Friday

Practical Applications of Temporal and Event Reasoning: 

Practical Applications of Temporal and Event Reasoning James Pustejovsky, Brandeis Graham Katz, Osnabrück Rob Gaizauskas, Sheffield ESSLLI 2003 Vienna, Austria August 25-29, 2003

Course Outline: 

Course Outline Monday- Theoretical and Computational Motivations Overview of Annotation Task Events and Temporal Expressions Tuesday Anchoring Events to Times Relations between Events Wednesday Syntax of TimeML Tags Semantic Interpretations of TimeML Relating Annotations Temporal Closure Thursday Automatic Identification of Expressions Automatic Link Construction Friday- Outstanding Problems

Friday Topics: 

Friday Topics Events with Argument Binding TimeML German Fragment Outstanding Problems TimeML-enabled Applications

Features in TimeML 2.0: 

Features in TimeML 2.0 Argument binding into Events Pred feature in EVENT General types with like entailments Vendler classification: Scope of Negation and Modality: Represented on TLINK

Argument Binding into Events: 

Argument Binding into Events

Syntax of Entity: 

Syntax of Entity <ENTITY> attributes ::= aid type agreement det aid ::= ID {aid ::= argumentID argumentID ::= a<integer>} type ::= <named entity type> agreement ::= ??? det ::= ‘a’|’the’|’possessive’|’quant’

Syntax of Arglink: 

Syntax of Arglink <ARGLINK> attributes ::= [lid] [origin] eventInstanceID (relatedEventInstanceID | relatedArgumentID) preposition lid ::= ID {lid ::= LinkID LinkID ::= l<integer>} origin ::= CDATA eventInstanceID ::= IDREF {eventInstanceID ::= EventInstanceID} relatedEventInstanceID ::= IDREF {eventInstanceID ::= EventInstanceID} relatedArgumentID ::= IDREF {argumentID ::= argumentID} preposition ::= CDATA

Example of Arguments 1: 

Example of Arguments 1 John left on Saturday. <ENTITY aid=”a1”> John </ENTITY> <EVENT eid=”e1” class=”OCCURRENCE”> left </EVENT> <MAKEINSTANCE eiid=”ei1” eid=”e1” tense=”past” aspect=”none”/> <SIGNAL sid=“s1”> on </SIGNAL> <TIMEX3 tid=“t1” value=20011014T112713> Saturday </TIMEX> <ARGLINK eventInstanceID=”ei1” relatedEntityID=”a1” /> <TLINK eventInstanceId=“ei1” relatedToTime=“t1” RelType = IS_INCLUDED/>

Example of Arguments: 

Example of Arguments Police arrested the suspect in the airport on Tuesday. <ENTITY aid=”a1”> Police </ENTITY> <EVENT eid=”e1” class=”OCCURRENCE”> arrested </EVENT> <MAKEINSTANCE eiid=”ei1” eid=”e1” tense=”past” aspect=”none”/> <ENTITY aid=”a2” det=”the”> the suspect </ENTITY> in <ENTITY aid=”a3” det=”the”> the airport </ENTITY> <SIGNAL> on </SIGNAL> <TIMEX3 tid=“t1” value=20011014T112713> Saturday </TIMEX> <ARGLINK eventInstanceID=”ei1” relatedEntityID=”a1” /> <ARGLINK eventInstanceID=”ei1” relatedEntityID=”a2” /> <ARGLINK eventInstanceID=”ei1” relatedEntityID=”a3” prepostion=”in”/> <TLINK eventInstanceId=“ei1” relatedToTime=“t1” RelType = IS_INCLUDED/>

Negation over Events: Currently: 

Negation over Events: Currently Survivors were not <EVENT eid="e1" class="OCCURRENCE" pred="TEACH"> found </EVENT> <SIGNAL sid="s1"> on </SIGNAL> <TIMEX3 tid="t2" type="DATE" value="XXXX-WXX-1"> Monday </TIMEX3> <MAKEINSTANCE eventID="e1" eventInstanceID="ei1" " negation=“TRUE"> <TLINK eventInstanceID="ei1" signalID="s1" relatedToTime="t2" relType="IS_INCLUDED"/> No survivors were found. <MAKEINSTANCE eventID="e1" eventInstanceID="ei1" " negation=“FASLE"> <TLINK eventInstanceID="ei1" signalID="s1" relatedToTime="t2" relType="IS_INCLUDED"/>

Quantifiers and Negation: 1: 

Quantifiers and Negation: 1 Survivors were not found on Monday. <ENTITY aid=”a1”> Survivors </ENTITY> Were <SIGNAL > not </SIGNAL> <EVENT eid=”e1” class=”OCCURRENCE” pred=“find” tense=”past” aspect=”none”> found </EVENT> <SIGNAL sid= “s1”> on </SIGNAL> <TIMEX3 tid=“t1” value=20011014T112713> Monday </TIMEX> <ARGLINK eventID=”ei1” relatedEntityID=”a1” /> <TLINK eventId=“ei1” relatedToTime=“t1” Polarity= “NEG” RelType = IS_INCLUDED/>

Quantifiers and Negation: 2: 

Quantifiers and Negation: 2 No survivors were found on Monday. <ENTITY aid=”a1” quant=“NO”> No survivors </ENTITY> were <EVENT eid=”e1” class=”OCCURRENCE” pred=“find” tense=”past” aspect=”none”> found </EVENT> <SIGNAL sid= “s1”> on </SIGNAL> <TIMEX3 tid=“t1” value=20011014T112713> Monday </TIMEX> INTENDED INTERPRETATION <ARGLINK eventID=”ei1” relatedEntityID=”a1” /> <TLINK eventId=“ei1” relatedToTime=“t1” Polarity= “FALSE” RelType = IS_INCLUDED/> Reference to the Argument (“no survivors”) provides a resource to the interpretation function for determining the polarity of the TLINK.

TimeML German Fragment: 

TimeML German Fragment (Due to Frank Schilder, ms. 2003)

TimeML in German: 

TimeML in German Corpus study in German focussing on the preposition in. Ca. 100 occurrences of the preposition in extracted from taz articles Marked with simplified TimeML: Only TLINKS Different Aspect specification Marked with additional features (see below) Goal: definition of a semantics for the proposition in considering: Aspectual classes Granularity Reference time Schilder (2003)

German temporal and event expressions: 

German temporal and event expressions Different tense and aspect system: Usage of tenses: Present tense is ambiguous wrt. Present/future tense. No progressive form (Past perfect preferred tense in spoken language for expressing past events) Aspectual information not morphologically encoded in a consistent way Different Aktionsarten: Ingressive: verlieben (to fall in love) Exgressive: verblühen (to wither) Semelfactive: husten (to cough) Iterative: hüsteln (coughing) … No imperfective/perfective aspect Schilder (2003)

German temporal and event expressions: 

German temporal and event expressions Different syntactical structure: Prefix-verbs: ausschließen (exclude) / schließen (close): Die Bedingungen schließen einen Verkauf aus Reflexive verbs: sich entwickeln (come out) / entwickeln (develop) Complex verb constructions: sich in der Lage sehen etwas zu tun (feel capable of doing something) Sah sich die Polizei schon bisher nicht in der Lage … , dass die die Polizei sich schon bisher nicht in der Lage sah … Normally, Verbs are at position 2, but Participle verbs come at the end of a clause and Subordinate clauses and relative clauses: end of clause Schilder (2003)

Slide17: 

“Schröder hatte bereits am Wochenende signalisiert, dass er eine dritte Amtszeit anstrebt.” 29.8.03 <ENTITY aid=”a1”> Schröder </ENTITY> hatte bereits <SIGNAL> am </SIGNAL> <TIMEX3 tid=“t1” type = “date” value=YYYY-XX-WE> Wochenende </TIMEX> <EVENT eid=”e1” class=”OCCURRENCE”> signalisiert </EVENT> <MAKEINSTANCE eiid=”ei1” eid=”e1” tense=”past” aspect=”none”/> <SLINK eventInstanceID="ei1" subordinatedEvent="ei2 relType="MODAL"/> <TLINK eventInstanceId=“ei1” relatedToTime=“t1” RelType = IS_INCLUDED/> dass <ENTITY aid=”a2” > er </ENTITY> <ENTITY aid=”a3” det=”eine”> eine dritte Amtszeit </ENTITY> <EVENT eid=”e1” class=”OCCURRENCE”> anstrebt </EVENT> <MAKEINSTANCE eiid=”ei2” eid=”e2” tense=”present” aspect=”none”/> <ARGLINK eventInstanceID=”ei1” relatedEntityID=”a1” /> <ARGLINK eventInstanceID=”ei1” relatedEventID=”ei2” /> <ARGLINK eventInstanceID=”ei2” relatedEntityID=”a2” /> <ARGLINK eventInstanceID=”ei2” relatedEntityID=”a3” />

Outstanding Problems: 

Outstanding Problems

Semantics of TimeML: 

Semantics of TimeML A text T is satisfied by a model M iff there are functions fe: Dome -> Pow(E), and fei: Domei -> E ft: Domt -> I , such that for all tags t Tag(T), t is satisfied by fe fei and ft in M. A tag t is satisfied by fe,ft, and fei in M iff if t has the form “<EVENT eid =  class =  pred=  >” then fe() = Val() “<TIMEX3 tid =  type = DATE value=  >” then ft() = Val() “<TLINK eventInstanceID =  relatedtoTime =  relType= ‘IS_INCLUDED’>” then (fei())  ft ( )

Problems for Interpretation: 

Problems for Interpretation Negation John didn’t teach on Tuesday <EVENT eid = “ei1”…pred=“TEACH”> <MAKEINSTANCE eiid =“eii1”…negation=“true”> <TIMEX3 tid=“t1” val=“XXXX-WXX-2”> <TLINK relatedToTime =“t1” eventInstance=“eii1” relation=“IS_INCLUDED”> -> SCOPE for negation Multiple Events John taught twice on Tuesday <EVENT eid = “ei1”…pred=“TEACH”> <MAKEINSTANCE eiid =“eii1” … event=“e1” cardinality=2> <TIMEX3 tid=“t1” val=“XXXX-WXX-2”> <TLINK relatedToTime =“t1” eventInstance=“eii1” relation=“IS_INCLUDED”> “<MAKEINSTANCE eiid =  eid =  negation=‘FALSE’ modal = ‘’ cardinality=>” then fei()  fe() <MAKEINSTANCE eiid =“eii1” … event=“e1” > <MAKEINSTANCE eiid =“eii2” … event=“e1” > Condition on Embedding Functions

Problems for TimeML: 

Problems for TimeML Set-valued Times John taught <TIMEX3 tid=”t4” type=”SET” value=”P1M” quant=”EVERY” freq=”3D”> three days every month</TIMEX3> PROBLEM: the temporal identifier can’t be interpreted as denoting a particular interval of time, it must be a set of intervals (or even a set of sets of intervals!) Disjunction John taught on Monday or on Wednesday

Some Solutions: 

Some Solutions Negation: Use TLINK as a scope domain, eliminate MAKEINSTANCE John didn’t teach on Tuesday <EVENT eid = “ei1”…pred=“TEACH”> <TIMEX3 tid=“t1” val=“XXXX-WXX-2”> <TLINK relatedToTime =“t1” eventID=“ei1” relation=“IS_INCLUDED” Polarity=“false”> New TLINK Rules “<TLINK eventID =  relatedtoTime =  relType= “IS_INCLUDED” Polarity=“true”>” there is an e  E such that e  fe() and (e)  ft ( ) “<TLINK eventID =  relatedtoTime =  relType= “IS_INCLUDED” Polarity=“false”>” there is no e  E such that e  fe() and (e)  ft ( )

Some Solutions: 

Some Solutions Multiple events Add cardinality element to the TLINK John taught twice on Tuesday <EVENT eid = “ei1”…pred=“TEACH”> <TIMEX3 tid=“t1” val=“XXXX-WXX-2”> <TLINK relatedToTime =“t1” eventID=“ei1” relation=“IS_INCLUDED” Polarity=“true” cardinality=“2”> “<TLINK eventID =  relatedtoTime =  relType= “IS_INCLUDED” Polarity=“true” cardinality=>” is satisfied iff there are Val() distinct e  E such that e  fe() and (e)  ft ( )

Harder Problems: 

Harder Problems Vagueness When he left, shortly after 5 am Tuesday, he discovered someone had smashed a window. Appavu has been involved with healthcare standards development for about a decade, an interest he developed shortly after he began working with information systems at Cook County. Domino's Pizza of Washington reported that they delivered "In excess" of 100 large pizzas to the White House late this afternoon. It was then,early in December of 1977, that he went to the NORML conference.

Vagueness: 

Vagueness Current Treatment: <TIMEX3 tid=“t1” val=“20030829TAFT” mod=“END”> late this afternoon </TIMEX3> <TIMEX3 tid=“t1” val=“197712” mod=“START”> early in December of 1977 </TIMEX3> Problem: No semantics for mod attributes means no possibility for doing reasoning. It was then,early in December of 1977, that he went to the NORML conference. Two weeks later he was a convert. Before or after Christmas? We might fake a solution by being overly general: Interpret START to mean “the first half of”

Current Treatment: 

Current Treatment No general solution for mod values: Shortly after 5am -> minutes Shortly after he began working -> weeks or months

Semantic Weakness: 

Semantic Weakness Simple annotation of temporal relations is too week: President John F. Kennedy's gravesite at Arlington National Cemetery has been restored to its original condition, after someone tried unsuccessfully to dig up some of its granite paving stones. South Africa, after losing the toss, were bowled out for 107 against England. How long after? Days or weeks An hour or two. This is not generally encoded overtly.

Context Dependent Vagueness: 

Context Dependent Vagueness If we did code this, lots of world-knowledge based information could be encoded by annotators: They ate lunch early on Monday. They ate dinner early on Monday. They ate breakfast early on Monday. Probably before noon in the early evening in the very early morning

Questions: 

Questions How to talk about a “likely distribution” in time? How to compare such annotations?

TimeML-enabled Applications: 

TimeML-enabled Applications

Web-basedTemporal Reasoning: 

Web-based Temporal Reasoning Web Negotiation Agents (Brokers) Scheduling Programs

Semantic Web: 

Semantic Web Delivery within five business days. <EVENT eid="e1" class="OCCURRENCE" tense="NONE" aspect="NONE"> order </EVENT> <MAKEINSTANCE eiid="ei1" eventID="e1"/> <EVENT eid="e2" class="OCCURRENCE" tense=nil aspect="PERFECTIVE"> delivery </EVENT> <MAKEINSTANCE eiid="ei2" eventID="e2"/> <SIGNAL sid="s1"> within </SIGNAL> <TIMEX3 tid="t1" type="DURATION" value=nil temporalFunction="true"> five business days </TIMEX3> <TLINK eventInstanceID="ei2" signalID="s1" relatedToEvent="ei1” relType="AFTER" /> <TLINK eventInstanceID="ei2" signalID="s1" relatedToTime="t1” relType="IS_INCLUDED" /> <TLINK eventInstanceID="ei1" signalID="s1" relatedToTime="t1” relType=" BEGINS" />

Scheduling Issues: 

Scheduling Issues Mary teaches on Mondays and Wednesdays in the fall. Sophie goes to daycare on Thursday and Friday at 4:00pm in October.

www.TimeML.org: 

www.TimeML.org