
presupposition (“All the men on Mars drink Coca-Cola.”)
fuzziness (“The milk is warm.”)
Semantic Analysis: Adjusting the Representation
Mapping Syntax to Semantics (J&M Chapter 18))
We want to compute the semantic representation of a sentence from the parse tree. Because the parse tree provides a structural framework, we will use a compositional, syntax-driven translation process. This means that we will associate a (partial) semantic interpretation with some or all of the nodes of the parse tree, to be computed (using a rule) from the interpretations of its children.
We could embed this translation in a procedure associated with each type of node. Alternatively, one can formalize this by a set of rules associated with the productions of the grammar (J&M sec. 18.1 and 2). The grammar will be extended to add a SEM feature, representing the semantic interpretation of a node. Each production will then incorporate the rule for computing its SEM value, and the SEM of the root will be the interpretation of the sentence.
The semantics of a verb phrase is essentially the semantics of a clause, with one argument (the subject) missing … a predicate with one unbound argument. We can represent this by a lambda expression (p. 587). Lambda expressions are commonly used to capture the rules for composing the semantics. (See p. 559 for a presentation of lambda notation.)
For the process of translating syntactic to semantic forms, it is convenient to introduce restricted quantifiers, of the form
(forall x: C(x))
These do not add any power to predicate calculus; they can be rewritten
(exists x: C(x))
(forall x: C(x)) P(x) = (forall x) (C(x) => P(x))
Roughly speaking, a noun phrase can be translated to a constant or a restricted quantifier.
(exists x: C(x)) P(x) = (exists x) (C(x) & P(x))
One source of ambiguity is quantifier scope:
A woman gives birth in the United States every five minutes.
We can represent the two readings in conventional predicate calculus using different quantifier scopes. If we explicitly represent all the semantic ambiguities in a sentence in this way, we may have very many readings. It is therefore practical to initially produce (from the parse) a representation which captures multiple readings … which encodes (some of) the ambiguity. (And hope that this ambiguity can be resolved at a later stage of semantic analysis.)
In particular, we can use complex terms (J&M First edition p. 555)
with the understanding that
P() = (Quantifier x: C(x)) P(x)
If an expression contains several complex terms, the scope of the quantifiers is indeterminate. Semantic analysis will generate such quasi-logical forms, with a separate step then determining the quantifier scope and generating a predicate calculus expression.
How can one learn patterns for information extraction without having to annotate a large amount of text? Can we learn from raw (unannotated) text? Several approaches have been developed for such semi-supervised learning. These often are 'bootstrapping' methods, which start with a few 'seed' patterns and gradually discover additional patterns.
One such procedure, described by Sergei Brin, discovers patterns for binary relations. It works as follows
In the automatic procedure, there is a risk that incorrect patterns can lead to incorrect pairs, so the errors can grow rapidly. However, the manual equivalent procedure (discussed in Lecture 8, using a search engine) allows for review as patterns are added and so is less vulnerable to errors.
start with a few pairs of names involved in the relation of interest, and a set of patterns (initially empty)
collect (from a large corpus) examples of patterns which connect several of these pairs
add best pattern(s) to set
collect additional pairs matching these patterns