Event Coreference for Information Extraction 
Kevin Humphreys and Robert Gaizauskas and Saliha Azzam 
Department of Computer Science 
The University of Sheffield 
Regent Court, 211 Portobello Street 
Sheffield S1 4DP UK 
{K. Humphreys, R. Gaizauskas, S. Azzam}@dcs. shef. ac. uk 
Abstract 
We propose a general approach for per- 
forming event coreference and for con- 
structing complex event representations, 
such as those required for information ex- 
traction tasks. Our approach is based on 
a representation which allows a tight coup- 
ling between world or conceptual model- 
ling and discourse modelling. The rep- 
resentation and the coreference mechanism 
are fully implemented within the LaSIE 
information extraction system where the 
mechanism is used for both object (noun 
phrase) and event coreference resolution. 
Indirect evaluation of the approach shows 
small, but significant benefit, for informa- 
tion extraction tasks. 
1 Introduction 
Much recent work on anaphora has concen- 
trated on coreference between objects referred 
to by noun phrases or pronouns (see, e.g., 
Botley and McEnery (1997)). But coreference in- 
volving events, expressed via verbs or nominalised 
verb forms, is also common, and can play an import- 
ant role in practical applications of natural language 
processing (NLP) systems. 
One application area of increasing interest is 
information extraction (IE) (see, e.g., Cowie and 
Lehnert (1996)). Information extraction systems 
attempt to fill predefined template structures with 
information extracted from short natural language 
texts, such as newswire articles. The prototypical 
IE tasks are those specified in the Message Under- 
standing Conference (MUC) evaluations (DARPA, 
1995; Grishman and Sundheim, 1996). In these ex- 
ercises the main template filling task centres around 
a 'scenario' which is defined in terms of a key event 
type and various roles pertaining to it. Examples 
of scenarios used in previous MUCs include joint 
venture announcements, microprocessor product an- 
nouncements, terrorist attacks, labour negotiations, 
and management succession events. In order not to 
spuriously overgenerate event instances and to prop- 
erly acquire all available role information, it is cru- 
cial that multiple references to the same event be 
correctly identified and merged. While these con- 
cerns are of central importance to IE systems, they 
are clearly of significance for any NLP system, and 
more broadly for any computational model of nat- 
ural language. 
A few concrete examples will make the issues 
clearer 1. A management succession event (as used 
in MUC-6) may involve the two separate events of a 
corporate position being vacated by one person and 
then filled by another. For an event to be considered 
reportable for the IE task, the post, the company 
and at least one person (either incoming or outgo- 
ing) must all be identifiable in the text. 
The first thing to note here is that while manage- 
ment succession events are sometimes reported as 
single, simple events, as in 
(1) Mr. Jones succeeds M. James Bird, 50, as pres- 
ident off Wholistic Therapy. 
more frequently multiple aspects or sub-events of 
a single succession event are identified in separ- 
ate clauses by separate verb phrases or nominalised 
forms: 
(2) Daniel Wood was named president and chief ex- 
ecutive officer off EFC Records Group, a unit off 
London's Spear EFC PLC. He succeeds Charles 
Paulson, who was recently made chairman and 
chief executive officer off EFC Records Group 
North America. 
1All examples in this paper are taken from the MUC- 
6 Wall Street Journal corpus with names of individuals 
and companies changed. 
75 
(3) The sell-o# followed the resignation late 
Monday o\] Freddie Heller, the president o/ 
Renard Broadcasting Co. Yesterday, Renard 
named Susan B. Kempham, chairman o/ 
Renard Inc. 's television production arm, to suc- 
ceed him. 
Both of these pairs of sentences refer to a single 
management succession event (though the second 
sentence in 2 also identifies a further one). Such 
event/sub-event relations are similar to the famil- 
iar part-whole or related-object anaphora exempli- 
fied in sentences such as The airplane crashed a~ter 
the wings/ell off or When John entered the kitchen 
the stove was on (Allen, 1987). 
The second thing to note is the variety of surface 
forms used to refer to events. Events are referred to 
by verb phrases in main clauses (1 above), and in re- 
lative clauses (second sentence in 2) or subordinate 
clauses. They may be referred to through nominal- 
ised forms (resignation in 3 above) or through infin- 
itival forms in control sentences (second sentence in 
3). When there are multiple references to the same 
event, antecedent and anaphor appear to be able to 
adopt all combinations of these forms 2. 
This paper discusses an approach to hand- 
ling event coreference as implemented in the 
LaSIE information extraction system (Gaiza- 
uskas et al., 1995; Gaizauskas and Humphreys, 
1997b). Within this system, event coreference 
is handled as a natural extension to object core- 
ference, outlined here and described in detail in 
Gaizauskas and Humphreys (1997a). Both mechan- 
isms are handled within a general approach to dis- 
course and world modelling. 
In the next section we give a brief overview of the 
LaSIE system. Section 3 describes in more detail the 
approach to world and discourse modelling within 
LaSIE and Section 4 details our coreference proced- 
ure. In Section 5 we discuss a particular example 
in detail and show how our approach enables us to 
correctly corefer multiple event references. Section 
6 presents results of an approach to evaluating the 
the approach and Section 7 concludes the paper with 
some general discussion. 
2 LaSIE Overview 
The Large Scale Information Extraction system 
(LaSIE) has been designed as a general purpose IE 
research system, initially geared towards, but not 
2While no extended study has been carried out, it 
appears that in newswire texts nominalised forms are 
less likely to appear in the first reference to an event, 
and more likely to appear in subsequent references. 
solely restricted to, carrying out the tasks specified 
in MUC-6: named entity recognition, coreference 
resolution, template element filling, and scenario 
template filling tasks (see DARPA (1995) for further 
details of the task descriptions). In addition, the 
system can generate a brief natural language sum- 
mary of any scenario it has detected in a text. All 
these tasks are carried out by building a single rich 
discourse model of the text from which the various 
results are read off. The system is a pipelined archi- 
tecture which processes a text one sentence at a time 
and consists of three principal processing stages: lex- 
ical preprocessing, parsing plus semantic interpreta- 
tion, and discourse interpretation. The overall con- 
tributions of these stages may be briefly described 
as follows (see Gaizauskas et al. (1995) for further 
details): 
lexical preprocessing reads and tokenises the raw 
input text, tags the tokens with parts-of- 
speech, performs morphological analysis, per- 
forms phrasal matching against lists of proper 
names; 
parsing and semantic interpretation 
builds lexical and phrasal chart edges in a 
feature-based formalism then does two pass 
chart parsing, pass one with a special named en- 
tity grammar, pass two with a general grammar, 
and, after selecting a 'best parse', constructs a 
predicate-argument representation of the cur- 
rent sentence; 
discourse interpretation adds the information 
from the predicate-argument representation to 
a hierarchically structured semantic net which 
encodes the system's world model, adds addi- 
tional information presupposed by the input, 
performs coreference resolution between new 
and existing instances in the world model, and 
adds any information consequent upon the new 
input. 
2.1 MUC-6 Coreference Performance 
MUC-6 included a quantitatively evaluated corefer- 
ence task, which required participating systems to 
propose coreference annotations for a set of texts. 
These annotations were then automatically scored 
against manually produced annotations for the same 
texts. The performance of the LaSIE system in this 
coreference task was 51% recall and 71% precision. 
This compares favourably with the highest scoring 
MUC-6 systems: the highest recall system scored 
63% recall and 63% precision; the highest precision 
system scored 59% recall and 72% precision. Recall 
76 
here is a measure of how many correct (i.e. manually 
annotated) coreferences the system actually found, 
and precision is a measure of how many coreferences 
the system proposed were actually correct. For ex- 
ample, suppose there are 100 real coreference rela- 
tions in a corpus and a system proposes 75, of which 
50 are correct. Then its recall is 50/100 or 50% and 
its precision is 50/75 or 66.7%. 
The MUC-6 definition of the coreference task in- 
cluded several forms of NP coreference, not only pro- 
nominal relations. However, it did not include event 
coreference, which can be measured only indirectly 
via the information extraction task results, a topic 
to which we return in Section 6. 
3 Discourse Interpretation 
The LaSIE system's 'world' or domain of interest is 
modelled by an inheritance-based semantic graph, 
using the XI knowledge representation language 
(Gaizauskas, 1995). In the graph classes of objects, 
events, and attributes appear as nodes; each node 
may have associated with it an attribute-value struc- 
ture and these structures are inherited down the 
graph. The higher levels of the graph, or ontology, 
for the management succession task have the struc- 
ture shown in Figure 1. Two simple attribute-value 
structures are also shown in the graph, connected 
by dashed lines to the nodes with which they are 
associated. 
Attribute-value structures are just sets of 
attribute:value pairs where the value for an 
attribute may either be static, as in the pair 
animate:yes, which is associated with the person 
node, or dynamic, where the value is dependent on 
various conditions, the evaluation of which makes 
reference to other information in the model. Cer- 
tain special attribute types, presupposition and 
consequence, may also return values which are used 
at specific points to modify the current state of the 
model. 
As a discourse is processed, discourse entities (ob- 
jects and events introduced by the text) are added as 
new nodes in the graph beneath their parent class 
and have associated with them an attribute-value 
structure containing both inherited and discourse- 
supplied attributes. This process may involve hypo- 
thesising new implicit entities if they are not avail- 
able explicitly in the text, or have not been dis- 
covered by the parser, but are required role play- 
ers for a given event type. Knowledge about re- 
quired roles is represented via attributes in the 
world model. For example, in Figure 1 we see that 
a retire event requires a logical subject of type 
person and an entity of this type will be hypothes- 
ised if it is not available from the text. 
4 Coreference Resolution 
After each sentence in a text is added to the 
'world model', gradually forming a discourse-specific 
model, a coreference procedure is applied to attempt 
to resolve, or merge, each of the newly added in- 
stances with instances currently in the discourse 
model. Coreference resolution is performed by com- 
paring instances from several candidate sets, each of 
which is a set of pairs of instances where one element 
is an instance from the current input sentence and 
the other an instance occurring earlier in the text, 
which may be coreferential. The algorithm proceeds 
as follows for each instance pair being considered: 
1. Ensure semantic type consistency 
To determine semantic consistency requires es- 
tablishing a path in the semantic graph between 
the semantic types of the two instances. If a 
path can be found a semantic similarity score 
is calculated using the inverse of the length of 
the path (measured in nodes) between the two 
types. 
For event instances, a path is valid if both event 
types are dominated by a task-specific top node, 
i.e. both types must be potential sub-events of 
an event required by the current IE template. 
For example, 'hire' and 'retire' are both sub- 
events of the 'succession' event in the ontology 
sketched above. 
For instances of the object class, a path is valid 
if the two types stand in a dominance relation, 
i.e. the types are ordered in the ontology. For 
example, 'company' is a sub-class of 'organisa- 
tion' so these type are ordered (and have a se- 
mantic similarity score of 0.5). 
If no valid path can be found the attempt to 
resolve the two instances is abandoned. 
. Ensure attribute consistency 
Certain attributes, e.g. animate and time, are 
specified in the ontology as taking a single fixed 
value for any particular instance. If two in- 
stances being compared have a common attrib- 
ute of this type, the values must be identical or 
the attempted resolution is abandoned. 
Type specific coreference constraints are then 
examined by attempting to inherit a distinct 
attribute. If this attribute can be derived from 
any of the instances' superclasses the attempted 
resolution of the current pair is abandoned. 
77 
person 
entity 
object event 
organisation date succession single-valued / \ /\ /\ 
company government incoming outgoing animate count /\ 
retire resign 
animate: yes I 
I 
attribute 
multi-valued /\ 
name near 
lsubj_type: person 
Figure 1: Upper ontology for the management succession task 
Constraints on the various event types are de- 
tailed in the following section. 
. Calculate a similarity score 
The semantic similarity score is summed with 
an attribute similarity score to give an overall 
score for the current pair of instances. The at- 
tribute similarity score is established by finding 
the ratio of the number of shared multi-valued 
attributes with compatible values, against the 
total number of the instances' attributes. 
After each pair m a candidate set has either been 
assigned a similarity score or has been rejected on 
grounds of inconsistency, the highest scoring pair (if 
any score at all) are merged in the discourse model. 
If several pairs have equal similarity scores then the 
pair with the closest realisations in the text will be 
merged. The merging of instances involves the re- 
moval of the least specific instance (i.e. the highest 
in the ontology) and the addition of all its attributes 
to the other instance. 
4.1 Event Coreference 
The constraints on events as used in Step 2 of the 
general coreference algorithm above can be associ- 
ated with any event node in the ontology, and will 
then be inherited by all instances of all sub-event 
types. The constraints currently used can be cat- 
egorised in the following way: 
1. General task-independent constraints are as- 
sociated with the top-level event node. For 
example two event instances are defined as 
distinct (i.e. not coreferential) if they have 
incompatible times. At present this simply 
means that two events with different tenses can- 
not be resolved, but clearly a more detailed 
model of event times is required, particularly as 
Crowe (1996) shows how temporal phrases are 
consistently useful in distinguishing and recog- 
nising events 3. 
2. General task-specific constraints are, for the 
management succession task, associated with 
the succession_event node. For example, the 
constraint that two instances must be distinct if 
they involve different organisations or different 
management positions. 
3. More specific constraints are represented at 
lower and possibly verb-specific nodes. For ex- 
ample, an incoming_event (e.g. hire, promote) 
is distinct from a changeover_event (e.g. re- 
place, succeed) if the former's logical object is 
distinct from the latter's logical subject. 
The determination of distinct or compatible event 
roles requires the application of the coreference 
mechanism to instances of the object class (the role 
players in the event). The same algorithm is used 
but the inherited constraints will be those associ- 
ated with the object nodes in the ontology. For ex- 
3It is possible to represent a time scale within the 
current XI formalism and then associate each input event 
with a point on the scale. Each point can be treated as a 
potential interval and be expanded to include the times 
of sub-events. The representation and use of this more 
detailed model is currently under investigation. 
78 
ample, indefinite noun phrases cannot be anaphors, 
pronouns should be resolved within the current para- 
graph, definite noun phrases within the last two 
paragraphs, etc. Full details and an evaluation of 
the coreference constraints on object instances can 
be found in Gaizauskas and Humphreys (1997a). 
The constraints above are similar to those used in 
the FASTUS IE system (Appelt et al., 1995) and by 
Sown (1984), where the merging takes place between 
template structures, considering special conditions 
for the unification of variables in template slots. 
However, the general approach here has more in 
common with Whittemore and Macpherson (1991) 
or Zarri (1992), where event merging is carried out 
within the underlying knowledge representation. 
5 A Worked Example 
This section describes the operation of the general 
coreference mechanism for the example (3) presen- 
ted in the introduction, concentrating on the effect 
of the various constraints on event instances. We 
reproduce the two sentences in (3) here: 
(3a) The sell-off followed the resignation late 
Monday of Freddie Heller, the president of 
Renard Broadcasting Co. 
(3b) Yesterday, Renard named Susan B. Kempham, 
chairman of Renard Inc. % television production 
arm, to succeed him to succeed him. 
The full semantic representation of these sentences 
as produced by the parser/semantic interpreter for 
input to the discourse interpreter is: 
Sentence 3a 
sell-off(e2), number(e2,sing), det(e2,the), 
follow(el), time(el,past), 
lsubj(el,e2), lobj(el,e3), 
resignation(e3), number(e3,sing), 
det(e3,the), 
date(e5), name(e5,'Monday'), 
person(eT), name(eT,'Freddie Heller'), 
title(eS,president), 
company(el0), 
name(el0,'Renard Broadcasting Co.') 
Sentence 3b 
yesterday(ell), number(ell,sing), 
name(e13,'Renard'), 
name(el2), time(el2,past), 
lsubj(e12,e13), 
person(el5), name(el5,'SusanB. Kempham'), 
apposed(el5,el6), 
title(elG,chairman), 
arm(el8), number(el8,sing), 
qual (el8, el9) , 
production(el9), number(el9, sing), 
qual(el9,e20), of(el9,e21), 
television (e20) , (~, 
company(e21), name(e21,'Renard Inc.'), 
succeed(el4) , time(el4,present), 
isubj(el4,el5), lobj(e14,e22), 
pronoun (e22 ,him) 
The nominalisation of the verb resign in (3a) leads 
to the presupposition of an outgoing_even% which 
in turn leads to hypothesised objects for a related 
person, post and organisation (these presuppositions 
are stored as attributes of the outgoing_event in 
the world model). The coreference mechanism will 
then be applied to these objects and, in this case, 
will be able to resolve all three within the same sen- 
tence. The resign event therefore forms a complete 
succession event for the management succession IE 
task. 
Both verbs in (3b), the incoming_event name 
and the changeover_event succeed, will cause the 
introduction of succession event instances into the 
discourse model, each of which will cause the hy- 
pothesis of a related person, post and organisation. 
Attributes of the name event will add additional con- 
straints to its hypothesised objects, including the 
specification that the organisation should be a po- 
tential subject of the verb, the person a potential 
logical object, and the post a potential complement. 
Objects with the required features will be found by 
the coreference mechanism for the organisation and 
person, but not the post. The succeed event will 
also cause the hypothesis of an additional person, 
with the constraints that one must be incoming, 
and a potential logical subject of the verb, and the 
other outgoing, and a potential logical object. The 
succeed event's hypothesised organisation and post 
will be resolved with the same objects as the resign 
event from the previous sentence. 
The general constraints on coreferential succession 
events are therefore satisfied for the succeed and 
resign events, and the restrictions on the more spe- 
cific subclasses must then be considered. The rel- 
evant restriction here is that a changeover_event 
must share its logical object with the logical subject 
of an outgoing_event. This will require the applic- 
ation of the coreference mechanism for objects to 
resolve the pronoun him. A correct resolution with 
Freddie Heller will then allow the two events to be 
resolved. 
The succeed and name events will also be 
resolved similarly, using the restriction that a 
changeover_event must share its logical subject 
79 
Succession Events Overall 
Recall Precision Recall , Precision I Combined 
No Event Corer 66% 72% 42% 59% 48.88% 
With Events Coref 65% 77% 42% ~ 60% \] 49.40% 
Table 1: Upper Ontology.for the management succession task 
with the logical object of an incoming_event. In 
this case the infinitive form of the succeed verb will 
have no explicit logical subject, but one will be hypo- 
thesised and resolved with the best antecedent of the 
required type (person), here Susan B. Kempham. 
The two events can therefore be merged, to result in 
the representation of a single succession event with 
Freddie Heller outgoing and Susan B. Kempham in- 
coming. 
6 Evaluation 
We have not been able to carry out direct evaluation 
of our approach to event coreference. To do so would 
require manually annotating coreferential events in 
a corpus of significant size, and we have not had the 
resources to do so. However, we have attempted to 
gain some indirect measure of the successfulness of 
the approach by toggling event coreference on and 
off and observing the effect on the ability of the sys- 
tem to fill MUC-6 management succession templates 
correctly. The hypothesis here is that effective event 
coreference will lead to higher scores in the template 
filling task for at least two reasons. First, role play- 
ers in events (which become slot fillers in the scored 
templates, e.g. persons and organisations) should 
become available due to event coreference. Second, 
spurious succession events should be eliminated due 
to proper event coreference. 
The MUC-6 management succession scenario 
task involved filling an object-oriented template 
consisting of five objects, each with associated 
slots (twenty slots in total). The top level ob- 
ject was a template object and contained one 
or more succession_event objects which in turn 
contained an organization object and one or 
more in_and_out objects, themselves containing 
organization and person objects (a precise defin- 
ition of the template and the task can be found in 
DARPA (1995)). 
Table 1 shows the gross results of running the sys- 
tem against the 100 articles in the MUC-6 scenario 
task test corpus. Our system is easily reconfigured 
to run with or without attempting event corefer- 
ence. The two rows in the table show the effects 
without and with event coreference. The 'Overall' 
column show the effects on the overall scenario tem- 
plate filling task, i.e., on recall and precision scores 
for all objects and slots in the templates. The 'Suc- 
cession Events' column shows the effect just for the 
succession_event objects in the templates, and is 
therefore a more direct measure of template filling 
performance where we might expect event corefer- 
ence to have an effect. 
As can be seen from the table, the effect over- 
all is not particularly significant. However, the ef- 
fect on succession events alone is more substan- 
tial, with precision going up five percentage points 
and recall dropping only one, when event corefer- 
ence is switched on. Closer examination revealed 
that the event coreference mechanism successfully 
avoided the proposal of 11 spurious succession events 
in the evaluation corpus, which included 196 possible 
events. 
We stress that this is a crude measure of our 
event coreference algorithm - really just an indica- 
tion of its utility in the information extraction task. 
However, even as such, it shows that the algorithm 
is performing correctly, on balance, and that event 
coreference is worth addressing in an IE system. 
7 Conclusion 
Event coreference is more complex than object core- 
ference because of the requirement that objects 
filling particular event roles in two possibly corefer- 
ential events must themselves be coreferential. Core- 
ferring events is therefore logically secondary to core- 
ferring objects 4. 
The approach we describe here provides a 
very general and powerful mechanism for per- 
forming event coreference and for constructing 
complex event representations, such as those re- 
quired for information extraction tasks. Within 
information extraction the problem has typic- 
ally been addressed by attempting to merge, or 
unify, extracted templates (e.g. Sown (1984) or 
Appelt et al. (1995)), but a more generally useful 
4Of course in some events, roles may be filled by other 
events, but this complication does not affect the basic 
point that object coreference is primary and event core- 
ference dependent upon it. 
80 
mechanism will operate within a more general rep- 
resentation. Our approach can be compared to that 
of Whittemore and Macpherson (1991) who discuss 
incremental building of event representations within 
a modified form of DRT (Kamp, 1981). However, 
the representation used here is preferred because it 
allows a tighter coupling between world or concep- 
tual modelling and discourse modelling. 
The representation and the coreference mechan- 
ism are fully implemented within the LaSIE in- 
formation extraction system and are currently be- 
ing extended to make use of a richer model of event 
times, the importance of which is demonstrated in 
Crowe (1996). The mechanism described here is 
used in the LaSIE system for both object and event 
coreference, treating the different types simply as in- 
stances subject to differing constraints, where con- 
straints can be easily represented at any level of gen- 
erality. Our evaluation, while far from exhaustive, 
shows that addressing event coreference can clearly 
result in real benefits for IE systems. 
8 Acknowledgements 
We thank the UK EPSRC (Grant: GR/K25267) and 
the European Commission Telematics Programme 
(ECRAN and AVENTINUS projects) for funding 
which has made the development of VIE/LaSIE and 
GATE possible. 

References 
Allen, J. 1987. Natural Language Understanding. 
Benjamin/Cummings, Menlo Park, CA, 1st edi- 
tion. 
Appelt, D., J. Hobbs, J. Bear, D. Israel, 
M. Kameyama, A. Kehler, D. Martin, K. Myers, 
and M. Tyson. 1995. SRI International FAS- 
TUS system: MUC-6 Test Results and Analysis. 
In Proceedings of the Sixth Message Understand- 
ing Conference (MUC-6), pages 237-248. Morgan 
Kaufmann. 
Botley, S. and T. McEnery, editors. 1997. Dis- 
course Anaphora and Anaphor Resolution. Uni- 
versity College London Press, London. In press. 
Cowie, J. and W. Lehnert. 1996. Information Ex- 
traction. Communications o/the ACM, 39(1):80- 
91. 
Crowe, J. 1996. Constraint Based Event Recog- 
nition for Information Extraction. Ph.D. thesis, 
University of Edinburgh. 
DARPA. 1995. Proceedings o/ the Sixth Mes- 
sage Understanding Conference (MUC-6). Mor- 
gan Kaufmann. 
Gaizauskas, R. 1995. XI: A Knowledge Representa- 
tion Language Based on Cross-Classification and 
Inheritance. Technical Report CS-95-24, Depart- 
ment of Computer Science, University of Sheffield. 
Gaizauskas, R. and K. Humphreys. 1997a. Quant- 
ative Evaluation of Coreference Algorithms in an 
Information Extraction System. In S. Botley and 
T. McEnery, editors, Discourse Anaphora and 
Anaphor Resolution. University College London 
Press. In press. 
Gaizauskas, R. and K. Humphreys. 1997b. Us- 
ing a Semantic Network for Information Extrac- 
tion. Journal of Natural Language Engineering. 
In press. 
Gaizauskas, R., T. Wakao, K Humphreys, H. Cun- 
ningham, and Y. Wilks. 1995. Description of the 
LaSIE System as Used for MUC-6. In Proceedings 
of the Sixth Message Understanding Conference 
(MUC-6). Morgan Kaufmann. 
Grishman, R. and B. Sundheim. 1996. Message 
Understanding Conference - 6: A Brief History. In 
Proceedings of the 16th International Conference 
on Computational Linguistics, Copenhagen, June. 
Kamp, H. 1981. A Theory of Truth and Semantic 
Representation. In Formal Methods in The Study 
o/ Language. J. Groenendijk, Jannsen, T, and 
Stokhof, M. 
Sowa, J.F. 1984. Conceptual Structures : Inform- 
ation Processing in Mind and Machine. Reading 
(MA): Addison-Wesley. 
Whittemore, S and M. Macpherson. 1991. Event- 
building through Role-Filling and Anaphora Res- 
olution. In Proceedings of the 29th meeting of 
the Asssociation for Computational Linguistics, 
Berkeley, CA. 
Zarri, G.P. 1992. Semantic Modeling of the Content 
of (Normative) Natural Language Documents. In 
Proceedings of the specialized conference on Nat- 
ural Language Processing and its applications, 
Avignon. 
