Structure from Anarchy: Meta Level Representation of Expert System 
Propositions for Natural Language Interfaces. 1 
Galina Datskovsky Moerdler 
Department of Computer Science 
Columbia University 
New York, N.Y. 10027 
phone: (212) 280-8177 
Arpanet: galina@ columbia.arpa.edu 
Abstract: 
In this paper we describe a meta level 
representation used for mapping natural language 
input into propositions of an expert system. This 
representation is based on verb classes that are 
structured hierarchically, with more general 
information encoded in the top level nodes and more 
specific information in the lower level nodes. 
Because of its structure, the representation is able to 
provide a detailed classification of the propositions, 
supplying a basis for defming semantics. It allows 
the system to answer questions about relationships 
between propositions without inferencing, as well as 
to answer questions the expert system could not 
previously handle. 
1 Introduction 
A great deal of work has been done in 
constructing natural language interfaces to well 
structured underlying systems, such as data base 
systems. These natural language interfaces generally 
make use of an assumed system structure, such as a 
schema, to define semantics \[Martin, Appelt and 
Pereira 83; Grosz et. al. 85\] \[Woods et. al. 72; Woods 
73\] \[Kaplan 79\]. On the other hand, almost no effort 
has been made in constructing natural language 
interfaces to systems that do not have such an 
extensive description, e.g. expert systems 2. The lack 
of such a schema means that the~ is no easy way to 
obtain information about propositions 3 of the 
underlying system. Thus, in order to build a natural 
2Throughout thlt paper we •m only ooncemed with expert 
systems that must communicate with their uteri in ot'der to gather 
data befon~ giving advice, inch It Mycin \[Shot~liffe 76\] 
3In the rule of the form IF A and B them C, A, B and C an~ 
propositions. The terms pmlxmtion and fact an~ used 
interch~geably throughout thl. paper. 
language interface to expert systems the semantic 
interpreter must be able to provide the necessary 
structure. In an earlier paper \[Datskovsky Moerdler 
eLal. 87\] we briefly described a semantic interpreter 
that maps user statements into facts of an expert 
system, as well an inference engine for expert 
systems that can efficiendy utilize this input. In this 
paper we discuss the meta level description of the 
expert system propositions, similar to a schema of a 
data base, utilized by the semantic interpreter and 
show how this structure is used in processing of user 
questions. 
Our structure consists of a group of hierarchies 
which are formed from verb categories. The 
hierarchies provide a grouping of the propositions of 
an expert system by topic. For example, all 
propositions that deal with interpersonal relationships 
are grouped under one hierarchy, while those dealing 
with transfer of possession are grouped under 
another. The meaning of a proposition is specified 
step by step, as the hierarchy is traversed, thus 
allowing for mapping of various sentences, or parts 
of sentences into the propositions. To test our 
theories, the approach is currendy being implemented 
as a front end to a small expert system that deals with 
personal income tax matters 4. 
2 Expert Systems vs. Data Base Systems 
Many techniques used in building natural 
language interfaces for data base systems can not 
carry over into the expert system domain because of 
the differences between the two underlying systems. 
'In particular, we are implememin8 our ideat on one module of 
Taxlmrt\[Entor et. tL 85\], m expert syttem designed in 
conjunction with AT&T Bell Laboratories. Thlt module helps 
uteri d~ ~_~mmine whether they can or can not claim an individual as 
• dependent \[Datskovsky Moerdler et.al. 87\] 
tThis research was partially supported by Office of Naval Research grant N00014-82-K-0256. 
121 
A semantic interpreter for a data base system usually 
relies on the regular structure of the dam base as 
encoded in the schema describing it. The schema 
usually describes the fields and tables of a given data 
base and provides such information as the key field 
of a table, the type of data found in each field, 
relationships between the fields (e.g. all the fields of 
a given table describe its key field), etc.. The 
relationships between tables are indicated by 
similarities and differences of their fields. A typical 
natural language interface associates semantics of 
nouns, adjectives and verb phrases of a natural 
language with fields of a data base. Verbs of the 
natural language are also associated with actions that 
can be performed on the tables of a given data base, 
such as Find and Join. 
In contrast, no schema or description is 
available for expert systems. The propositions of an 
expert system may have arbitrary meanings. No 
relationship between the propositions is clearly 
defined. Although meta level structures have been 
built by systems such as Theresias \[Davis 78\], these 
structures are inadequate for defming semantics. 
Theresias provides such information as the 
relationships between antecedents and consequents of 
rules, groupings of rules by their left hand sides, etc.. 
Only one type of representation, schemata, actually 
gives a shallow (3 levels) description of propositions 
(as opposed to rules). However, this information is 
not sufficient for complete semantic definition and a 
more complex structure is required. 
Another major difference is in the function of 
the two systems. A data base system is not expected 
to know or solve a user's problem, but only supply 
the information that the user requests. Consequently, 
an interface to a dam base system must be able to 
retrieve information requested by the user. In figure 
1 we present a typical interaction between 
Lifer-Ladder \[Hendrix eL al. 78\] and a use,. The 
questions here involve retrieval of information from a 
naval data base. 
user:. What is the length mad hull numb= of the 
Constellation? 
system: ....... 
user. the home port? 
system: ... 
Figure 1: Interaction between • us= and 
LIFER-LADDER (Taken from \[Tennmt 81\]) 
On the other hand, an expert system is 
designed to be a problem solver. A use~ consults it 
about an issue and it must gather information in order 
to advise him. In figure 2 we present a typical 
interaction between a user and the Mycin \[Shortliffe 
76\] system (taken from transcripts generated by the 
author). First, in questions 1-6 the system gathers 
information about the patient, such as age, sex, lab 
analysis, etc., and then, after many more questions 
not shown in the figure, makes a recommendation 
based on the gathered data. Note that the menu 
interface predefines the order in which information is 
entered into the system, whereas with a natural 
language interface, information is entered in no 
particular order, i.e it may be imbedded in every user 
input. The addition of new information with every 
user statement means that the expert system has to 
pose fewer questions and that the natural language 
interface must be responsible for managing all the 
new information. Further, the interface may have to 
derive information not only from user statements, but 
also from questions. This means that it has to derive 
the problem to solve, as well as facts that can be used 
for its solution from any given question, and add 
these facts to the data base (or working memory). 
The action of extracting a goal 5 and adding facts at 
the same time has no analogy in a data base system, 
but would be similar to allowing the user to query 
and UlXiate the data base at the same time. 
I) Patient's name: 
** PT244 
2) Age: 
** 8 0 YEARS 
3) Sex: 
* * MALE 
"'6) From what site was the specimen 
for CULTURE-I taken? 
** SPUTUM 
More similar Questions Follow and finally 
a recommendation is made 
\[Determining which drugs are 
desirable for use against the 
(Klebsiella- pneumoniae...\] 
\[REC-I\] My preferred therapy 
recommendation is as follows: 
Figure 2: Mycin Transcript 
~'hmugheut thht paper seth refer to the grab the expert system 
mutt prove, not Ion 8 tam met goal,. 
122 
3 The Structure in More Detail 
To translate user input into facts and goals of 
an underlying expert system, a structure that is able 
to provide a foundation for the translation is 
necessary. This structure must provide the meaning 
of the expert system propositions, relationships 
between them and supply a means of mapping 
semantics of words and phrases into those 
propositions. It is also desirable that such a structure 
he general, and hence to some extent transportable 
from one system to another. 
<Transfer of possession> \[hmn/org, ", ",'\] 
Non Phys.oioj\[-,abstract,','\] Phys. 0hi \[-,concrete, ",'\] 
Money -,m~0n0tary., o1 
Donation \[-,-,0rg,'\] Income \[--,hmn,'\] /\ 
Tax \[-,-,-,paymen~earn~\] Non tax \[-,-,-,payment/given\] 
I~'~_~ 6 .~ (.'~ ~ a~_~_~ .~a ~ 
Figure 3: Partial Tree formed forl~'le 
Transfer of possession category 6 
Our structure consists of a group of hierarchies 
formed from classes of verbs. We have analyzed 
over 90 verbs most common to our domain and 
classified them into 13 categories 7. These categories 
can be used in any domain that requires the verbs 
belonging to them, because they are derived from 
general properties of the verbs s, thus allowing for a 
degree of transportability. Each verb category is 
organized hierarchically where each node of a 
hierarchy is derived from the meanings of one or 
more verbs. A number of selectional restrictions is 
attached to each node indicating constraints on the 
the figure, * stanch for wild card, and . means that the 
feature is inherited flora the patent node. 
7This is not an absolute number. Mo~ ca~4oci~ may be 
n~_,9~__ in other domains wbem • grater nmnbm, of ved~ is 
necessary. 
SThe.~ camgoriea am beusd on works m lingumics, e. 8. 
\[Osso~d 791 and an Roller's ~. Fix • mo~ 
d~cription of tim caleSoriej see \[Datakovsky Moeldler eLM. 87\] 
agent, patient, object and modifier of an input 
sentence (not all four restrictions are specified for 
every category). The hierarchies group propositions 
of an expert system by topic. The leaves of the 
hierarchies contain either expert system facts or 
pointers to other hierarchies, thus forming a 
connectexl forest. The top level nodes of the 
hierarchies provide general classes into which a 
group of propositions of an expert system might fall. 
At the lower levels of the hierarchies the propositions 
are separated into more specific subclasses of a given 
parent node, thus further specifying their meanings. 
At the lowest level, each node points to only one 
proposition thus uniquely def'ming it within its class. 
For example, figure 3 shows the partial 
hierarchy for the Transfer of Possession category. 
The top level node of the hierarchy is derived from 
the properties of the verbs of the general class of 
Transfer of Possession. Verbs from that class have 
pointers to this node and all the propositions that deal 
with transfer of possession can be accommodated by 
this node and the nodes below it. The selectional 
restrictions on this node indicate that the transfer is 
initiated by either a human or an organization and 
that the beneficiary of the transfer, the object being 
transferred, as well as any modifiers can be 
unspecified until some lower level. The two nodes at 
the next level further divide the class of transfer of 
possession verbs and predicates into those dealing 
with physical object transfers and non physical object 
transfers. The \[-\] in the selectional restrictions 
indicate that the feature is inherited from the parent 
node. The restrictions on the two nodes also further 
specify that the object being transferred must be 
concrete in order to take the Plays Obj link and 
abstract in order to take the Non Phys Obj link. At 
the next level, the concept of physical object transfers 
(as embodied by the Plays Obj node) is further 
specified. In this example only one of its children, the 
Money node is shown 9. Again, verbs dealing 
specifically with money transfers may point directly 
to this node. The restriction on the object of the 
transfer must be monetary in order for this node to be 
chosen during parsing. This node is further 
subdivided into Donation and Income, where the 
distinction is made based on the recipient of the 
transfer, since donations are normally given to 
organizations, and income to people. Next, Income 
can come in two forms, Taxable and Non Taxable, as 
indicated by the selectional restrictions of the objects 
9The node I~m 2 other chil&en in the complete tree. 
123 
of the transfer, and finally, the bottom level of the 
hierarchy contains expert system propositions. The 
propositions (?dependent is gross_income ?income) 
and (?dependent is amount_of_support ?support) 
belong to a general class of Transfer of Possession, 
and a more specific class Income, indicating that both 
propositions describe a type of income that is 
generally transferred from one party to another. 
However, because one deals with taxable income and 
the other with non taxable income, these propositions 
are further subdivided into subclasses at the next 
level. 
This kind of gradual division of propositions 
into subclasses not only provides a means for 
mapping user input into facts and goals of an expert 
system, but also allows the system m answer 
questions about relationships between the 
propositions, often without any infercncing. In 
addition, it allows the system to make meta level 
inferences it could not make without the structure. In 
the next section we present a brief description of the 
parsing algorithm and illustrate it with an example. 
3.1 Parsing Algorithm: Overview and 
Example. 
During parsing, an appropriate hierarchy is 
selected according to the definition of the verb in the 
system's dictionary, where each verb can point to any 
level in a hierarchy, and a selectional restriction 
based algorithm is used to traverse the hierarchy with 
the nouns of the sentence guiding the parser down the 
hierarchy, until an expert system proposition is 
reached. The information for this algorithm is 
encoded into each hierarchy, with the restrictions on 
the arguments of the verbs based on noun featm'es 
derived from Roget's thesaurus. The system is 
currently being implemented in Common lisp on a 
Symbolics Lisp Machine. It uses an ATN parser 
which has been modified to call the semantics at 
various points before deciding which a~ to take next. 
Syntax and semantk:s run in parallel, with syntax 
providing a deep struclxn, e of a sentence, and 
semantics supplying infommtion for modifier 
attachment. Although the verb hierarchies are the 
primary source of facts, some facts are derived 
directly from the noun features. 
As an example of how the natural language 
interface derives both Wopositions and goals from 
Yes~No questions posed by the user consider the 
question Can i claim my son who earns a salary of 
$2000?. A trace of the system execution of this 
sentence is shown in appendix I. The trace shows the 
nodes of the different hierarchies considered by the 
algorithm and where the interaction between syntax 
and semantics occurs. It also shows all the predicates 
derived by the system and a complete syntactic parse. 
In yes/no questions the goal is generally indicated by 
the main verb. The syntactic parser identifies claim 
as the main verb of the sentence. The verb claim is 
defined in the system's dictionary as Classification 
<+> Dependency 1°, indicating that the verb belongs 
to the general category of Classification and a more 
specific subnode of that category, Dependency. The 
<+> indicates that the syntactic subject of the 
sentence is the semantic agent. Based on the 
definition of the verb the algorithm enters the 
ClassOeu:ation hierarchy at the Dependency node, as 
demonstrated in stam~ents 1 and 2 of the system 
trace, thus limiting the choice of propositions that this 
input can map into to the general category of 
ClassO~cation and the subclass Dependency (see 
figure 4). Since only one proposition, (?user 
can_claim ?dependent), falls into this classification, it 
is derived as the goal, indicating that the user wants 
to know whether he can or can not claim a dependent 
(the variables of the proposition will later be 
instantiated with the appropriate values). 
The additional information in the relative 
clause states that the dependent earns a salary of 
$2000, or (?dependent is gross_income ?income). To 
derive this additional information, the system selects 
a hierarchy based on the meaning of the verb of the 
relative clause. The verb to earn is defined in the 
dictionary as Transfer of possession <+>, so the 
algorithm enters the Transfer of Possession hierarchy 
(shown in figure 3). The choice of propositions that 
this input can map into is now limited to those in the 
general class of Transfer of Possession. Next, 
because of the feature concrete of the object (two 
thousand dollars) of the sentence the algorithm 
selects Phys Obj as the next node to consider. Based 
on the feature monetary of the word dollars the 
Money node is selected next. The Income node is 
chosen because the recipient of the money has the 
feature human, and finally, because salary is defined 
as payment~earned, the node Tax is selected, since 
earned payments are generally taxable. Finally 
(?dependent is gross_income ?income) is added to 
the working memory. The variables ?dependent and 
t°A/lhough there are other meaninp of the verb, this is the most 
fiequmtly used mcemng in the tax domain, so the system tries this 
caliph/fma. 
124 
?income are later instantiated with son and $2000 
respectively. The derivation of this predicate can be 
seen in statement 5-13 of the system trace in 
appendix I. 
Propositions can also be derived from certain 
noun phrases. In this example, the phrase my son 
indicates the existence of a child-parent relationship. 
The system then checks for agreement between the 
head pronoun I and the possessive my and once this 
agreement is verifies maps the representation of this 
relationship into the proposition (?dependent is 
son_of ?user), as shown in statement 4 of the trace. 
<Classification> \[hmn/org, ", ",'\] 
Secrecy\[-,',secret,'l Categorization{.,', ",'\] 
PeoDle .,h~mn,.,.\] N~ ot)j \[-,obj,','\] j- 
Oepencfency\[-,.,subjection,'\] 
(?user can_claim ?dependent) 
Figure 4: Partial Tree formed for the ClassCication 
category. 
The mapping of natural language into 
propositions of the expert system as demonstrated 
above is possible because of the classification of 
propositions and descriptions of their meanings 
provided by the hierarchies. Note that the hierarchies 
are used to def'me semantics of words of the natural 
language e.g. the verb to earn is directly related to 
the meta level structure, or the Transfer of Possession 
hierarchy. The strncture given by the hierarchies also 
provides a description of the propositions and gives 
similarities and differences between them. For 
example, both propositions (?dependent is 
gross_income ?income) and (?depeadeat is 
amount_of..support ?support) would have the general 
properties of the class Income, with unique features 
of their particular subclasses Tax and Non Tax. This 
unique classification allows for the mapping of the 
input in the above example into the aplxopriate 
proposition. It also allows the system to answer 
questions about the differences twaween the two 
propositions, as shown in the next section. Another 
benefit of this representation is that it provides the 
system with a way of dealing with input sentences 
like My son earns $2000, that do not completely 
specify a particular proposition. The sentence 
indicates that the desired proposition is in the class 
Income, and the system can proceed to specify the 
appropriate subclass by posing questions to the user 
without any additional inferencing on the part of the 
expert system. This particular capability of the 
algorithm will be discussed in greater detail in future 
work. 
3.2 Other Questions that can be Answered 
from the Hierarchies 
The hierarchies allow the system to handle a 
number of questions that could not be previously 
handled by the expert system, and answer other 
questions without invoking the inference process. In 
particular, these include questions that deal with 
relationships between facts and comparisons between 
sessions, as well as questions requiring general 
information. 
User: My daughter receives a stipend of $5000, 
while my son gets a salary of $2000. 
WHY is my daughter's tax situation different 
from my son's? 
System: Your daughter's stipend is non taxable 
income. 
\[Answered by looking at the Income node of the 
Transfer of Possession hierarchy, where the 
two paths diverged.\] 
Figure 5: A Question Answered from the 
Transfer of Possession Hierarchy 
AS an example of questions that can be 
answered without invoking the inference process, 
consider the hypothetical example in figure 4 where 
the user tells the expert that his daughter receives a 
stipend of $5000, which translates into the 
proposition (daughter is amount_of..support 5000), 
since stipend is defined in the dictionary as 
payment.given. The fact that his son has a salary of 
$2000 translates into the proposition (son is 
gross income 2000). To answer the WHY question 
the system could check where the derivation paths for 
the two sets of inputs diverged, and the difference 
between the two subclasses would constitutes the 
answer. In this example the paths diverge at the 
Income node of the Transfer of possession hierarchy, 
thus the answer can be supplied by simply examining 
the hierarchy. 
125 
The question in the first example required both 
a comparison between two derivation paths as well as 
the knowledge of the differences between two 
propositions. As a second example consider the 
question What kinds of family relationships are 
recognized by the tax code? This question is about 
general properties of the tax code and could not be 
handled by the expert system without the natural 
language interface, even though all the necessary 
information was already available in the system. To 
answer this question it is enough to search the 
hierarchies for a Relationship node with a child node 
that describes family relationships. Such a parent- 
child pair is found in the Possession hierarchy (see 
figure 6). The answer returned would consist of all 
the children found under this pair. 
<Possession> \[l~mn/org, ", "\] 
t(~al \[-,-,l~.rol\] Family \[hmn,mL'\] Business{-,.'\] 
Spouse Parent... Child 
Figure 6: Partial Tree formed for the Possession 
category. 
The question handling algorithm is currently 
under design. To process WH questions the system 
must ftrst be able to deter\[nine whether it can be 
answered from the hierm~hies, or whether the 
inference engine of the expert system should be 
invoked. Many of the necessary clues that indicate 
the question type have been identified, however there 
is still some more work to be done on this, as well as 
on the implementation of the module. It is clear, 
however, that the hierarchies give the system the 
ability to handle many more types of questions than 
the expert system alone could handle, and in many 
instances allow questions to be answered without 
invoking the inference process of the expert system. 
4 Comparison with Previous Work: NLIs 
to Expert Systems and Other Work in 
Semantics 
There has been some effort to construct 
natural language interfaces to expert systems, namely 
Prospector \[Duda eL al. 79\] and Xcaiibur \[Cartxmell 
eLal. 83; Carbonell and Hayes 84\]. Prospector is one 
of the fast expert systems to communicate with its 
users in natural language. During the consultation 
the user simply describes what has been discovered at 
a given site by using patterns, built with the help of 
the Lifer \[Hendrix et. al. 78\] system, of the form 
"There is <deposit>', "There may be <deposit>", etc. 
There is not much published information that 
describes Prospector's natural language module. We 
can only hypothesize that a very simple and limited 
set of sentences is accepted by the system based on 
sample system sessions. 
Xcalibur's interaction with the user greatly 
resembles that of a natural lang~mge interface to a 
data base system. Unlike systems such as Mycin, 
Xcalibur does not do most of the asking. It is not 
responsible for solving the user's problem, but rather 
the user has to know what he wants and query 
accordingly. Most expert systems are designed to 
solve a user's problem, and this property must be 
reflected in the interface. Xcalibur does not seem to 
be suitable as an interface for such systems because it 
is designed to retrieve information rather than solve a 
problem. 
4.1 Other work in Semantics 
Our work draws on Palmer's \[Palmer 85\], but 
is different from it in several ways. Palmer's 
Inference-driven sexnantic analysis is specifically 
designed for a finite, well-defined, i.e. limited 
domain. The main element of her approach is a set of 
partially instantiated logical terms, or semantic 
propositions, which capture the different 
relationships that can occur in a given domain. 
Unlike Palmer's work, our interpreter deals with a 
complex real world domain. It also makes a greater 
separation between domain specific and domain 
independent knowledge to allow for a degree of 
transportability. Also, while our semantics provides 
a hierarchical organization, Palmer's does not. 
Other work that has influenced our own also 
includes that of Graeme Hirst \[Hirst 83\] and Steve 
Lytinen tLytinen 84\]. One of the main differences 
between our work and the work mentioned above 
(including Palmer's) is that our semantics imposes a 
structure on top of an unstructured underlying 
system, which is not the goal of the work mentioned 
above. 
126 
5 Possible Automation of Hierarchy 
Design 
The lack of automatic construction of the 
hierarchies and automatic classification of 
propositions in them is currently a limitation in our 
system. If, for a given domain, a certain tree has to 
be extended, such extension will have to be done by 
hand. Also, propositions have to be hand encoded in 
the hierarchies. This makes transportability to other 
domains more difficult. After the top level categories 
are selected, the rest of the nodes of the hierarchies 
and the propositions, as well as the selectional 
restrictions can not be done interactively. However, 
we feel that the hierarchies lend themselves to 
automation construction by an Expert System Expert, 
because they are based on the linguistic properties of 
the verbs in the domain, as well as on the knowledge 
of the meanings of propositions. 
In the future, we would like to design a 
customization phase similar to that of Team \[Martin, 
Appelt and Pereira 83; Grosz et al. 85\] and 
Teli \[-Ballard 86\]. With such a customization phase, a 
given expert, such as an Expert Systems Expert, can 
spend several horn's automatically building up the 
necessary parse Irees for a given domain. We feel 
that such a module would geatly enhance the system 
and make it much more usable. 
6 Conclusions and Future Research 
In this paper we presented a slructure for 
expert systems, similar to a dam base schema, that 
facilitates construction of natural language interfaces. 
This structure is based on verb classification and 
hierarchical structuring within each categocy. The 
hierarchies provide a grouping of expert system 
propositions into classes, thus capturing the 
similarities and differences betweea the pmlx~itions. 
This grouping provides a mal~ing between user input 
and the propositions of the expert system, as well as a 
mechanism for dealing with several types of 
questions without additional ~pert system 
inferencing. The structme provides a mechanism for 
answering questions that could not be previously 
handled by the expert system. It also provides a 
flexible and somewhat general mapping allowing for 
a degree of wansportability. 
One of our primary goals is to complete the 
implementation of our ideas. Processing of 
statements and yes/no questions has been fully 
implemented and the work on petagraph parsing and 
handling of semanUcally incomplete input is our 
current focus. In the future we plan to add such 
features as complete WH question processing and an 
automatic hierarchy construction algorithm. 
7 Acknowledgments 
I would like to thank my advisor, Kathleen 
McKeown for all her help and guidance in this work 
and Robert Ensor of AT&T for his helpful comments. 
Appendix I 
(process '((can I claim my son who earns a salary of 
twothousand donars))) 
1. In Tree: CLASSIFY 
2. Considering the children of DEPENDENCY 
3. the proposition that was derived is 
((?USER ICAN_CLAIMI ?DEPENDENT)) 
back to syntax... 
4. the proposition derived from the noun phrase 
(MY SON) is 
(?DEPENDENT IS ISON_OFI ?USER) 
5. In Tree: TRANS OF_POS 
6. Considering the children of TRANS OF POS 
beck to syntax... 
7. Considering the children of TRANS_OF__POS 
back to syntax... 
8. Considering the children of TRANS_OF..POS 
9. Considering the children of IPHYS_OBJI 
10. Considering the children of MONEY 
11. Considering the children of INCOME 
12. Considering the children of TAX 
13. the proposition that was derived is 
((?DEPENDENT IS IGROSS_INCOMEI 
?INCOME)) 
back to syntax... 
((S (SLrBJ (NP (DET NIL) (DESCRIBERS NIL) (HEAD ((PRON 
I))) (NUMBER SING)(CONJ NIL) (gEM (HUMAN) NIL)) (QUAI.2FIERS 
NIL) (Q~ON NO)(CASE OIMECTIVE))) (AUXS (CAN)) (TENSE 
PRES) (MAINVERB CLAIM) (SEM-~ ((CLASSIFY ÷ 
DEPENDENCY))) (ADVERB NIL) (IND-OBJ NIL) (SUBCONJ NIL) 
(D-ore (NP (Dirt MY) (D~ERS NIL)(HEAD ((NOUN SON))) 
(NUMBER Sn~G)(CONJ Nn.) (SEM ((HUMAN MALE RELA'nVE 
127 
CHILD) NIL)) (QUALIHF, RS ((S (SUBJ (NP (DET MY) (DESCRIBERS 
NIL) (HEAD ((NOUN SON))) (SEM ((HUMAN MALE RELATIVE 
CIiXLD) NIL)) (QUESTION NO))XAUXS NIL) (TENSE PRES) 
(MAINVERB EARN) (SEM-MVERB ((TRANS_OF_POS +))) (ADVERB 
NIL) (BrD-OBJ NIL)(SUBCONJ Nn.)(D-OBJ (NP (DET A) 
(DESCRmERS NIL) (HEAD ((NOUN SALARY))) (NUMBER 
SING)(CONJ NIL) (SEM ((CONCRETE MONETARY PAYMENT- 
EARNED) Nn.)) (QUAL~RS (0'I' (PREP OF) (PRF.P-OW (NP (DET 
~'~I.) (DESOtmERS tTWOTHOUSAND)) (HEAD ((NOUN DOLLAR))) 
(NUMBER PLURALXCONJ NIL) (SEM ((MONETARY) (CONCRETE 
NUMBER))) (QUALu.\[F_.R$ NILXQUESTION NO) (CASE 
OBJECTIVE)))))) (QUESTION NO) (CASE OBJECTIVE))XADJCOMP 
NIL) (MODIF NIL) (TYPE WH-RELAT\[VE) (QUF~TION-k'\[x~Mlgsrr 
NILXVOICE At=rIVE) (CONJ NIL) (CSENT NIL)))) (QUESTION NO) 
(CASE POSSESSIVE))) (ADJCOMP N1L)(MODn ~ NIL) (TYPE 
\[~TERRGO&TIVE) (QUESTION-ELEMENT (YES-NO)) (VOICE 
ACT\[VE)(CONJ N\]L)(CSENT NIL))) 
THE GOAL IS: ('RISER EAN_O.AIM\] 7D~EYr) 

References 

Ballard B. Semantic Acquisition in 
TELI: A Transportable, User-Customized Natural 
Language Processor. In Proceedings of twenty-first ACL 
conference. 1986. 

Carbonell J., Hayes, P. 
Recovery Strategies Parsing Extragrammatical 
Language. Technical Report CMU-CS-84-107, 
Carngie-Mellon University, 1984. 

Carbonell J., Boggs W.M., Mauldin 
M., Anick, P. Xcalibur Project Report 1. 
Technical Report CMU-CS-83-143, Carngie-MeLlon 
University, Digital Equipment Corp., 1983. 

Clancey, W. The Epistinxalogy of a 
Rule-Based Expert System - a Framework for Explanation. 
Art~tcial Intelligence 20,1983. 

G. Datskovsky Moerdler, 
K. McKeown, J.R. Ensor. Building Nat, u-al Language 
Interface to Expert Systems. In Proceedings of the IJCAI. 
1987. 

Davis, R. Knowledge Acquisition in 
Rule-Based Systems-Knowledge About Representation as a 
Basis for System Construction and Maintanance. Panern 
Directed Inference Systems. Academic Press" 1978. 

Duda, R., Gasching, J., Hart, P. Model 
Design in the Prospector Consultant System for Mineral 
Exploration. In Michie, D. (editor), F..wert Systona in the 
micro-electronic age. Edinburgh University Press, 1979. 

Ensor, Gabbe and Blumenthal. Taxpert - A Framework for Exploring Interactions Among Experts. 
1985. in preparation. 

Grosz, B., Martin, P., AIvelt, D., 
Pereira, F.,Team: An Experiment in the Design of 
Transportable Natured Language Interfaces. Technical 
Report, SKI International 1985. 

Henddx, G., Sacerdoti, E., Sagalowicz 
D., Slocum Jl)eveloping a Natural Language Interface to 
Complex Data. ACM Transactiona on Database 
Systems, 1978. 

Hitst, G. Semantic Interpretation 
Against Ambiguity. PhD thesis, Brown University, 1983. 

Kaplan, SJ. Cooperative Responses 
From a Portable Natural Language Data Base Query 
System. PhD thesis, University of Pennsylvania, 1979. 

Levin, B. Lexical Semantics in Review: 
An Introduction. In Levin, B. (editor), Le.xical Semantics 
in Review. M1T, 1985. 

Lytinen S.L. The Organization of 
Knowledge in a Multi-lingual Integrated Parser. PhD 
thesis, Yale University, 1984. 

Matin, P., Appelt, D., 
Peteira, F. Transportability and Generality in a 
Natural Language Interface System. In 
Proceedings Of lJCAI. 1983. 

Osgood, Charles, E. Focus on Meaning 
Volume I: Explorcaions in Semantic Space. Mouton 
Publishers, 1979. 

Palmer, M. Inference-Driven Semantic 
Analysis. In Proceedings of the AAAI. 1983. 

Stone Palmer, M. Driving Semantics for 
a Limited Domain. PhD thesis, University of Edinburg, 
1985. 

PoLlack, M.E. Generating Expert 
Answers Through Goal Inference. Technical Report, SRI 
International, 1983. 

ShoRliffe, E.H. Mycin: A rule-based 
computer program for advising physiciana regarding 
anilimicrobial therapy selection. PhD thesis, Stanford 
University, 1976. 

Tennant, H. Nat"red Language 
Processing. PetoceLli Books, USA, 1981. 

Nash-Webber, B., Verbs of 
Composition. 1971.Harvard University, 1971. 

Woods" W.A. An Experimental parsing 
System for Transition Network Grammars. In Rustin 
(editor), Natural Language Processing. Algorithmic Press, 
1973. 

Woods W., Kaplan R., Nash-Webber B. 
The Lunar Sciences Natured Lnaguage Information System: 
Final Report. Technical Report 2378, BBN, Cambridge, 
Mass" 1972. 
