There are four basic annotation types: events, times, temporal relations and event identity. We briefly introduce these here.
Events Intuitively, an event is something that happens, something that one can imagine placing on a time line. Events can be viewed as conceptually instantaneous or as happening over a period of time and events can stand in coreference relationships to each other.
Times Events happen at or during certain times. Times can be viewed as either points or intervals.
Temporal Relations Events and events as well as events and times stand in certain temporal relations to each other, for example before and after.
Event Identity In newspaper articles, events are often introduced and come back to later to give more information about them.
Anything else, we do not want to annotate. For example states (non-essential relations between entities or holding of a non-essential attribute like The plane can carry four people) and details of events (like logical subject). The ultimate test of whether something is to be annotated as an event is whether it is anchorable in time and whether one would like to place it on a time map.
NOTE: Headlines are not to be annotated
at the moment either !
NOTE: Events, times and the signals (see below) indicating the
temporal relation between them have unique IDs
&nbs
p;
throughout the text. To make the annotation easier, I created an ID counter
that just counts from one up.
&nbs
p;
This counter does not ditinguish between IDs for events or times or signals,
but at least if you stick with
&nbs
p;
it you will assign unique IDs to everything. If an annotation is deleted,
the counter does not take this into
&nbs
p;
account either, so there will be gaps (which does not matter for the
annoation).
If an already annotated
&nbs
p;
text is loaded into the tool, then the highest ID (plus one) is ID that
is displayed (so here too might be
&nbs
p;
gaps in the numbers, which is fine). You don't have to use this ID counter
! Some annotators said they
&nbs
p;
would find it helpful, so I built it in.
Events must be anchorable in time, even though we might not be able
to accurately do so with the information in the article. To get a better
idea what events are, we can compare them with states like
States are either not anchorable in time, like the first example,
or are true over time spans longer than that covered by the article (second
and third example).
We distinguish different classes of events, the reason being that those classes influence the `temporal interpretation' of the associated events. This is best illustrated with the following example.
The plane hitting the water is what we call occurrence event, one
of the event classes we are interested in placing on a time map. The fisherman
seeing the crash is also an occurrence event,
but is related to the crash event in a particular way. The fisherman saw
the crash, which means that both events happened at (roughly) the same
time. To be able to exploit this special relationship (it helps to locate
the crash event in time), we distinguish perception
events (the fisherman observing) from occurrence
events (the plane crashing). The third class of events is called
reporting
events (said Petty Officer Jeff Fenn), which serve a particular
purpose in the genre of newspaper texts, that of giving the source of the
information.
The final class of events we distinguish is aspectual events, such as
Events can stand in certain coreference relation to each other and the
one we want to annotate is event identity.
It is impossible to identify the full range of ways in which this can be
expressed in text, although coreference relation between participants,
animate or inanimate, are good indicators.
The following (simplified) examples show the basics. They are simplified, because they do not show attributes, we will come back to that later.
eid The event ID uniquely
identifies the event in the text.
&nbs
p;
NOTE: Please make sure that you are using UNIQUE
identifiers for each event mentioned in the text, ie.
&nbs
p;
each time you annotate something as an event. Identical events mentioned
more than once in the
&nbs
p;
text get a different identifier for each time they are mentioned. And event
identity is expressed by
&nbs
p;
using the attributes "relatedToEvent" and "eventRelType".
&nbs
p;
For example, if we find the following in the text, then both times the
crash is mentioned a
&nbs
p;
DIFFERENT ID is given:
&nbs
p;
The plane crashed, <event eid=2> killing
</event>
all passengers on board. After the <event
eid=3>
&nbs
p;
crash </event>
search
teams ...
potential
values: positive integers
optional:
no
example:
The
plane <event eid=3> crashed </event>
class An event belongs to one of these classes.
potential
values: OCCURRENCE, PERCEPTION, REPORTING, ASPECTUAL
optional:
no
See next attribute
for example.
argid Reporting, perception and aspectual events
usually have another event or other events as an
&nbs
p;
attribute and the argument ID identifies these. Multiple event IDs separated
by a comma can
be filled in.
potential
values: positive integers
optional:
yes
example:
The
MoD <event eid=5 class=reporting argid=7>
announced </event> that the jet
&nbs
p;
fighter <event eid=7 class=occurrence>
ploughed
</event> into the mountain.
&nbs
p;
The <event eid=8 class=occurrence> search
</event>
&nbs
p;
<event eid=9 class=aspectual type=start argid=8
relatedToTime=6
&nbs
p;
&n
bsp;
timeRelType=is_included>
&nbs
p;
began</event>
<signal sid=3> on
</signal> <timex tid=6> Thursday
</timex>
tense The tense of the indicating verb plays a crucial
role in determining the time the event took place
&nbs
p;
and the simple tense of it is recorded in this attribute. This does not
hold for nominalisations
&nbs
p;
and non-finite clauses, since these are tenseless by definition.
The annotator can fill in the
&nbs
p;
tense if it is clear whether the event conveyed happened in the past, is
happening or will happen
&nbs
p;
in the future, but this is not necessary. For events conveyed by
finite clauses, however, the tense
&nbs
p;
SHOULD BE FILLED IN.
potential
values: PAST, PRESENT, FUTURE
optional:
yes
example:
The
plane <event eid=3 tense=past> crashed </event>.
&nbs
p;
Mr Smith will <event eid=11 tense=future>
make
</event> an announcement
&nbs
p;
tomorrow afternoon.
aspect It is sometimes helpful to know whether the
aspect of the verb is progressive or perfective,
&nbs
p;
which can be indicated using the attribute aspect.
potential
values: PROGRESSIVE, PERFECTIVE
optional:
yes
example:
...several
vessels and a helicopter were <event eid=13
tense=past
&nbs
p;
class=PROGRESSIVE> combing </event> the
area.
relatedToEvent The ID of the event the current event is
related to is stored here.
potential
values: positive integers
optional:
yes
See next
attribute for example.
eventRelType The type of temporal relation those
two events are related by is stored in this attribute.
potential
values: BEFORE, AFTER, INCLUDES, IS_INCLUDED, SIMULTANEOUS
optional:
yes
example:
All
75 people on board the Aeroflot Airbus <event
eid=4 relatedToEvent=5
 
;
eventRelType=simultaneous signalID=7> died </event>
&nbs
p;
<signal sid=7> when </signal>
it <event eid=5> ploughed </event>
into
a Siberian
&nbs
p;
mountain.
relatedToTime The ID of the time-object the current
event is related to is stored here.
potential
values: positive integers
optional:
yes
See next attribute
for example.
timeRelType The type of temporal relation the event and
time-object are related by is stored in this
&nbs
p;
attribute.
potential
values: BEFORE, AFTER, INCLUDES, IS_INCLUDED, SIMULTANEOUS
&nbs
p;
&n
bsp;
(see section 4
for a description of these relations)
optional:
yes
example:
A
small single-engine plane <event eid=9 relatedToTime=5
timeRelType=is_included
&nbs
p;
signalID=9> crashed </event> into
the Atlantic Ocean
<signal sid=9 > on
</signal>
&nbs
p;
<timex tid=5> Wednesday
</timex>.
signalID The ID of the text span that signals the temporal
relation holding between two entities can be
&nbs
p;
kept in this attribute.
potential
values: positive integers
optional:
yes
See relatedToEvent
and relatedToTime for example.
Note about event class:
Causal and subevent relationships would come to mind as relationships
one might want to annotate. But they are very difficult to define and to
distinguish, which is why they are not included as an event class. If the
annotator encounters such a relation then this can only be annotated using
the attribute eventRelType and choosing the
appropriate temporal relation (usually cause precedes effect and subevents
are temporally included in their `container' events). Examples for
what can be feasibly interpreted as a causality or subevent relation are
the following:
Following general convention, and the approach taken in MUC, we distinguish between two classes of time objects, dates and times, units which are larger or smaller than a day, respectively. In addition, we flag complex time objects, as in the example before, as complex.
Like events, times have a unique ID so they can be uniquely identified in the text. All time expressions, simple and complex, have the following attributes:
tid The ID uniquely identifies the time
object in the text.
potential
values: positive integers
optional:
no
type The type of the time object.
potential
values: DATE, TIME, COMPLEX
optional:
no
calDate The calendrical date represented by the time
object.
potential
values: [[[[[HH]MM]DDMM]YYYY or (`SPR'|`SUM'|`AUT'|`WIN')YYYY (24 hour
clock)
optional:
no
Examples of time expressions are:
eid The ID of the event the time interval is
related to.
potential
values: positive integers
optional:
yes
signalID The ID of the signal representing
the type of temporal relation.
potential
values: positive integers
optional:
yes
relType The temporal relation holding between
the time interval and the event.
potential
values: BEFORE, AFTER, INCLUDES, IS_INCLUDED, SIMULTANEOUS
optional:
yes
An example of an annotated complex time expression is the following:
Miscellaneous Non-Time Expressions
Special days, such as holidays referred to by name All Saints' Day
are to be tagged.
Temporal relations are either expressed explicitly or implicitly. Explicitly expressed are those where, for example, temporal prepositional phrases, temporal adverbial phrases and temporal subordinate conjunctions are used. The first example shows how an event is related to another event (by using the subordinate conjunction when) and then how an event is related to a time (by using the temporal preposition in).
Events can stand in certain coreference relationships to each other and the relation we want to annotate is identity, as in the following example.
The full set of temporal relations we suppose at present is:
is_included The plane crashed on Wednesday.
  ; & nbsp; The event is completely included in the time or other event.
includes By midafternoon, several vessels were combing the area.
 
; &
nbsp;
This is the reverse of is_included, where we know that the
combing
event includes
&nbs
p;
midafternoon completely.
after The plane crashed after the pilot and his crew ejected.
 
; &
nbsp;
This relation, as well as the following before, is chosen when an
event is clearly after
&nbs
p;
(or before) another event.
before ...before the craft fell, its three rotor blades shot off.
simultaneous All 75 people on board the Aeroflot Airbus
died when it ploughed into a Siberian
&
nbsp; &nb
sp;
mountain in March 1994.
 
; &
nbsp;
Events can happen at exactly the same time (like identical events) or overlap
and we
&nbs
p;
know how they overlap exactly. But often we only know that two events happen
&nbs
p;
`roughly at the same time' without knowing exactly whether and how they
overlap.
&nbs
p;
In both cases (when we know exactly how they are overlapping or we only
roughly
&nbs
p;
know) we choose the relation simultaneous.
<
br>
If the word or text span signaling two events being related is realised, as in the example below, then not only is the signal annotated but also the ID of the signal is stored as an attribute of the event, so the link between the signaling word and the events is not lost. If an event is related to a time object then the event carries the ID of the time object and the type of relation in its attributes. Should the signaling word be realised, again as in the example, then the ID of the signal becomes an attribute of the event. The signal has only one atribute, the unique identifier sid.
The following examples illustrate the approach.
It is recommended to go through 5 phases when annotating.
NAVY BLAMES SHOWING OFF FOR JET CRASH
WASHINGTON The Navy
saide1Fridayt1
that a pilot was probably
showing offe2 for his parents
whens1
he crashede3 an F-14A jet fighter in
Nashville ins2 Januaryt2,
killinge4
himself, a fellow officer and three
people on the ground.
(e1) Event: eid=1 class=reporting tense=past argEvent=2
(s1) Time: tid=1 type=date calDate=29111996
(e2) Event: eid=2 class=occurrence tense=past aspect=progressive relatedToEvent=3
&nbs p; eventRelType=simultaneous
(s1) Signal: sid=2
(e3) Event: eid=3 class=occurrence tense=past relatedToTime=2 timeRelType=included
&nbs p; signalID=2
(s2) Signal: sid=2
(t2) Time: tid=2 type=date caLDate=011996
(e4) Event: eid=4 class=occurrence tense=past relatedToEvent=3
&nbs p; eventRelType=is_included
&nbs p; NOTE: I would have annotated this as a subevent, but since I can't do this I
&nbs p; i ncluded the temporal relation "is_included".
The pilot, Lt. Comdr. John Stacy Bates, had been
groundede5
for a
month ins3 April 1995t3
afters4
he loste6 control of another F-14A
afters5
taking offe7 from the aircraft carrier
USS Lincoln. The plane crashede8
afters6 the pilot and his crew
ejectede9.
The crew members were
rescuede10.
( e5) Event: eid=5 class=occurrence tense=past aspect=perfective relatedToTime=3Rear Adm. Bernard Smith, who investigatede11 the Nashville crashe12,
&nbs p; timeRelType=is_included signalID=3
( s3) Signal: sid=3
( t3) Time: tid=3 type=date calDate=041995
( s4) Signal: sid=4
( e6) Event: eid=6 class=occurrence tense=past relatedToEvent=5 eventRelTyp=before signalID=4
( s5) Signal: sid=5
( e7) Event: eid=7 class=occurrence tense=past relatedToEvent=6 eventRelType=before
( e8) Event: eid=8 class=occurrence tense=past
&nbs p; NOTE: This is not the same crash event as the one in the first paragraph!
( s6) Signal: sid=6
( e9) Event: eid=9 class=occurrence tense=past relatedToEvent=8 eventRelType=before
(e10) Event: eid=10 class=occurrence tense=past
(e11) Event: eid=11 class=occurrence tense=pastThe admiral saide13 that ins7 taking offe14 from the Nashville airport
(e12) Event: eid=12 class=occurrence tense=past relatedToEvent=8 eventRelType=identity
(e13) Event: eid=13 class=reporting tense=past argEvent=14,15,16Afters8 the near-perpendicular takeoffe17, in the clouds, Smith saide18,
( s7) Signal: sid=7
(e14) Event: eid=14 class=occurrence tense=past aspect=progressive
&nbs p; NOTE: This is not the same take off as event 7!
(e15) Event: eid=15 class=occurrence tense=past relatedToEvent=14
&nbs p; eventRelType=included signalID=7
(e16) Event: eid=16 class=occurrence tense=past aspect=progressive relatedToEvent=15
&nbs p; eventRelType=simultaneous
( s8) Signal: sid=8By then, the admiral said, it was too late to prevent the plane
(e17) Event: eid=17 class=occurrence tense=past relatedToEvent=14
&nbs p; eventRelType=identity
(e18) Event: eid=18 class=reporting tense=past argEvent:17,19,20,21,22
(e19) Event: eid=19 class=occurrence tense=past relatedToEvent=17 eventRelType=after
&nbs p; signalID=8
(e20) Event: eid=20 class=occurrence aspect=progessive tense=past relatedToEvent=17
&nbs p; eventRelType=after
(e21) Event: eid=21 class=occurrence tense=past aspect=progressive
( s9) Signal: sid=9
(e22) Event: eid=22 class=occurrence tense=past relatedToEvent=21 eventRelType=after
&nbs p; signalID=9
(e23) Event: eid=23 class=occurrence tense=past aspect=progessive relatedToEvent=3In addition to Bates, 33, and his radar intercept officer, Lt.
&nbs p; eventRelType=is_includedNOTE: "prevent" could be interpreted as an event, but is hypothetical and thus excluded. And because "prevent" is not annotated, the reporting event "said" is not annotated because reporting events are only marked up if their argument events are marked up.
(e24) Event: eid=24 class=occurrence tense=past relatedToEvent=3The crash would never have happened had the pilot made a normal
&nbs p; eventRelType=is_included
&nbs p; NOTE: I interpret the killing event as a subevent of the crash, so the
&nbs p; &n bsp; temporal relation is included
(e25) Event: eid=25class=occurrence tense=past aspect=progressive relatedToEvent=24
&nbs p; eventRelType=simultaneous
&nbs p; NOTE: This relation would be open for discussion. One could say that the
&nbs p; &n bsp; visiting event includes the crash event but I would argue that that
&n bsp; is not all that clear and thus I put "simultaneous".
NOTE: No events are annotated here for the same reason as above (events being hypothetical).
04-12-96
(doa) This is the date of the article. No attributes involved.