Proceedings of the Workshop on Annotating and Reasoning about Time and Events, pages 30–37,
Sydney, July 2006. c©2006 Association for Computational Linguistics
Evaluating Knowledge-based Approaches to the Multilingual Extension of
a Temporal Expression Normalizer
Matteo Negri
ITC-irst
Povo - Trento, Italy
negri@itc.it
Estela Saquete, Patricio Mart´ınez-Barco, Rafael Mu˜noz
DLSI, University of Alicante
Alicante, Spain
{stela,patricio,rafael}@dlsi.ua.es
Abstract
The extension to new languages is a well
known bottleneck for rule-based systems.
Considerable human effort, which typi-
cally consists in re-writing from scratch
huge amounts of rules, is in fact required
to transfer the knowledge available to the
system from one language to a new one.
Provided sufficient annotated data, ma-
chine learning algorithms allow to mini-
mize the costs of such knowledge trans-
fer but, up to date, proved to be ineffec-
tive for some specific tasks. Among these,
the recognition and normalization of tem-
poral expressions still remains out of their
reach. Focusing on this task, and still ad-
hering to the rule-based framework, this
paper presents a bunch of experiments on
the automatic porting to Italian of a system
originally developed for Spanish. Differ-
ent automatic rule translation strategies are
evaluated and discussed, providing a com-
prehensive overview of the challenge.
1 Introduction
In recent years, inspired by the success of MUC
evaluations, a growing number of initiatives (e.g.
TREC1, CLEF2, CoNLL3, Senseval4) have been
developed to boost research towards the automatic
understanding of textual data. Since 1999, the Au-
tomatic Content Extraction (ACE) program5 has
been contributing to broaden the varied scenario
of evaluation campaigns by proposing three main
1http://trec.nist.gov
2http://clef-campaign.org
3http://www.cnts.ua.ac.be/conll
4http://www.senseval.org
5http://www.nist.gov/speech/tests/ace
tasks, namely the recognition of entities, rela-
tions, and events. In 2004, the Timex2 Detec-
tion and Recognition task6 (also known as TERN,
for Time Expression Recognition and Normaliza-
tion) has been added to the ACE program, making
the whole evaluation exercise more complete. The
main goal of the task was to foster research on sys-
tems capable of automatically detecting temporal
expressions (TEs) present in an English text, and
normalizing them with respect to a specifically de-
fined annotation standard.
Within the above mentioned evaluation exer-
cises, the research activity on monolingual tasks
has gradually been complemented by a consid-
erable interest towards multilingual and cross-
language capabilities of NLP systems. This trend
confirms how portability across languages has
now become one of the key challenges for Natu-
ral Language Processing research, in the effort of
breaking the language barrier hampering systems’
application in many real use scenarios. In this di-
rection, machine learning techniques have become
the standard approach in many NLP areas. This
is motivated by several reasons, including i) the
fact that considerable amounts of annotated data,
indispensable to train ML-based algorithms, are
now available for many tasks, and ii) the difficulty,
inherent to rule-based approaches, of porting lan-
guage models from one language to new ones. In
fact, while supervised ML algorithms can be eas-
ily extended to new languages given an annotated
training corpus, rule-based approaches require to
redefine the set of rules, adapting them to each new
language. This is a time consuming and costly
work, as it usually consists in manually rewriting
from scratch huge amounts of rules.
6http://timex2.mitre.org
30
In spite of their effectiveness for some tasks,
ML techniques still fall short from providing ef-
fective solutions for others. This is confirmed by
the outcomes of the TERN 2004 evaluation, which
provide a clear picture of the situation. In spite
of the good results obtained in the TE recognition
task (Hacioglu et al., 2005), the normalization by
means of ML techniques has not been tackled yet,
and still remains an unresolved problem.
Considering the inadequacy of ML techniques
to deal with the normalization problem, and fo-
cusing on portability across languages, this pa-
per extends and completes the previous work pre-
sented in (Saquete et al., 2006b) and (Saquete et
al., 2006a). More specifically, we address the fol-
lowing crucial issue: how to minimize the costs
of building a rule-based TE recognition system
for a new language, given an already existing sys-
tem for another language. Our goal is to experi-
ment with different automatic porting procedures
to build temporal models for new languages, start-
ing from previously defined ones. Still adhering
to the rule-based paradigm, we analyse different
porting methodologies that automatically learn the
TE recognition model used by the system in one
language, adjusting the set of normalization rules
for the new target language.
In order to provide a clear and comprehen-
sive overview of the challenge, an incremental ap-
proach is proposed. Starting from the architecture
of an existing system developed for Spanish (Sa-
quete et al., 2005), we present a bunch of exper-
iments which take advantage of different knowl-
edge sources to build an homologous system for
Italian. Building on top of each other, such exper-
iments aim at incrementally analyzing the contri-
bution of additional information to attack the TE
normalization task. More specifically, the follow-
ing information will be considered:
• The output of online translators;
• The information mined from a manually an-
notated corpus;
• A combination of the two.
2 The task: TE recognition and
normalization
The TERN task consists in automatically detect-
ing, bracketing, and normalizing all the time ex-
pressions mentioned within an English text. The
recognized TEs are then annotated according to
the TIMEX2 annotation standard described in
(Ferro et al., 2005). Markable TEs include both
absolute (or explicit) expressions (e.g. “April 15,
2006”), and relative (or anaphoric) expressions
(e.g. “three years ago”). Also markable are du-
rations (e.g. “two weeks”), event-anchored ex-
pressions (e.g. “two days before departure”), and
sets of times (e.g. “every week”). Detection and
bracketing concern systems’ capability to recog-
nize TEs within an input text, and correctly deter-
mine their extension. Normalization concerns the
ability of the system to correctly assign, for each
detected TE, the correct values to the TIMEX2
normalization attributes. The meaning of these at-
tributes can be summarized as follows:
• VAL: contains the normalized value of a TE
(e.g. “2004-05-06” for “May 6th, 2004”)
• ANCHOR VAL: contains a normalized form
of an anchoring date-time.
• ANCHOR DIR: captures the relative
direction-orientation between VAL and
ANCHOR VAL.
• MOD: captures temporal modifiers (pos-
sible values include: “approximately”,
“more than”, “less than”)
• SET: identifies expressions denoting sets of
times (e.g. “every year”).
2.1 The evaluation benchmark
Moving to a new language, an evaluation bench-
mark is necessary to test systems performances.
For this purpose, the temporal annotations of the
Italian Content Annotation Bank (I-CAB-temp7)
have been selected.
I-CAB consists of 525 news documents
taken from the Italian newspaper L’Adige
(http://www.adige.it), and contains around
182,500 words. Its 3,830 temporal expressions
(2,393 in the training part of the corpus, and 1,437
in the test part) have been manually annotated
following the TIMEX2 standard with some adap-
tations to the specific morpho-syntactic features
of Italian, which has a far richer morphology than
English (see (Magnini et al., 2006) for further
details).
7I-CAB is being developed as part of the three-year
project ONTOTEXT funded by the Provincia Autonoma di
Trento, Italy. See http://tcc.itc.it/projects/ontotext
31
3 The starting point: TERSEO
As a starting point for our experiments we used
TERSEO, a system originally developed for the
automatic annotation of TEs appearing in a Span-
ish written text in compliance with the TIMEX2
standard (see (Saquete, 2005) for a thorough de-
scription of TERSEO’s main features and func-
tionalities).
TEXT
POS 
TAGGER
RECOGNITION: 
PARSER
Lexical and
morphological
information
Temporal 
expression
recognition
DATE
ESTIMATION
Dictionary
Temporal
Expression
Grammar
TEMPORAL
EXPRESSION
NORMALIZATION
EVENT 
ORDERING
ORDERED
TEXT
Documental 
DataBase
Figure 1: System’s architecture.
Basically (see Figure 1), the TE recognition
and normalization process is carried out in two
phases. The first phase (recognition) includes a
pre-processing of the input text, which is tagged
with lexical and morphological information that
will be used as input to a temporal parser. The
temporal parser is implemented using an as-
cending technique (chart parser) and relies on a
language-specific temporal grammar. As TEs can
be divided into absolute and relative ones, such
grammar is tuned for discriminating between the
two groups. On the one hand, absolute TEs di-
rectly provide and fully describe a date. On the
other hand, relative TEs require some degree of
reasoning (as in the case of anaphora resolution).
In the second phase of the process, in order to
translate these expressions into their normalized
form, the lexical context in which they occur is
considered. At this stage, a normalization unit
is in charge of determining the appropriate refer-
ence date (anchor) associated to each anaphoric
TE, calculating its value, and finally generating the
corresponding TIMEX2 tag.
¿From a multilingual perspective, an impor-
tant feature of TERSEO is the distinction between
recognition rules, which are language-specific,
and normalization rules, which are language-
independent and potentially reusable for any other
language. Taking the most from the modular ar-
chitecture of the system, a first multilingual exten-
sion has been evaluated over the English TERN
2004 test set. In that extension, the English
temporal model was automatically obtained from
the Spanish one, through the automatic transla-
tion into English8 of the Spanish TEs recognized
by the system (Saquete et al., 2004). The re-
sulting English TEs were then mapped onto the
corresponding language-independent normaliza-
tion rules, with good results (compared with other
participants to the competition) both in terms of
precision and recall. These results are shown in
Table 1.
Prec Rec F
timex2 0.673 0.728 0.699
anchor dir 0.658 0.877 0.752
anchor val 0.684 0.912 0.782
set 0.800 0.667 0.727
text 0.770 0.620 0.690
val 0.757 0.735 0.746
Table 1: Evaluation of English-TERSEO over the
TERN 2004 test set
The positive results of this experience demon-
strated the viability of the adopted solutions, and
motivate our further investigation with Italian as a
new target language.
4 Porting TERSEO to Italian
Due to the separation between language-specific
recognition rules and language-independent nor-
malization rules, the bulk of the porting process
relies on the adaptation of the recognition rules
to the new target language. Taking advantage of
different knowledge sources (either alone or in
combination), an incremental approach has been
adopted, in order to determine the contribution of
additional information on the performance of the
resulting system for Italian.
8Altavista Babel Fish Translation has been used for this
purpose (http://world.altavista.com).
32
4.1 Using online translators
As a first experiment, the same procedure adopted
for the extension to English has been followed.
This represents the simplest approach for porting
TERSEO to other languages, and will be consid-
ered as a baseline for comparison with the results
achieved in further experiments. The only minor
difference with respect to the original procedure
is that now, since two aligned sets of recognition
rules (i.e. for Spanish and for English) are avail-
able, both models have been used. The reason for
considering both models is the fact that they com-
plement each other: on the one hand, the Span-
ish model was obtained manually and showed high
precision values in detection (88%); on the other
hand, although the English model showed lower
precision results in detection (77%), the on-line
translators from English to Italian perform better
than translators from Spanish to Italian.
The process is carried out in the following four
steps.
1. Eng-Ita translation. All the English TEs
known by the system are translated into Ital-
ian9. Starting English, the probability of ob-
taining higher quality translations is maxi-
mized.
2. Spa-Ita translation. For each English TE
without an Italian translation, the correspond-
ing Spanish expression is translated into Ital-
ian. Also the Spanish TEs that do not have an
English equivalent are translated from Span-
ish10 into Italian. This way, the coverage
of the resulting model is maximized, becom-
ing comparable to the hand-crafted Spanish
model.
3. TE Filtering. A filtering module is used to
guarantee the correctness of the translations.
For this purpose, the translated expressions
are searched in the Web with Google. If an
expression is not found by Google it is given
up; otherwise it is considered as a valid Ital-
ian TE. The inconvenience of adopting this
simple filtering strategy occurs in case of am-
biguous expressions, i.e. when a correct ex-
pression is obtained through translation, and
9Also for English to Italian translation, Altavista Babel
Fish Translation has been used
10Using the Spanish-Italian translator available at
http://www.tranexp.com:2000/Translate/result.shtml
Google returns at least on document contain-
ing it, but the expression is not a tempo-
ral one. In these cases the system will er-
roneously store in its database non-temporal
expressions. In this experiment the results
returned by Google have not been analyzed
(only the number of hits has been taken into
account), nor the impact of these errors has
been estimated. A more precise analysis of
the output of the web search has been left as
a future improvement direction.
4. Normalization rules assignment. Finally,
the resulting Italian translations are mapped
onto the language-independent normalization
rules associated with the original English and
Spanish TEs.
The development of this first automatic porting
procedure required one person/week for software
implementation, and less than an hour to obtain
the new model for Italian. The performance of the
resulting system, evaluated over the test set of I-
CAB, is shown in table 2.
Prec Rec F
timex2 0.725 0.833 0.775
anchor dir 0.211 0.593 0.311
anchor val 0.203 0.571 0.300
set 0.152 1.000 0.263
text 0.217 0.249 0.232
val 0.364 0.351 0.357
Table 2: Porting to Italian based on translations
The results achieved by the translation-based
approach are controversial. On the one hand, we
observe a detection performance in line with the
English version of the system. The timex2 at-
tribute, which indicates the proportion of detected
TEs11, has even higher scores, both in terms of
precision (+5%) and recall (+11%), with respect
to the English system. On the other hand, both
bracketing (see the text attribute, which indicates
the quality of extent recognition) and normaliza-
tion (described by the other attributes) show a per-
formance drop. Unfortunately, the reasons of this
drop are still unclear. One possible explanation
is that, due to the intrinsic difficulties presented
by the Italian language, the translation-based ap-
proach falls short from providing an adequate cov-
erage of the many possible TE variants. While
11At least one overlapping character in the extent of the
reference and the system output is required for tag aligment.
33
the presence of lexical triggers denoting a TE ap-
pearing in a text (e.g. the Italian translations of
“years”, “Monday”, “afternoon”, “yesterday”) can
be easily captured by this approach, the complex-
ity of many language-specific constructs is out of
its reach.
4.2 Using an annotated corpus
In a second experiment, the annotations of the
training portion of I-CAB have been used as a pri-
mary knowledge source. The main purpose of this
approach is to maximize the coverage of the Ital-
ian TEs, starting from language-specific knowl-
edge mined from the corpus. The basic hypothe-
sis is that a bottom-up porting methodology, led by
knowledge in the target language, is more effective
than the top-down approach based on knowledge
derived from models built for other languages.
The former, in fact, is in principle more suitable to
capture language-specific TE variations. In order
to test the validity of ths hypothesis, the following
two-step process has been set up:
1. TE Collection and translation. The Italian ex-
pressions are collected from the I-CAB train-
ing portion, and translated both into Spanish
and English.
2. Normalization rules assignment. Italian TEs
are assigned to the appropriate normalization
rules. For each Italian TE mined from the
corpus, the selection is done considering the
normalization rules assigned to its transla-
tions. If both the Spanish and English ex-
pressions are found in their respective mod-
els, and are associated with the same normal-
ization rule, then this rule is assigned also to
the Italian expression. Also, when only one
of the translated expressions is found in the
existing models, the normalization rule is as-
signed. In case of discrepancies, i.e. if both
expressions are found, but are not associated
to the same normalization rule, then one of
the languages must be prioritized. Since the
manually obtained Spanish model has shown
a higher precision, Spanish rules are pre-
ferred.
As the corpus-based approach is mostly built on
the same software used for the translation-based
porting procedure, it did not require additional
time for implementation. Also in this case, the
new model for Italian has been obtained in less
than one hour. Performance results calculated over
the I-CAB test set are reported in Table 3.
Prec Rec F
timex2 0.730 0.839 0.781
anchor dir 0.412 0.414 0.413
anchor val 0.339 0.340 0.339
set 0.030 1.000 0.059
text 0.222 0.255 0.238
val 0.285 0.274 0.279
Table 3: Porting based on corpus annotations
These results partially confirm our working hy-
pothesis, showing a performance increase in terms
of the Italian TEs correctly recognized by the sys-
tem. In fact, both the timex2 attribute, which
indicates the coverage of the system (detection),
and the text attribute, which refers to the TEs
extent determination (bracketing), are slightly in-
creased. This may lead to the conclusion that auto-
matic porting procedures can actually benefit from
language-specific knowledge derived from a cor-
pus.
However, looking at the other TIMEX2 at-
tributes, the situation is not so clear due to the less
coherent behaviour of the system on normaliza-
tion. While for two attributes (anchor dir and an-
chor val) the system performs better, for the other
two (set and val) a performance drop is observed.
A possible reason for that could be related to the
limited number of TE examples that can be ex-
tracted from the Italian corpus (whose dimensions
are relatively small compared to the annotated cor-
pora available for English). In fact, compared to
the sum of English and Spanish examples used for
the translation-based porting procedure, the Ital-
ian expressions present in the corpus are fewer and
repetitive. For instance, with 131, 140, and 30 oc-
currences, the expressions “oggi” (“today”), “ieri”
(“yesterday”), and “domani” (“tomorrow”) repre-
sent around 12.5% of the 2,393 Italian TEs con-
tained in the I-CAB training set.
4.3 Combining online translators and an
annotated corpus
In light of the previous considerations, a third ex-
periment has been conducted combining the top-
down approach proposed in Section 4.1 and the
bottom-up approach proposed in Section 4.2. The
underlying hypothesis is that the induction of an
effective temporal model for Italian can bene-
fit from the combination of the large amount of
examples coming from translations on the one
34
side, and from the more precise language-specific
knowledge derived from the corpus on the other.
To check the validity of this hypothesis, the pro-
cess described in Section 4.2 has been modified
adding an additional phase. In this phase, the set
of TEs derived from I-CAB is augmented with the
expressions already available in the Spanish and
English TE sets. The new porting process is car-
ried out in the following steps:
1. TE Collection and translation. The Italian ex-
pressions are collected from the I-CAB train-
ing portion, and translated both into Spanish
and English.
2. Normalization rules assignment. With the
same methodology described in Section 4.2
(step 2), the Italian TEs mined from the cor-
pus are mapped onto the appropriate normal-
ization rules assigned to their translations.
3. TE set augmentation. The set of Italian TEs
is automatically augmented with new expres-
sions derived from the Spanish and English
TE sets. As described in Section 4.1, these
expressions are first translated into Italian us-
ing on-line translators, then filtered through
Web searches. The remaining TEs are in-
cluded in the Italian model, and related to the
same normalization rules assigned to the cor-
responding Spanish or English TEs.
Also this porting experiment was carried out
with minimal modifications of the existing code.
The automatic acquisition of the new model for
Italian required around one hour. Evaluation re-
sults, calculated over the I-CAB test set are pre-
sented in Table 4.
Prec Rec F
timex2 0.726 0.834 0.776
anchor dir 0.578 0.475 0.521
anchor val 0.516 0.424 0.465
set 0.182 1.000 0.308
text 0.258 0.296 0.276
val 0.564 0.545 0.555
Table 4: Porting based on corpus annotations and
online translators
As can be seen from the table, the combina-
tion of the two approaches leads to an overall per-
formance improvement with respect to the previ-
ous experiments. Apart from a slight decrease in
terms of detection (timex2 attribute), both brack-
eting and normalization performance benefit from
such combination. The improvement on bracket-
ing (text attribute) is around 4% with respect to
both the previous experiments. On average, the
improvement for the normalization attributes is
around 15% with respect to the translation-based
method (ranging from +4,5% for the set attribute,
to +20% for the val attribute), and 20% with re-
spect to the corpus-based method (ranging from
+11% for the anchor dir attribute, to +30% for
the set attribute). These performance improve-
ments are summarized in Table 5, which reports
the F-Measure scores achieved by the three port-
ing approaches.
F-Tran. F-Corpus F-Comb.
timex2 0.775 0.781 0.776
anchor dir 0.311 0.413 0.521
anchor val 0.300 0.339 0.465
set 0.152 0.059 0.308
text 0.263 0.238 0.276
val 0.232 0.279 0.555
Table 5: F-Measure scores comparison
These results confirm the validity of our work-
ing hypothesis, showing that:
• taken in isolation, both the knowledge de-
rived from models built for other languages,
and the language-specific knowledge derived
from an annotated corpus, have a limited im-
pact on the system’s performance;
• taken in combination, the top-down and the
bottom-up approaches can complement each
other, allowing to cope with the complexity
of the porting task.
5 Comparing TERSEO with a
language-specific system
For the sake of completeness, the results achieved
by our combined porting procedure have been
compared with those achieved, over the I-CAB
test set, by a system specifically designed for
Italian. The ITA-Chronos system (Negri and
Marseglia, 2004), a multilingual system for the
recognition and normalization of TEs in Italian
and English, has been used for this purpose. Up to
date, being among the two top performing systems
at TERN 2004, Chronos represents the state-of-
the-art with respect to the TERN task. In addition,
to the best of our knowledge, this is the only sys-
tem effectively dealing with the Italian language.
35
Like all the other state-of-the-art systems ad-
dressing the recognition/normalization task, ITA-
Chronos is a rule-based system. From a design
point of view, it shares with TERSEO a rather
similar architecture which relies on different sets
of rules. These are regular expressions that check
for specific features of the input text, such as the
presence of particular word senses, lemmas, parts
of speech, symbols, or strings satisfying specific
predicates12. Each set of rules is in charge of
dealing with different aspects of the problem. In
particular, a set of around 350 rules is designed
for TE recognition and is capable of recognizing
with high Precision/Recall rates a broad variety of
TEs. Other sets of regular expressions, for a total
of around 700 rules, are used in the normalization
phase, and are in charge of handling each specific
TIMEX2 normalization attribute. The results ob-
tained by the Italian version of Chronos over the
I-CAB test set are shown in Table 6.
Prec Rec F F-Comb
timex2 0.925 0.908 0.917 0.776 (-14%)
anchor dir 0.733 0.636 0.681 0.521 (-16%)
anchor val 0.495 0.462 0.478 0.465 (-1.3%)
set 0.616 0.500 0.552 0.308 (-24%)
text 0.859 0.843 0.851 0.276 (-57%)
val 0.636 0.673 0.654 0.555 (-10%)
Table 6: Evaluation of ITA-Chronos over the I-
CAB test set
As expected, the distance between the results
obtained by ITA-Chronos and the best Italian sys-
tem automatically obtained from TERSEO (F-
Comb) is considerable. On average, in terms of
F-Measure, the scores obtained by ITA-TERSEO
are 20% lower, ranging from -1.3% for the an-
chor val attribute, to -57% for the text attribute.
However, going beyond the raw numbers, a com-
prehensive evaluation must also take into account
the great difference, in terms of the required time,
effort, and resources deployed in the development
of the two systems. While the implementation of
the manual one took several months, the automatic
porting procedure of TERSEO to Italian (in all the
three modalities described in this paper) is a very
fast process that can be accomplished in less than
an hour. Considering the trade-off between per-
formance and effort required for system’s devel-
12For instance, the predicates “Weekday-p” and
“Time Unit-p” are respectively satisfied by strings such
as “Monday”, “Tuesday”, ..., “Sunday”, and “second”,
“minute”, “hour”, “day”, ..., “century”. Of course, this also
holds for the Italian equivalents of these expressions
opment, the proposed methodology represents a
viable solution to attack the porting problem.
6 Conclusions
In this paper, the problem of automatically extend-
ing to new languages a rule-based system for TE
recognition and normalization has been addressed.
Adopting an incremental approach, different port-
ing strategies, for the creation of an Italian system
starting from an already available Spanish system,
have been evaluated and discussed. Each exper-
iment has been carried out considering the con-
tribution of different knowledge sources for rules
translation. Firstly, the contribution given by the
output of online translators has been evaluated,
showing detection performances in line with a pre-
viously developed English extension of the sys-
tem, but a performance drop in terms of normal-
ization performance. Then, the contribution of
knowledge mined from an annotated corpus has
been considered. Results show a performance in-
crease in terms of detection and bracketing, but
a less coherent behaviour in terms of normaliza-
tion. Finally, a combined approach has been ex-
perimented, resulting in an overall performance
increase. System’s performance is still far from
the results obtained by a state-of-the-art system for
Italian but, considering the trade-off between per-
formance and effort required for system’s devel-
opment, results are encouraging.

References
L. Ferro, L. Gerber, I. Mani, B. Sundheim, and G. Wil-
son. 2005. Tides.2005 standard for the annotation
of temporal expressions. Technical report, MITRE.
K. Hacioglu, Y. Chen, and B. Douglas. 2005. Time
Expression Labeling for English and Chinese Text.
In Proceedings of CICLing 2005, pages 548–559.
B. Magnini, E. Pianta, C. Girardi, M. Negri, L. Ro-
mano, M. Speranza, and R. Sprugnoli. 2006. I-
CAB: the Italian Content Annotation Bank. In Pro-
ceedings of LREC 2006. To appear.
M. Negri and L. Marseglia. 2004. Recognition and
normalization of time expressions: Itc-irst at tern
2004. Technical report, ITC-irst, Trento.
E. Saquete, P. Martnez-Barco, and R. Muoz. 2004.
Evaluation of the automatic multilinguality for time
expression resolution. In DEXA Workshops, pages
25–30. IEEE Computer Society.
E. Saquete, R. Muoz, and P. Martnez-Barco. 2005.
Event ordering using TERSEO system. Data and
Knowledge Engineering Journal, page (To be pub-
lished).
E. Saquete, P. Martinez-Barco, R. Munoz R., M. Ne-
gri, M. Speranza, and R. Sprugnoli R. 2006a. Au-
tomatic resolution rule assignment to multilingual
temporal expressions using annotated corpora. In
Proceedings of the TIME 2006 International Sym-
posium on Temporal Representation and Reasoning.
To Appear.
E. Saquete, P. Martinez-Barco, R. Munoz R., M. Negri,
M. Speranza, and R. Sprugnoli R. 2006b. Multilin-
gual Extension of a Temporal Expression Normal-
izer using Annotated Corpora. In Proceedings of the
EACL Workshop on Cross-Language Knowledge In-
duction.
E. Saquete. 2005. Temporal information Resolution
and its application to Temporal Question Answer-
ing. Phd, Departamento de Lenguages y Sistemas
Inform´aticos. Universidad de Alicante, June.
