Txt/Hyp term similarities

The rows are the txt words. The columns are hyp words.
Resource summary:
Acronym: AcronymLexicalResource
BasicWN: BasicWNLexicalResource
Country: CountryLexicalResource
Cyc: null
DekangLin: DekangLinLexicalResource
Google: null
InfoMap: InfoMapLexicalResource
NomBank: NomBankLexicalResource
Number: NumberLexicalResource
Ordinal: OrdinalLexicalResource
Preposition: PrepositionLexicalResource
Ravichandran: RavichandranLexicalResource
ResnikWN: ResnikWNLexicalResource
StringSim: StringSimLexicalResource


Inference ID: Brandeis-01.1

Txt: StatesWest Airlines withdrew its offer to acquire Mesa Airlines.

Hyp: An offer was made to acquire Mesa Airline . (Yes)

An
DT		 offer
NN		 was
VBD		 made
VBN		 to
TO		 acquire
VB		 Mesa_Airline
NNP		 .
.
StatesWest_Airlines:NNS 20.50 10.46 15.17 15.46 20.50 15.46   9.47 20.50
withdrew:VBD 20.00 12.60 10.00   7.27 20.00   5.67 15.46 20.00
its:PRP$ 20.00 12.00 15.00 15.00 20.00 15.00 12.50 20.00
offer:NN 20.00   0.00 15.00 12.60 20.00 10.39 10.46 19.35
to:TO 10.00 20.00 20.00 20.00   0.00 20.00 20.50 10.00
acquire:VB 20.00 10.39 10.00   6.84 20.00   0.00 15.46 20.00
Mesa_Airlines:NNS 20.50 10.46 13.67 15.46 20.50 15.46   0.00 20.50
.:. 10.00 19.35 20.00 17.07 10.00 20.00 20.50   0.00
NO_WORD   1.00 10.00   1.00 10.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -13.2714
Features matched: NullPunisher.aux: was; NullPunisher.article: An; RootEntailment.poorlyAlignedRoot: "made" aligned badly to "withdrew"; Structure.parentsMismatch: args have different parents, different relations: text "acquire" <-infmod-- "offer" vs. hyp "acquire" <-xcomp-- "made", which aligned to text "withdrew"
Hand-tuned score: -4.1500
Threshold: -11.4590

Inference ID: Brandeis-01.2

Txt: StatesWest Airlines withdrew its offer to acquire Mesa Airlines.

Hyp: StatesWest did plan to acquire Mesa Airlines . (Yes)

StatesWest
NNP		 did
VBD		 plan
NN		 to
TO		 acquire
VB		 Mesa_Airlines
NNS		 .
.
StatesWest_Airlines:NNS   0.00 15.46   9.52 20.50 15.46   0.00 20.50
withdrew:VBD 15.46   7.45 12.60 20.00   5.67 15.46 20.00
its:PRP$ 12.50 15.00 12.00 20.00 15.00 12.50 20.00
offer:NN 10.46 12.85   5.11 20.00 10.39 10.46 19.35
to:TO 20.50 20.00 20.00   0.00 20.00 20.50 10.00
acquire:VB 15.46   9.97 11.85 20.00   0.00 15.46 20.00
Mesa_Airlines:NNS   9.96 15.46   9.52 20.50 15.46   0.00 20.50
.:. 20.50 17.99 19.69 10.00 20.00 20.50   0.00
NO_WORD 10.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -14.5600
Features matched: RootEntailment.poorlyAlignedRoot: "did" aligned badly to "withdrew"; Structure.parentsMismatch: args have different parents, different relations: text "acquire" <-infmod-- "offer" vs. hyp "acquire" <-xcomp-- "did", which aligned to text "withdrew"
Hand-tuned score: -4.0000
Threshold: -11.4590

Inference ID: Brandeis-01.3

Txt: StatesWest Airlines withdrew its offer to acquire Mesa Airlines.

Hyp: StatesWest did acquire Mesa Airlines . (No)

StatesWest
NNP		 did
VBD		 acquire
VB		 Mesa_Airlines
NNS		 .
.
StatesWest_Airlines:NNS   0.00 15.46 15.46   0.00 20.50
withdrew:VBD 15.46   7.45   5.67 15.46 20.00
its:PRP$ 12.50 15.00 15.00 12.50 20.00
offer:NN 10.46 12.85 10.39 10.46 19.35
to:TO 20.50 20.00 20.00 20.50 10.00
acquire:VB 15.46   9.97   0.00 15.46 20.00
Mesa_Airlines:NNS   9.96 15.46 15.46   0.00 20.50
.:. 20.50 17.99 20.00 20.50   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -5.0000
Features matched: Factive.inPositiveEmbedding: embedded positive text; NullPunisher.aux: did; Structure.argsMismatch: args have different parents but same relations: text "StatesWest_Airlines" <-nsubj-- "withdrew vs. hyp "StatesWest" <-nsubj-- "acquire", which aligned to text "acquire" args have different parents but same relations: text "." <-punct-- "withdrew vs. hyp "." <-punct-- "acquire", which aligned to text "acquire"
Hand-tuned score: -1.0500
Threshold: -11.4590

Inference ID: Brandeis-02

Txt: Steynar Gil welcomed the release of prisoners imprisoned during the October 15-16 riots.

Hyp: Prisoners imprisoned during October 15-16 riots have been released . (Yes)

Prisoners
NNS		 imprisoned
VBN		 October
NNP		 15-16
CD		 riots
NNS		 have
VBP		 been
VBN		 released
VBN		 .
.
Steynar_Gil:NNP 10.46 15.46 14.96 24.96 10.46 15.46 15.46 15.46 20.50
welcomed:VBD 15.00 10.00 14.19 20.46 12.10   4.61 10.00   7.10 18.83
the:DT 20.00 20.00 20.50 20.50 20.00 20.00 20.00 20.00 10.00
release:NN   8.95 15.00   9.19 20.46   7.10 12.80 12.74   0.50 19.68
prisoners:NNS   0.00 10.00 10.50 19.96   8.86 13.05 14.34 13.95 19.74
imprisoned:VBN 10.00   0.00 15.50 19.21 11.95 10.00 10.00   8.73 20.00
the:DT 20.00 20.00 20.50 20.50 20.00 20.00 20.00 20.00 10.00
October:NNP 10.50 15.50   0.00 24.96   9.19 15.50 15.50 14.19 20.50
15-16:CD 20.46 19.21 24.96   0.00 20.46 20.46 20.46 20.40 19.68
riots:NNS 10.00 11.95   9.19 20.46   0.00 13.07 13.07 12.10 20.00
.:. 20.00 20.00 20.50 19.68 20.00 20.00 20.00 20.00   0.00
NO_WORD 10.00 10.00 10.00 10.00 10.00   1.00   1.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -6.5000
Features matched: Date.matchDatesByNormForm: hyp/txt matching, by normalized form: 10/01/1000; NullPunisher.aux: have; NullPunisher.aux: been; Structure.argsMismatch: args have different parents but same relations: text "prisoners" <-nsubjpass-- "imprisoned vs. hyp "Prisoners" <-nsubjpass-- "released", which aligned to text "release" args have different parents but same relations: text "." <-punct-- "welcomed vs. hyp "." <-punct-- "released", which aligned to text "release" text "prisoners" is prep_of of "release" while hyp "Prisoners" is nsubjpass of "released" which aligned to text "release"
Hand-tuned score: -1.1000
Threshold: -11.4590

Inference ID: Brandeis-03

Txt: Program trading is hurting the market's efforts to bring back small investors.

Hyp: There are some efforts to bring_back small investors . (Yes)

There
EX		 are
VBP		 some
DT		 efforts
NNS		 to
TO		 bring_back
VB		 small
JJ		 investors
NNS		 .
.
Program_trading:NN 20.00 14.34 20.00   8.35 20.00 13.66 11.34 10.00 20.00
is:VBZ 20.00   0.50 20.00 15.00 20.00   8.35 11.34 14.34 20.00
hurting:VBG 20.00 10.00 20.00 12.03 20.00 10.00 10.16 12.70 18.27
the:DT 10.00 20.00 10.00 20.00 10.00 20.00 20.00 20.00 10.00
market:NN 20.00 15.00 20.00   7.72 20.00 12.62 10.69   4.93 19.55
efforts:NNS 20.00 15.00 20.00   0.00 20.00 13.86 10.38   8.17 18.92
to:TO 10.00 20.00 10.00 20.00   0.00 20.00 20.00 20.00 10.00
bring_back:VB 20.00 10.00 20.00 13.86 20.00   0.00 10.34 14.02 20.00
small:JJ 20.00 10.69 20.00 10.38 20.00 10.34   0.00 10.49 18.52
investors:NNS 20.00 15.00 20.00   8.17 20.00 14.02 10.49   0.00 20.00
.:. 10.00 20.00 10.00 18.92 10.00 20.00 18.52 20.00   0.00
NO_WORD   1.00 10.00 10.00 10.00 10.00 10.00   9.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -15.5000
Features matched: NullPunisher.other: some; NullPunisher.functionWord: There; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "hurting vs. hyp "." <-punct-- "are", which aligned to text "is" args have different parents, different relations: text "efforts" <-dobj-- "hurting" vs. hyp "efforts" <-nsubj-- "are", which aligned to text "is"
Hand-tuned score: -3.1000
Threshold: -11.4590

Inference ID: Brandeis-04

Txt: Thirty-two of the 159 U.N. members were honoring the sanctions against Iraq.

Hyp: There were sanctions against Iraq . (Yes)

There
EX		 were
VBD		 sanctions
NNS		 Iraq
NNP		 .
.
Thirty-two:NN 20.00 14.96   9.96 10.46 20.00
the:DT 10.00 20.00 20.00 20.50 10.00
159:NNP 20.00 14.96   9.28 10.46 18.60
U.N.:NNP 20.50 15.46 10.46 13.65 20.50
members:NNS 20.00 14.34   8.03   9.84 19.57
were:VBD 15.00   0.00 13.07 14.84 20.00
honoring:VBG 20.00 10.00 12.10 15.50 20.00
the:DT 10.00 20.00 20.00 20.50 10.00
sanctions:NNS 20.00 13.07   0.00 10.50 20.00
Iraq:NNP 20.50 14.84 10.50   0.00 20.50
.:. 10.00 20.00 20.00 20.50   0.00
NO_WORD   1.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -5.0000
Features matched: NullPunisher.functionWord: There; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "honoring vs. hyp "." <-punct-- "were", which aligned to text "were" args have different parents, different relations: text "sanctions" <-dobj-- "honoring" vs. hyp "sanctions" <-nsubj-- "were", which aligned to text "were"
Hand-tuned score: -2.1000
Threshold: -11.4590

Inference ID: Brandeis-05.1

Txt: Herbicide use in some areas of the U.S. was delayed earlier in the year by heavy rains.

Hyp: Herbicides were delayed in all areas of the U.S. this year . (No)

Herbicides
NNS		 were
VBD		 delayed
VBN		 all
DT		 areas
NNS		 the
DT		 U.S.
NNP		 this
DT		 year
NN		 .
.
Herbicide:JJ   3.00   9.34 12.00 20.00 11.34 20.00 11.84 20.00 12.00 20.00
use:NN 10.00   9.62 13.07 20.00   7.47 20.00 10.50 20.00   8.69 18.38
some:DT 20.00 20.00 20.00 10.00 20.00 10.00 20.50   8.49 20.00 10.00
areas:NNS   9.34 14.34 13.69 20.00   0.00 20.00   7.61 20.00   8.69 18.59
the:DT 20.00 20.00 20.00 10.00 20.00   0.00 20.50 10.00 20.00 10.00
U.S.:NNP   9.84 14.84 15.50 20.50   7.61 20.50   0.00 20.50   8.31 20.50
was:VBD 14.34   0.50 10.00 20.00 12.11 20.00 11.83 20.00 15.00 20.00
delayed:VBN 15.00   8.07   0.00 20.00 13.69 20.00 15.50 20.00 12.84 19.88
earlier:RB 14.96 19.96 17.64 20.00 14.96 20.00 15.46 20.00 12.09 19.27
the:DT 20.00 20.00 20.00 10.00 20.00   0.00 20.50 10.00 20.00 10.00
year:NN 10.00 15.00 12.84 20.00   8.69 20.00   8.31 20.00   0.00 17.10
heavy:JJ 11.34 11.34 10.35 20.00   9.47 20.00 11.84 20.00 12.00 18.07
rains:NNS   7.34 12.34 13.51 20.00   6.86 20.00   9.84 20.00   8.69 20.00
.:. 20.00 20.00 19.88 10.00 18.59 10.00 20.50 10.00 17.10   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00   1.00 10.00 10.00 10.00 10.00

Response: dontknow (INCORRECT)
Justification:
Alignment score: -29.5000
Features matched: Adjunct.dropPosCxt: text adjunct "earlier" of "delayed" dropped on aligned hyp word "delayed"; Antonym.samePol: matching polarity with antonyms: all & some; Quant.expand: [some,all]; Structure.argsMismatch: args have different parents but same relations: text "areas" <-prep_in-- "use vs. hyp "areas" <-prep_in-- "delayed", which aligned to text "delayed" args have different parents, different relations: text "Herbicide" <-amod-- "use" vs. hyp "Herbicides" <-nsubjpass-- "delayed", which aligned to text "delayed" args have different parents, different relations: text "year" <-prep_in-- "earlier" vs. hyp "year" <-tmod-- "delayed", which aligned to text "delayed"
Hand-tuned score: -16.5000
Threshold: -11.4590

Inference ID: Brandeis-05.2

Txt: Herbicide use in some areas of the U.S. was prevented this year because of heavy rains.

Hyp: Herbicides were used this year in the U.S. . (Yes)

Herbicides
NNS		 were
VBD		 used
VBN		 this
DT		 year
NN		 the
DT		 U.S.
NNP		 .
.
Herbicide:JJ   3.00   9.34 12.00 20.00 12.00 20.00 11.84 20.00
use:NN 10.00   9.62   0.50 20.00   8.69 20.00 10.50 18.38
some:DT 20.00 20.00 20.00   8.49 20.00 10.00 20.50 10.00
areas:NNS   9.34 14.34 12.47 20.00   8.69 20.00   7.61 18.59
the:DT 20.00 20.00 20.00 10.00 20.00   0.00 20.50 10.00
U.S.:NNP   9.84 14.84 15.50 20.50   8.31 20.50   0.00 20.50
was:VBD 14.34   0.50 10.00 20.00 15.00 20.00 11.83 20.00
prevented:VBN 15.00 10.00   6.99 20.00 15.00 20.00 15.50 18.96
this:DT 20.00 20.00 20.00   0.00 20.00 10.00 20.50 10.00
year:NN 10.00 15.00 13.69 20.00   0.00 20.00   8.31 17.10
heavy:JJ 11.34 11.34   9.47 20.00 12.00 20.00 11.84 18.07
rains:NNS   7.34 12.34 11.94 20.00   8.69 20.00   9.84 20.00
.:. 20.00 20.00 18.25 10.00 17.10 10.00 20.50   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00   1.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -12.5000
Features matched: Adjunct.dropPosCxt: text adjunct "rains" of "year" dropped on aligned hyp word "year"; NullPunisher.aux: were; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "prevented vs. hyp "." <-punct-- "used", which aligned to text "use" args have different parents, different relations: text "year" <-dobj-- "prevented" vs. hyp "year" <-tmod-- "used", which aligned to text "use" args have different parents, different relations: text "U.S." <-prep_of-- "areas" vs. hyp "U.S." <-prep_in-- "used", which aligned to text "use" text "Herbicide" is amod of "use" while hyp "Herbicides" is nsubjpass of "used" which aligned to text "use"
Hand-tuned score: -1.5500
Threshold: -11.4590

Inference ID: Brandeis-06

Txt: The US is bolstering its military presence in the Gulf.

Hyp: The US does have a military presence in the Gulf . (Yes)

The
DT		 US
NNP		 does
VBZ		 have
VB		 a
DT		 military
JJ		 presence
NN		 the
DT		 Gulf
NNP		 .
.
The:DT   0.00 20.50 20.00 20.00 10.00 20.00 20.00   0.00 20.50 10.00
US:NNP 20.50   0.00 14.84 14.84 20.50 12.50   7.61 20.50   7.52 20.50
is:VBZ 20.00 14.84   9.34   7.80 20.00 12.00 14.34 20.00 14.84 20.00
bolstering:VBG 20.00 14.84   8.95   7.62 20.00 10.71 13.03 20.00 14.84 19.81
its:PRP$ 20.00 12.50 15.00 15.00 20.00 15.00 12.00 20.00 12.50 20.00
military:JJ 20.00 12.50   9.59 12.00 20.00   0.00   7.88 20.00 12.50 20.00
presence:NN 20.00   7.61 14.33 14.34 20.00   7.88   0.00 20.00   8.02 19.91
the:DT   0.00 20.50 18.65 20.00 10.00 20.00 20.00   0.00 20.50 10.00
Gulf:NNP 20.50   7.52 14.84 14.84 20.50 12.50   8.02 20.50   0.00 20.50
.:. 10.00 20.50 20.00 20.00 10.00 20.00 19.91 10.00 20.50   0.00
NO_WORD   1.00 10.00   1.00 10.00   1.00   9.00 10.00   1.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -11.6174
Features matched: NullPunisher.aux: does; NullPunisher.article: a; RootEntailment.poorlyAlignedRoot: "have" aligned badly to "bolstering"
Hand-tuned score: -1.1500
Threshold: -11.4590

Inference ID: Brandeis-07

Txt: The unemployed took to the streets of the German capital, Berlin, mirroring protests around the country.

Hyp: The protests were only in Berlin . (No)

The
DT		 protests
NNS		 were
VBD		 only
RB		 Berlin
NNP		 .
.
The:DT   0.00 20.00 20.00 20.00 20.50 10.00
unemployed:JJ 20.00 10.82 11.96 11.96 12.46 20.00
took:VBD 20.00   9.63   4.35 19.96 15.50 18.92
the:DT   0.00 20.00 20.00 20.00 20.50 10.00
streets:NNS 20.00   5.29 14.34 14.96   8.53 19.83
the:DT   0.00 20.00 20.00 20.00 20.50 10.00
German_capital:NN 20.00 10.00 14.34 14.96   0.50 20.00
,:, 10.00 20.00 20.00 20.00 20.50   5.73
Berlin:NNP 20.50 10.50 14.84 15.46   0.00 20.50
,:, 10.00 20.00 20.00 20.00 20.50   5.73
mirroring:NN 20.00   8.51 12.74 14.96   7.62 20.00
protests:NNS 20.00   0.00 15.00 14.96 10.50 20.00
the:DT   0.00 20.00 20.00 20.00 20.50 10.00
country:NN 20.00   6.89 14.34 14.96   6.57 18.25
.:. 10.00 20.00 20.00 20.00 20.50   0.00
NO_WORD   1.00 10.00 10.00   9.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -18.3499
Features matched: Adjunct.addPosCxt: hyp added only[only-RB]; Adjunct.dropPosCxt: text adjunct "country" of "took" dropped on aligned hyp word "were"; NullPunisher.article: The; NullPunisher.other: only; RootEntailment.poorlyAlignedRoot: "were" aligned badly to "took"; Structure.argsMismatch: args have different parents but same relations: text "protests" <-appos-- "Berlin vs. hyp "protests" <-nsubj-- "were", which aligned to text "took" args have different parents, different relations: text "Berlin" <-appos-- "German_capital" vs. hyp "Berlin" <-prep_in-- "were", which aligned to text "took"
Hand-tuned score: -5.6000
Threshold: -11.4590

Inference ID: Brandeis-08.1

Txt: The court nullified a standstill agreement between DPC Acquisition and Dataproducts.

Hyp: There was some agreement between DPC Acquisition and Dataproducts . (Yes)

There
EX		 was
VBD		 some
DT		 agreement
NN		 DPC_Acquisition
NNP		 Dataproducts
NNP		 .
.
The:DT 10.00 20.00 10.00 20.00 20.50 20.50 10.00
court:NN 20.00 12.11 20.00   8.35   9.68 10.46 20.00
nullified:VBN 20.00 10.00 20.00 14.05 15.46 15.46 19.40
a:DT 10.00 20.00 10.00 20.00 20.50 20.50 10.00
standstill:JJ 20.00 12.00 20.00   8.69 12.46 12.46 19.59
agreement:NN 20.00 15.00 20.00   0.00   8.56 10.46 19.30
DPC_Acquisition:NNP 20.50 15.46 20.50   8.56   0.00 14.96 20.50
Dataproducts:NNP 20.50 15.46 20.50 10.46 14.96   0.00 20.50
.:. 10.00 20.00 10.00 19.30 20.50 20.50   0.00
NO_WORD   1.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -23.0000
Features matched: Adjunct.dropPosCxt: text adjunct "standstill" of "agreement" dropped on aligned hyp word "agreement"; NullPunisher.functionWord: There; Quant.contract: [a,some]; RootEntailment.poorlyAlignedRoot: "was" aligned badly to "nullified"; Structure.relMismatch: text "agreement" is dobj of "nullified" while hyp "agreement" is nsubj of "was" which aligned to text "nullified"
Hand-tuned score: -0.6000
Threshold: -11.4590

Inference ID: Brandeis-08.2

Txt: The court nullified a standstill agreement between DPC Acquisition and Dataproducts.

Hyp: There are some agreements between DCP Acquisition and Dataproducts . (MULTIPLE ANSWERS)

There
EX		 are
VBP		 some
DT		 agreements
NNS		 DCP_Acquisition
NNP		 Dataproducts
NNP		 .
.
The:DT 10.00 20.00 10.00 20.00 20.50 20.50 10.00
court:NN 20.00 15.00 20.00   8.35   9.68 10.46 20.00
nullified:VBN 20.00 10.00 20.00 13.34 15.46 15.46 19.40
a:DT 10.00 20.00 10.00 20.00 20.50 20.50 10.00
standstill:JJ 20.00 12.00 20.00   8.73 12.46 12.46 19.59
agreement:NN 20.00 13.69 20.00   0.50   8.56 10.46 19.30
DPC_Acquisition:NNP 20.50 15.46 20.50   8.56   0.00 14.96 20.50
Dataproducts:NNP 20.50 15.46 20.50 10.46 14.96   0.00 20.50
.:. 10.00 20.00 10.00 18.89 20.50 20.50   0.00
NO_WORD   1.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -23.5000
Features matched: Adjunct.dropPosCxt: text adjunct "standstill" of "agreement" dropped on aligned hyp word "agreements"; NullPunisher.other: some; NullPunisher.functionWord: There; Quant.contract: [a,some]; RootEntailment.poorlyAlignedRoot: "are" aligned badly to "nullified"; Structure.relMismatch: text "agreement" is dobj of "nullified" while hyp "agreements" is nsubj of "are" which aligned to text "nullified"
Hand-tuned score: -1.6000
Threshold: -11.4590

Inference ID: Brandeis-09.1

Txt: Sardar Patel faced imprisonment for the first time when he was assisting Gandhiji in the Salt Satyagraha.

Hyp: Sardar Patel was in prison . (MULTIPLE ANSWERS)

Sardar_Patel
NNP		 was
VBD		 prison
NN		 .
.
Sardar_Patel:NNP   0.00 15.46 10.46 20.50
faced:VBD 15.46   7.52 12.14 19.26
imprisonment:NN 10.46 15.00   2.91 20.00
the:DT 20.50 20.00 20.00 10.00
first:JJ 12.46 11.34   9.44 20.00
time:NN 10.46 15.00   6.95 17.52
when:WRB 20.46 19.96 19.96 10.00
he:PRP 12.50 15.00 12.00 20.00
was:VBD 15.46   0.00 14.34 20.00
assisting:VBG 15.46 10.00 12.37 19.30
Gandhiji:NNP   9.96 15.46 10.46 20.50
the:DT 20.50 20.00 20.00 10.00
Salt_Satyagraha:NNP 14.96 15.17 10.17 20.50
.:. 20.50 20.00 19.98   0.00
NO_WORD 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -8.9131
Features matched: Location.mismatch: no clear info of matching: be(X, prep_in); Structure.argsMismatch: args have different parents but same relations: text "Sardar_Patel" <-nsubj-- "faced vs. hyp "Sardar_Patel" <-nsubj-- "was", which aligned to text "was" args have different parents but same relations: text "." <-punct-- "faced vs. hyp "." <-punct-- "was", which aligned to text "was" args have different parents, different relations: text "imprisonment" <-dobj-- "faced" vs. hyp "prison" <-prep_in-- "was", which aligned to text "was"
Hand-tuned score: -4.0000
Threshold: -11.4590

Inference ID: Brandeis-09.2

Txt: Sardar Patel faced imprisonment for the first time when he was assisting Gandhiji in the Salt Satyagraha.

Hyp: Sardar Patel was convicted of a crime . (MULTIPLE ANSWERS)

Sardar_Patel
NNP		 was
VBD		 convicted
VBN		 a
DT		 crime
NN		 .
.
Sardar_Patel:NNP   0.00 15.46 15.46 20.50 10.46 20.50
faced:VBD 15.46   7.52   5.77 20.00 12.38 19.26
imprisonment:NN 10.46 15.00   7.95 20.00   6.74 20.00
the:DT 20.50 20.00 20.00 10.00 20.00 10.00
first:JJ 12.46 11.34 10.95 20.00   9.72 20.00
time:NN 10.46 15.00 12.60 20.00   5.00 17.52
when:WRB 20.46 19.96 19.96 10.00 19.96 10.00
he:PRP 12.50 15.00 15.00 20.00 12.00 20.00
was:VBD 15.46   0.00   9.34 20.00 15.00 20.00
assisting:VBG 15.46 10.00   6.15 20.00 11.90 19.30
Gandhiji:NNP   9.96 15.46 15.46 20.50 10.46 20.50
the:DT 20.50 20.00 20.00 10.00 20.00 10.00
Salt_Satyagraha:NNP 14.96 15.17 15.17 20.50   9.68 20.50
.:. 20.50 20.00 20.00 10.00 19.93   0.00
NO_WORD 10.00   1.00 10.00   1.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -15.7673
Features matched: Adjunct.dropPosCxt: text adjunct "first" of "time" dropped on aligned hyp word "crime"; NullPunisher.article: a; NullPunisher.aux: was; Quant.contract: [the,a]; RootEntailment.poorlyAlignedRoot: "convicted" aligned badly to "faced"; Structure.relMismatch: text "Sardar_Patel" is nsubj of "faced" while hyp "Sardar_Patel" is nsubjpass of "convicted" which aligned to text "faced" text "time" is prep_for of "faced" while hyp "crime" is prep_of of "convicted" which aligned to text "faced"
Hand-tuned score: -0.6500
Threshold: -11.4590

Inference ID: Brandeis-10.1

Txt: Darryl Strawberry recently avoided imprisonment when a judge sentenced him to a drug treatment center for violating his probation.

Hyp: Darryl Strawberry was in prison . (No)

Darryl_Strawberry
NNP		 was
VBD		 prison
NN		 .
.
Darryl_Strawberry:NNP   0.00 15.17   9.97 20.50
recently:RB 15.46 19.96 14.85 17.72
avoided:VBD 15.46 10.00 14.66 18.85
imprisonment:NN 10.46 15.00   2.91 20.00
when:WRB 20.46 19.96 19.96 10.00
a:DT 20.50 20.00 20.00 10.00
judge:NN   9.55 14.34   5.62 20.00
sentenced:VBD 14.84 10.00   6.40 20.00
him:PRP 12.50 15.00 12.00 20.00
a:DT 20.50 20.00 20.00 10.00
drug:NN 10.17 14.34   9.34 19.56
treatment:NN   9.26 15.00   8.13 19.73
center:NN   9.17 12.11   4.75 20.00
for:IN 20.50 20.00 20.00 20.00
violating:VBG 15.46 10.00 11.89 19.34
his:PRP$ 12.50 15.00 12.00 20.00
probation:NN   9.84 15.00   2.62 20.00
.:. 20.50 20.00 19.98   0.00
NO_WORD 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -14.6238
Features matched: Adjunct.dropPosCxt: text adjunct "recently" of "avoided" dropped on aligned hyp word "was"; Location.mismatch: no clear info of matching: be(X, prep_in); RootEntailment.poorlyAlignedRoot: "was" aligned badly to "avoided"; Structure.parentsMismatch: args have different parents, different relations: text "probation" <-dobj-- "violating" vs. hyp "prison" <-prep_in-- "was", which aligned to text "avoided"
Hand-tuned score: -5.5000
Threshold: -11.4590

Inference ID: Brandeis-10.2

Txt: Darryl Strawberry recently avoided imprisonment when a judge sentenced him to a drug treatment center for violating his probation.

Hyp: Darryl was imprisoned for violating his probation . (No)

Darryl
NNP		 was
VBD		 imprisoned
VBN		 for
IN		 violating
VBG		 his
PRP$		 probation
NN		 .
.
Darryl_Strawberry:NNP   0.00 15.17 15.46 20.50 15.46 12.50   9.84 20.50
recently:RB 15.46 19.96 19.11 20.00 18.39 15.00 14.90 17.72
avoided:VBD 15.46 10.00   9.18 20.00   8.04 15.00 14.84 18.85
imprisonment:NN 10.46 15.00   1.00 20.00 12.08 12.00   4.19 20.00
when:WRB 20.46 19.96 19.96 20.00 19.96 20.00 19.96 10.00
a:DT 20.50 20.00 20.00 18.59 20.00 20.00 20.00 10.00
judge:NN 10.46 14.34 14.03 20.00   9.47 12.00   5.61 20.00
sentenced:VBD 15.46 10.00   3.67 20.00   6.30 15.00   7.38 20.00
him:PRP 12.50 15.00 15.00 20.00 15.00   7.59 12.00 20.00
a:DT 20.50 20.00 20.00 18.59 20.00 20.00 20.00 10.00
drug:NN 10.46 14.34 14.92 20.00 12.79 12.00   9.92 19.56
treatment:NN 10.46 15.00 14.20 20.00 14.34 12.00   8.35 19.73
center:NN 10.46 12.11 15.00 20.00 15.00 12.00   8.35 20.00
for:IN 20.50 20.00 20.00   0.00 20.00 20.00 20.00 20.00
violating:VBG 15.46 10.00   7.81 20.00   0.00 15.00 11.69 19.34
his:PRP$ 12.50 15.00 15.00 20.00 15.00   0.00 12.00 20.00
probation:NN 10.46 15.00 10.81 20.00 11.69 12.00   0.00 20.00
.:. 20.50 20.00 20.00 20.00 19.34 20.00 20.00   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -8.0000
Features matched: NullPunisher.aux: was; Structure.argsMismatch: args have different parents but same relations: text "for" <-prep-- "sentenced vs. hyp "for" <-prep-- "imprisoned", which aligned to text "imprisonment" args have different parents but same relations: text "." <-punct-- "avoided vs. hyp "." <-punct-- "imprisoned", which aligned to text "imprisonment" args have different parents, different relations: text "Darryl_Strawberry" <-nsubj-- "avoided" vs. hyp "Darryl" <-nsubjpass-- "imprisoned", which aligned to text "imprisonment"
Hand-tuned score: -2.0500
Threshold: -11.4590

Inference ID: Brandeis-11

Txt: Three patients declined further surgery.

Hyp: The patients did have some surgeries . (Yes)

The
DT		 patients
NNS		 did
VBD		 have
VB		 some
DT		 surgeries
NNS		 .
.
Three:CD 20.50 20.50 19.19 20.50 20.50 20.50 20.50
patients:NNS 20.00   0.00 14.67 13.05 20.00   3.28 20.00
declined:VBD 20.00 13.95   6.25   7.61 20.00 13.95 17.40
further:JJ 20.00 12.00 12.00 12.00 20.00 12.00 20.00
surgery:NN 20.00   2.43 13.05 13.95 20.00   0.50 19.89
.:. 10.00 20.00 17.99 20.00 10.00 20.00   0.00
NO_WORD   1.00 10.00   1.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -20.1142
Features matched: Adjunct.dropPosCxt: text adjunct "further" of "surgery" dropped on aligned hyp word "surgeries"; Antonym.samePol: matching polarity with antonyms: have & declined; NullPunisher.article: The; NullPunisher.aux: did; NullPunisher.other: some; RootEntailment.poorlyAlignedRoot: "have" aligned badly to "declined"
Hand-tuned score: -9.6500
Threshold: -11.4590

Inference ID: Brandeis-12

Txt: Foodstuffs are being blocked from entry into Iraq.

Hyp: Foodstuffs are getting into Iraq . (No)

Foodstuffs
NNS		 are
VBP		 getting
VBG		 Iraq
NNP		 .
.
Foodstuffs:NNS   0.00 15.00 15.00   9.84 20.00
are:VBP 15.00   0.00 10.00 15.50 20.00
being:VBG 13.95   0.50 10.00 14.84 20.00
blocked:VBN 13.03   6.09   7.38 12.61 19.67
entry:NN   8.03 13.69 12.38   9.84 20.00
Iraq:NNP   9.84 15.50 15.50   0.00 20.50
.:. 20.00 20.00 18.44 20.50   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -9.3754
Features matched: Adjunct.dropPosCxt: text adjunct "entry" of "blocked" dropped on aligned hyp word "getting"; RootEntailment.poorlyAlignedRoot: "getting" aligned badly to "blocked"; Structure.relMismatch: text "Foodstuffs" is nsubjpass of "blocked" while hyp "Foodstuffs" is nsubj of "getting" which aligned to text "blocked"
Hand-tuned score: -1.5000
Threshold: -11.4590

Inference ID: Brandeis-13

Txt: The new space allowed Compaq to increase the manufacturing capacity of its plant in Erskine, Scotland.

Hyp: Compaq did increase its manufacturing capacity in the Erskine plant . (Yes)

Compaq
NNP		 did
VBD		 increase
VB		 its
PRP$		 manufacturing
VBG		 capacity
NN		 the
DT		 Erskine
NNP		 plant
NN		 .
.
The:DT 20.50 20.00 20.00 20.00 20.00 20.00   0.00 20.50 20.00 10.00
new:JJ 12.46 11.96 11.96 15.00 11.96 11.96 20.00 12.46 11.91 20.00
space:NN 10.46   9.55 11.31 12.00 12.39   6.95 20.00 10.46   6.14 19.75
allowed:VBD 15.46   7.62   7.62 15.00 10.00 14.78 20.00 15.46 11.08 18.49
Compaq:NNP   0.00 15.46 15.46 12.50 15.46 10.46 20.50 14.96 10.46 20.50
to:TO 20.50 20.00 20.00 20.00 20.00 20.00 10.00 20.50 20.00 10.00
increase:VB 15.46   7.53   0.00 15.00   7.69 10.86 20.00 15.46 14.88 20.00
the:DT 20.50 20.00 20.00 20.00 20.00 20.00   0.00 20.50 20.00 10.00
manufacturing:VBG 15.46   7.32   7.69 15.00   0.00   9.91 20.00 15.46   8.96 19.20
capacity:NN 10.46 11.76 10.86 12.00   9.91   0.00 20.00 10.46   5.18 19.69
its:PRP$ 12.50 15.00 15.00   0.00 15.00 12.00 20.00 12.50 12.00 20.00
plant:NN 10.46 11.26 14.88 12.00   8.96   5.18 20.00 10.46   0.00 19.44
Erskine:NNP 14.96 15.46 15.46 12.50 15.46 10.46 20.50   0.00 10.46 20.50
,:, 20.50 19.80 19.86 20.00 19.93 20.00 10.00 20.50 19.64   5.73
Scotland:NNP 14.96 15.50 15.50 12.50 15.50 10.50 20.50   9.96   9.84 20.50
.:. 20.50 17.99 20.00 20.00 19.20 19.69 10.00 20.50 19.44   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00   1.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -10.0000
Features matched: Adjunct.dropPosCxt: text adjunct "Scotland" of "plant" dropped on aligned hyp word "plant"; NullPunisher.article: the; NullPunisher.aux: did; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "allowed vs. hyp "." <-punct-- "increase", which aligned to text "increase" args have different parents, different relations: text "plant" <-prep_of-- "capacity" vs. hyp "plant" <-prep_in-- "increase", which aligned to text "increase"
Hand-tuned score: -1.6500
Threshold: -11.4590

Inference ID: Brandeis-14

Txt: Holders of more than a majority of the stock of the company have approved the transaction by written consent.

Hyp: The transaction has taken_place . (Yes)

The
DT		 transaction
NN		 has
VBZ		 taken_place
VBN		 .
.
Holders:NNS 20.00 10.00 14.34 14.67 20.00
more_than:IN 20.00 20.00 20.00 20.00 20.00
a:DT 10.00 20.00 20.00 20.00 10.00
majority:NN 20.00   8.68 12.53 12.56 19.33
the:DT   0.00 20.00 20.00 20.00 10.00
stock:NN 20.00   6.69 13.69 12.22 19.80
the:DT   0.00 20.00 20.00 20.00 10.00
company:NN 20.00   5.29 14.34 13.72 18.58
have:VBP 20.00 15.00   0.50   5.92 20.00
approved:VBN 20.00 13.21 10.00   7.82 20.00
the:DT   0.00 20.00 20.00 20.00 10.00
transaction:NN 20.00   0.00 15.00 14.18 19.24
written:VBN 20.00 13.71 10.00   7.87 19.89
consent:NN 20.00   7.02 13.69 12.88 19.08
.:. 10.00 19.24 20.00 20.00   0.00
NO_WORD   1.00 10.00   1.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -10.3221
Features matched: Adjunct.dropPosCxt: text adjunct "consent" of "approved" dropped on aligned hyp word "taken_place"; RootEntailment.poorlyAlignedRoot: "taken_place" aligned badly to "approved"; Structure.relMismatch: text "transaction" is dobj of "approved" while hyp "transaction" is nsubj of "taken_place" which aligned to text "approved"
Hand-tuned score: -1.5000
Threshold: -11.4590

Inference ID: Brandeis-15

Txt: The transaction has been approved by Kyle's board but requires the approval of the company's shareholders.

Hyp: The transaction has taken_place . (No)

The
DT		 transaction
NN		 has
VBZ		 taken_place
VBN		 .
.
The:DT   0.00 20.00 20.00 20.00 10.00
transaction:NN 20.00   0.00 15.00 14.18 19.24
has:VBZ 20.00 15.00   0.00   8.76 20.00
been:VBN 20.00 15.00   9.34   7.17 20.00
approved:VBN 20.00 13.21 10.00   7.82 20.00
Kyle:NNP 20.50 10.46 15.46 15.46 20.50
board:NN 20.00   8.19 14.34 13.28 18.74
requires:VBZ 20.00 14.14 10.00   7.80 19.71
the:DT   0.00 20.00 20.00 20.00 10.00
approval:NN 20.00   7.49 13.69 13.72 19.52
the:DT   0.00 20.00 20.00 20.00 10.00
company:NN 20.00   5.29 14.34 13.72 18.58
shareholders:NNS 20.00   3.12 14.34 14.67 20.00
.:. 10.00 19.24 20.00 20.00   0.00
NO_WORD   1.00 10.00   1.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -9.8221
Features matched: RootEntailment.poorlyAlignedRoot: "taken_place" aligned badly to "approved"; Structure.relMismatch: text "transaction" is nsubjpass of "approved" while hyp "transaction" is nsubj of "taken_place" which aligned to text "approved"
Hand-tuned score: -2.0000
Threshold: -11.4590

Inference ID: Brandeis-16

Txt: The company's board authorized the purchase of an additional one million shares.

Hyp: The company has purchased additional shares . (MULTIPLE ANSWERS)

The
DT		 company
NN		 has
VBZ		 purchased
VBN		 additional
JJ		 shares
NNS		 .
.
The:DT   0.00 20.00 20.00 20.00 20.00 20.00 10.00
company:NN 20.00   0.00 14.34 13.24   7.63   7.80 18.58
board:NN 20.00   6.69 14.34 15.00   9.35   6.40 18.74
authorized:VBD 20.00 13.07 10.00   5.63   7.18 10.63 19.27
the:DT   0.00 20.00 20.00 20.00 20.00 20.00 10.00
purchase:NN 20.00   6.66 15.00   0.31   8.76   6.84 19.06
an:DT 10.00 20.00 20.00 20.00 20.00 20.00 10.00
additional:JJ 20.00   7.63 11.96   9.60   0.00   9.51 19.09
one:CD 20.50 20.50 19.19 20.50 20.46 19.19 20.50
million:CD 20.50 17.54 19.19 17.92 17.78 16.86 19.92
shares:NNS 20.00   7.80 13.69 11.84   9.51   0.00 19.44
.:. 10.00 18.58 20.00 19.15 19.09 19.44   0.00
NO_WORD   1.00 10.00   1.00 10.00   9.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -7.3094
Features matched: Adjunct.dropPosCxt: text adjunct "one" of "shares" dropped on aligned hyp word "shares"; NullPunisher.aux: has; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "authorized vs. hyp "." <-punct-- "purchased", which aligned to text "purchase" args have different parents, different relations: text "company" <-poss-- "board" vs. hyp "company" <-nsubj-- "purchased", which aligned to text "purchase" text "shares" is prep_of of "purchase" while hyp "shares" is dobj of "purchased" which aligned to text "purchase"
Hand-tuned score: -1.5500
Threshold: -11.4590

Inference ID: Brandeis-17

Txt: The purchase was subsequently rejected by regulators.

Hyp: The purchase did take_place . (No)

The
DT		 purchase
NN		 did
VBD		 take_place
JJ		 .
.
The:DT   0.00 20.00 20.00 20.00 10.00
purchase:NN 20.00   0.00 15.00   9.98 19.06
was:VBD 20.00 15.00 10.00 10.56 20.00
subsequently:RB 20.00 12.93 16.56 11.96 17.89
rejected:VBN 20.00 12.44 10.00 10.00 20.00
regulators:NNS 20.00   8.48 14.48 11.67 20.00
.:. 10.00 19.06 17.99 20.00   0.00
NO_WORD   1.00 10.00 10.00   9.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -22.0018
Features matched: Adjunct.dropPosCxt: text adjunct "subsequently" of "rejected" dropped on aligned hyp word "take_place"; NullPunisher.aux: did; RootEntailment.poorlyAlignedRoot: "take_place" aligned badly to "rejected"; Structure.relMismatch: text "purchase" is nsubjpass of "rejected" while hyp "purchase" is nsubj of "take_place" which aligned to text "rejected"
Hand-tuned score: -1.5500
Threshold: -11.4590

Inference ID: Brandeis-18

Txt: Analysts were predicting 1990 BellSouth earnings in the range of $3.90 a share.

Hyp: BellSouth has experienced earnings in 1990 . (Don't know)

BellSouth
NNP		 has
VBZ		 experienced
VBN		 earnings
NNS		 1990
CD		 .
.
Analysts:NNS 10.46 14.34 15.00 10.00 20.46 20.00
were:VBD 15.46   9.34   8.07 15.00 20.46 20.00
predicting:VBG 15.46 10.00   6.97 11.30 19.84 20.00
1990:CD 24.96 20.46 18.56 20.46   0.00 19.07
BellSouth:NNP   0.00 15.46 15.46 10.46 24.96 20.50
earnings:NNS 10.46 15.00 14.16   0.00 20.46 20.00
the:DT 20.50 20.00 20.00 20.00 20.50 10.00
range:NN 10.46 12.52 12.82   9.92 20.46 19.48
$:$ 25.00 20.50 20.50 20.50 24.16   9.91
3.90:CD 24.96 20.46 20.46 18.85 10.00 19.80
a:DT 20.50 20.00 20.00 20.00 20.50 10.00
share:NN 10.46 13.69 13.07   4.13 20.46 20.00
.:. 20.50 20.00 19.35 20.00 19.07   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -11.9661
Features matched: Adjunct.dropPosCxt: text adjunct "range" of "predicting" dropped on aligned hyp word "experienced"; Date.matchDatesByNormForm: hyp/txt matching, by normalized form: 01/01/1990; NullPunisher.aux: has; RootEntailment.poorlyAlignedRoot: "experienced" aligned badly to "predicting"; Structure.parentsMismatch: args have different parents, different relations: text "BellSouth" <-nn-- "earnings" vs. hyp "BellSouth" <-nsubj-- "experienced", which aligned to text "predicting"
Hand-tuned score: -2.5500
Threshold: -11.4590

Inference ID: Brandeis-19.1

Txt: Milton Roy disclosed in May that it was approached for a possible acquisition by Thermo Electron.

Hyp: Milton Roy was approached by Thermo Electron . (Yes)

Milton_Roy
NNP		 was
VBD		 approached
VBN		 Thermo_Electron
NNP		 .
.
Milton_Roy:NNP   0.00 15.17 14.97 14.47 20.50
disclosed:VBD 15.46 10.00   6.50 15.46 19.87
May:NNP 14.05 14.84 14.19 14.47 20.50
that:IN 20.50 20.00 20.00 20.50 20.00
it:PRP 12.50 15.00 15.00 12.50 20.00
was:VBD 15.17   0.00   7.52 15.17 20.00
approached:VBN 14.97   7.52   0.00 14.97 18.22
a:DT 20.50 20.00 20.00 20.50 10.00
possible:JJ 11.52 11.34 10.95 11.97 17.78
acquisition:NN 10.46 15.00 12.38 10.46 20.00
Thermo_Electron:NNP 14.47 15.17 14.97   0.00 20.50
.:. 20.50 20.00 18.22 20.50   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -4.0000
Features matched: Adjunct.dropPosCxt: text adjunct "acquisition" of "approached" dropped on aligned hyp word "approached"; Factive.inPositiveEmbedding: embedded positive text; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "disclosed vs. hyp "." <-punct-- "approached", which aligned to text "approached" args have different parents, different relations: text "Milton_Roy" <-nsubj-- "disclosed" vs. hyp "Milton_Roy" <-nsubjpass-- "approached", which aligned to text "approached"
Hand-tuned score: -0.5000
Threshold: -11.4590

Inference ID: Brandeis-19.2

Txt: Milton Roy disclosed in May that it was approached for a possible acquisition by Thermo Electron.

Hyp: Milton Roy has been acquired by Thermo Electron . (Don't know)

Milton_Roy
NNP		 has
VBZ		 been
VBN		 acquired
VBN		 Thermo_Electron
NNP		 .
.
Milton_Roy:NNP   0.00 15.17 15.17 15.46 14.47 20.50
disclosed:VBD 15.46 10.00   7.74   2.60 15.46 19.87
May:NNP 14.05 14.19 14.84 15.50 14.47 20.50
that:IN 20.50 20.00 20.00 20.00 20.50 20.00
it:PRP 12.50 15.00 15.00 15.00 12.50 20.00
was:VBD 15.17   7.52   0.50 10.00 15.17 20.00
approached:VBN 14.97   7.52   6.07   6.88 14.97 18.22
a:DT 20.50 20.00 20.00 20.00 20.50 10.00
possible:JJ 11.52 11.34 11.34 10.56 11.97 17.78
acquisition:NN 10.46 15.00 15.00   2.50 10.46 20.00
Thermo_Electron:NNP 14.47 15.17 15.17 15.46   0.00 20.50
.:. 20.50 20.00 20.00 20.00 20.50   0.00
NO_WORD 10.00   1.00   1.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -8.6049
Features matched: Adjunct.dropPosCxt: text adjunct "May" of "disclosed" dropped on aligned hyp word "acquired"; NullPunisher.aux: has; NullPunisher.aux: been; Structure.argsMismatch: passive struct without correct agent (Milton_Roy vs. Thermo_Electron); Structure.argsMismatch: args have different parents but same relations: text "Thermo_Electron" <-agent-- "approached vs. hyp "Thermo_Electron" <-agent-- "acquired", which aligned to text "disclosed" text "Milton_Roy" is nsubj of "disclosed" while hyp "Milton_Roy" is nsubjpass of "acquired" which aligned to text "disclosed"
Hand-tuned score: -4.6000
Threshold: -11.4590

Inference ID: Brandeis-20

Txt: Analysts noted that over the past 20 years, Mr. Fournier has built his company through astute stock-market activity.

Hyp: Mr. Fournier has built his company through astute stock-market activity . (MULTIPLE ANSWERS)

Mr._Fournier
NNP		 has
VBZ		 built
VBN		 his
PRP$		 company
NN		 astute
JJ		 stock-market
NN		 activity
NN		 .
.
Analysts:NNS 10.46 14.34 13.95 12.00   8.05 11.96   9.96 10.00 20.00
noted:VBD 15.46   7.53   7.53 15.00 12.44 10.73 14.36 12.44 18.67
that:IN 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00
the:DT 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 10.00
past:JJ 12.46 10.69 10.67 15.00 10.13   5.00 11.25 10.69 19.32
20:CD 24.96 19.19 18.98 20.50 19.06 20.46 20.42 19.19 19.23
years:NNS 10.46 13.69 12.10 12.00 10.00 11.36   9.57   8.69 19.49
,:, 20.50 20.00 19.36 20.00 18.72 20.00 19.65 20.00   5.73
Mr._Fournier:NNP   0.00 15.46 15.46 12.50 10.46 12.46 10.46 10.46 20.50
has:VBZ 15.46   0.00   7.53 15.00 14.34 11.96 14.96 12.53 20.00
built:VBN 15.46   7.53   0.00 15.00 13.07 10.31 14.63 12.53 19.53
his:PRP$ 12.50 15.00 15.00   0.00 12.00 15.00 12.00 12.00 20.00
company:NN 10.46 14.34 13.07 12.00   0.00 11.96   8.00   7.44 18.58
astute:JJ 12.46 11.96 10.31 15.00 11.96   0.00 11.09 11.96 19.87
stock-market:NN 10.46 14.96 14.63 12.00   8.00 11.09   0.00   7.59 19.68
activity:NN 10.46 12.53 12.53 12.00   7.44 11.96   7.59   0.00 19.47
.:. 20.50 20.00 19.53 20.00 18.58 19.87 19.68 19.47   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00   9.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -2.0000
Features matched: Adjunct.dropPosCxt: text adjunct "years" of "built" dropped on aligned hyp word "built"; Factive.inPositiveEmbedding: embedded positive text; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "noted vs. hyp "." <-punct-- "built", which aligned to text "built"
Hand-tuned score: -0.5000
Threshold: -11.4590

Inference ID: Brandeis-21.1

Txt: I was pleased that Ms. Currie's lawyers stated unambiguously this morning that she's not aware of any unethical conduct.

Hyp: Ms. Currie is aware of some unethical conduct on her part . (MULTIPLE ANSWERS)

Ms._Currie
NNP		 is
VBZ		 aware
JJ		 some
DT		 unethical
JJ		 conduct
NN		 her
PRP$		 part
NN		 .
.
I:PRP 12.50 15.00 15.00 20.00 15.00 12.00 10.00 12.00 20.00
was:VBD 15.17   0.50 11.96 20.00 11.96 15.00 15.00 12.52 20.00
pleased:JJ 12.46 12.00   4.23 20.00   8.36   9.90 15.00 10.59 17.64
that:IN 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00
Ms._Currie:NNP   0.00 15.17 12.46 20.50 12.46 10.46 12.50   9.97 20.50
lawyers:NNS   9.52 14.34   9.42 20.00   8.89   5.04 12.00   8.95 20.00
stated:VBD 15.17   5.51   9.10 20.00 10.25 10.50 15.00 12.44 19.01
unambiguously:RB 15.46 19.96 10.40 20.00   8.37 13.67 15.00 14.96 20.00
this:DT 20.50 20.00 20.00   8.49 20.00 20.00 20.00 20.00 10.00
morning:NN 10.46 15.00 11.33 20.00 11.96   8.69 12.00   7.82 17.77
that:IN 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00
she:PRP 12.50 15.00 15.00 20.00 15.00 12.00   0.50 12.00 20.00
's:VBZ 15.17   0.50 10.72 20.00 11.09 12.53 15.00 12.01 18.25
not:RB 15.46 19.96 11.96 20.00 11.96 14.96 15.00 14.96 20.00
aware:JJ 12.46 11.96   0.00 20.00   5.22   7.11 15.00 10.23 17.23
any:DT 20.50 20.00 20.00   8.51 20.00 20.00 20.00 20.00 10.00
unethical:JJ 12.46 11.96   5.22 20.00   0.00   6.22 15.00 11.83 19.87
conduct:NN 10.46 13.07   7.11 20.00   6.22   0.00 12.00   7.65 19.33
.:. 20.50 20.00 17.23 10.00 19.87 19.33 20.00 17.01   0.00
NO_WORD 10.00   1.00   9.00 10.00   9.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -25.1590
Features matched: Adjunct.diffPol: hyp and txt have different polarity; Factive.inPositiveEmbedding: embedded positive text; Polarity.txtNegMarker: "aware": neg; Quant.contract: [any,some]; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "pleased vs. hyp "." <-punct-- "aware", which aligned to text "aware" args have different parents, different relations: text "Ms._Currie" <-poss-- "lawyers" vs. hyp "Ms._Currie" <-nsubj-- "aware", which aligned to text "aware"
Hand-tuned score: -5.0000
Threshold: -11.4590

Inference ID: Brandeis-21.2

Txt: I was pleased that Ms. Currie's lawyers stated unambiguously this morning that she's not aware of any unethical conduct.

Hyp: Ms. Currie have s lawyers stated she s not aware of some unethical conduct . (Yes)

Ms._Currie
NNP		 have
VBP		 s
VBN		 lawyers
NNS		 stated
VBD		 she
PRP		 s
VBZ		 not
RB		 aware
JJ		 some
DT		 unethical
JJ		 conduct
NN		 .
.
I:PRP 12.50 15.00 15.00 12.00 15.00 10.00 15.00 20.00 15.00 20.00 15.00 12.00 20.00
was:VBD 15.17   9.34   9.34 14.34   5.13 15.00   9.34 19.96 11.96 20.00 11.96 15.00 20.00
pleased:JJ 12.46 12.00 12.00   9.72   9.90 15.00 12.00 11.96   4.23 20.00   8.36   9.90 17.64
that:IN 20.50 20.00 20.00 20.00 20.00 18.32 20.00 20.00 20.00 20.00 20.00 20.00 20.00
Ms._Currie:NNP   0.00 14.52 15.17   9.52 15.17 12.50 15.17 15.46 12.46 20.50 12.46 10.46 20.50
lawyers:NNS   9.52 13.05 13.67   0.00 10.93 12.00 13.67 14.96   9.42 20.00   8.89   5.04 20.00
stated:VBD 15.17   7.80   9.34 10.93   0.00 15.00   9.34 19.96   9.10 20.00 10.25 10.50 19.01
unambiguously:RB 15.46 19.96 19.91 14.96 18.54 15.00 19.91   9.96 10.40 20.00   8.37 13.67 20.00
this:DT 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00   8.49 20.00 20.00 10.00
morning:NN 10.46 15.00 12.84   9.93 12.90 12.00 12.84 14.96 11.33 20.00 11.96   8.69 17.77
that:IN 20.50 20.00 20.00 20.00 20.00 18.32 20.00 20.00 20.00 20.00 20.00 20.00 20.00
she:PRP 12.50 15.00 15.00 12.00 15.00   0.00 15.00 20.00 15.00 20.00 15.00 12.00 20.00
's:VBZ 15.17   9.34   0.00 14.34   8.60 15.00   0.00 19.96 10.72 20.00 11.09 12.53 18.25
not:RB 15.46 19.96 19.96 14.96 19.96 15.00 19.96   0.00 11.96 20.00 11.96 14.96 20.00
aware:JJ 12.46 11.96 11.96   9.42   9.10 15.00 11.96 11.96   0.00 20.00   5.22   7.11 17.23
any:DT 20.50 20.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00   8.51 20.00 20.00 10.00
unethical:JJ 12.46 11.96 11.29   8.89 10.25 15.00 11.29 11.96   5.22 20.00   0.00   6.22 19.87
conduct:NN 10.46 11.73 12.53   5.04 10.50 12.00 12.53 14.96   7.11 20.00   6.22   0.00 19.33
.:. 20.50 20.00 19.43 20.00 19.01 20.00 19.43 20.00 17.23 10.00 19.87 19.33   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00 10.00   9.00   9.00 10.00   9.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -17.5100
Features matched: Adjunct.dropPosCxt: text adjunct "morning" of "stated" dropped on aligned hyp word "stated"; Polarity.hypNegMarker: "aware": neg; Polarity.txtNegMarker: "aware": neg; NullPunisher.aux: have; Quant.contract: [any,some]; Structure.argsMismatch: args have different parents but same relations: text "stated" <-ccomp-- "pleased vs. hyp "stated" <-ccomp-- "s", which aligned to text "'s" args have different parents but same relations: text "." <-punct-- "pleased vs. hyp "." <-punct-- "s", which aligned to text "'s" args have different parents, different relations: text "Ms._Currie" <-poss-- "lawyers" vs. hyp "Ms._Currie" <-nsubj-- "s", which aligned to text "'s" ; Polarity.txtNegMarker&PolarityhypNegMarker:
Hand-tuned score: -0.5500
Threshold: -11.4590

Inference ID: Brandeis-22

Txt: President Clinton and Blair will stand together on arresting the terrorists suspected of blowing up Pan Am flight 103 over Scotland.

Hyp: There has been an explosion of a Pan Am flight over Scotland . (Yes)

There
EX		 has
VBZ		 been
VBN		 an
DT		 explosion
NN		 a
DT		 Pan
NNP		 Am
NNP		 flight
NN		 Scotland
NNP		 .
.
President_Clinton:NNP 20.00 13.76 14.34 20.00   9.18 20.00   9.45   9.68   8.95   8.67 20.00
Blair:NNP 20.50 14.84 14.84 20.50 10.50 20.50 13.95 14.34   9.45 14.34 20.50
will:MD 10.00 18.69 20.00 10.00 17.85 10.00 17.95 18.35 15.96 20.50 10.00
stand:VB 20.00   7.52   5.04 20.00 11.89 20.00 12.62 13.53 12.82 13.02 18.14
together:RB 20.00 19.96 19.96 20.00 14.96 20.00 15.46 15.46 14.96 15.46 20.00
on:IN 20.00 20.00 20.00 18.21 20.00 20.00 20.50 20.50 20.00 20.50 20.00
arresting:VBG 20.00 10.00 10.00 20.00 10.61 20.00 15.50 13.85 10.88 15.50 19.07
the:DT 10.00 20.00 20.00 10.00 20.00 10.00 20.50 20.50 20.00 20.50 10.00
terrorists:NNS 20.00 14.34 14.34 20.00   6.95 20.00   9.45   9.84   7.32   9.84 20.00
suspected:VBN 20.00   9.34   9.34 20.00 10.31 20.00 14.45 14.84 12.96 14.84 19.80
of:IN 20.00 20.00 20.00 18.57 20.00 20.00 20.50 20.50 20.00 20.50 20.00
blowing:VBG 20.00 10.00 10.00 20.00 10.90 20.00 15.50 15.50 11.94 15.50 19.78
Pan:NNP 20.50 14.84 14.84 18.50 10.50 20.50   0.00   9.34   4.66 14.34 20.50
Am:VBP 20.50   8.03   4.89 20.50 13.35 20.50 14.34   0.00 13.85 19.34 20.50
flight:NN 20.00 14.34 14.34 20.00   4.88 20.00   4.66   8.85   0.00   9.84 20.00
103:CD 20.50 20.46 20.46 20.50 20.01 20.50 24.96 24.96 19.83 24.96 18.11
Scotland:NNP 20.50 13.02 14.84 20.50 10.50 20.50 14.34 14.34   9.84   0.00 20.50
.:. 10.00 20.00 20.00 10.00 19.32 10.00 20.50 20.50 20.00 20.50   0.00
NO_WORD   1.00   1.00 10.00   1.00 10.00   1.00 10.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -23.9123
Features matched: Adjunct.dropPosCxt: text adjunct "103" of "flight" dropped on aligned hyp word "flight"; NullPunisher.article: a; NullPunisher.article: an; NullPunisher.functionWord: There; NullPunisher.aux: has; RootEntailment.poorlyAlignedRoot: "been" aligned badly to "stand"; Structure.argsMismatch: args have different parents but same relations: text "Scotland" <-prep_over-- "arresting vs. hyp "Scotland" <-prep_over-- "been", which aligned to text "stand" args have different parents, different relations: text "flight" <-prep_up-- "blowing" vs. hyp "explosion" <-dobj-- "been", which aligned to text "stand"
Hand-tuned score: -3.8500
Threshold: -11.4590

Inference ID: Brandeis-23

Txt: The accord helps RJR pay off debt.

Hyp: RJR does have some debt . (Yes)

RJR
NNP		 does
VBZ		 have
VB		 some
DT		 debt
NN		 .
.
The:DT 20.50 20.00 20.00 10.00 20.00 10.00
accord:NN 10.46 14.93 15.00 20.00   3.20 19.88
helps:VBZ 15.46   8.11   7.32 20.00 12.48 18.97
RJR:NNP   0.00 15.46 15.46 20.50 10.46 20.50
pay_off:VB 15.46   9.96   6.84 20.00 13.48 20.00
debt:NN 10.46 15.00 15.00 20.00   0.00 19.04
.:. 20.50 20.00 20.00 10.00 19.04   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -19.8406
Features matched: Factive.inPositiveEmbedding: embedded positive text; NullPunisher.aux: does; NullPunisher.other: some; RootEntailment.poorlyAlignedRoot: "have" aligned badly to "pay_off"; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "helps vs. hyp "." <-punct-- "have", which aligned to text "pay_off"
Hand-tuned score: -4.0500
Threshold: -11.4590

Inference ID: Brandeis-24

Txt: 48 Kuwaiti jet fighters managed to escape the Iraqi invasion.

Hyp: Some Kuwaiti jet fighters did escape the Iraqi invasion . (Yes)

Some
DT		 Kuwaiti
NNP		 jet
NN		 fighters
NNS		 did
VBD		 escape
VB		 the
DT		 Iraqi
JJ		 invasion
NN		 .
.
48:CD 20.50 24.96 20.46 20.46 20.15 20.46 20.50 24.96 19.56 18.60
Kuwaiti:NNP 20.50   0.00   8.58   8.55 15.50 13.61 20.50   4.08 10.50 20.50
jet:NN 20.00   8.58   0.00   3.27 12.69   6.38 20.00 10.58   3.94 20.00
fighters:NNS 20.00   8.55   3.27   0.00 14.77 11.39 20.00 10.55   5.48 20.00
managed:VBD 20.00 15.50 15.00 15.00   1.52   7.91 20.00 12.50 15.00 20.00
to:TO 10.00 20.50 20.00 20.00 20.00 20.00 10.00 20.50 20.00 10.00
escape:VB 20.00 13.61   6.38 11.39   7.69   0.00 20.00 10.61   8.94 18.79
the:DT 10.00 20.50 20.00 20.00 20.00 20.00   0.00 20.50 20.00 10.00
Iraqi:JJ 20.50   4.08 10.58 10.55 12.50 10.61 20.50   0.00 12.50 20.50
invasion:NN 20.00 10.50   3.94   5.48 12.69   8.94 20.00 12.50   0.00 20.00
.:. 10.00 20.50 20.00 20.00 17.99 18.79 10.00 20.50 20.00   0.00
NO_WORD 10.00 10.00 10.00 10.00   1.00 10.00   1.00   9.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -15.0000
Features matched: Adjunct.dropPosCxt: text adjunct "48" of "fighters" dropped on aligned hyp word "fighters"; NullPunisher.aux: did; NullPunisher.other: Some; Structure.argsMismatch: args have different parents but same relations: text "fighters" <-nsubj-- "managed vs. hyp "fighters" <-nsubj-- "escape", which aligned to text "escape" args have different parents but same relations: text "." <-punct-- "managed vs. hyp "." <-punct-- "escape", which aligned to text "escape"
Hand-tuned score: -2.5500
Threshold: -11.4590

Inference ID: Brandeis-25

Txt: It appears that allied troops haven't yet fully engaged Iraq's Republican Guard.

Hyp: Allied troops have engaged Iraq s Republican_Guard . (Don't know)

Allied
NNP		 troops
NNS		 have
VBP		 engaged
VBN		 Iraq
NNP		 s
VBZ		 Republican_Guard
JJ		 .
.
It:PRP 12.00 12.00 15.00 15.00 12.50 15.00 15.00 20.00
appears:VBZ 15.00 15.00   8.33   7.48 15.50 10.00 12.00 20.00
that:IN 20.00 20.00 20.00 20.00 20.50 20.00 20.00 20.00
allied:JJ   2.00   5.84   9.80   9.35 11.84 11.34   6.69 20.00
troops:NNS   7.59   0.00 15.00 14.06 10.50 14.82   9.59 20.00
have:VBP 12.80 15.00   0.00   6.84 14.84   9.34 11.02 20.00
n't:RB 14.96 14.59 19.96 19.96 15.46 19.56 11.96 17.90
yet:RB 14.96 14.96 19.96 19.96 15.46 19.96 11.96 20.00
fully:RB 14.96 13.64 19.96 19.47 15.46 19.96 11.96 18.33
engaged:VBN 13.07 14.06   6.84   0.00 15.50 10.00 11.03 18.98
Iraq:NNP   9.84 10.50 14.84 15.50   0.00 14.84 12.17 20.50
Republican_Guard:NN   7.19   8.09 14.52 14.53 14.67 13.98   0.50 20.50
.:. 20.00 20.00 20.00 18.98 20.50 19.43 20.00   0.00
NO_WORD 10.00 10.00   1.00 10.00 10.00 10.00   9.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -20.5000
Features matched: Adjunct.diffPol: hyp and txt have different polarity; Factive.unknownPassage: non factive text -- unknown: appears-VBZ; Polarity.txtNegMarker: "engaged": neg; NullPunisher.other: s; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "appears vs. hyp "." <-punct-- "engaged", which aligned to text "engaged" text "Republican_Guard" is dobj of "engaged" while hyp "Republican_Guard" is ccomp of "engaged" which aligned to text "engaged"
Hand-tuned score: -7.5000
Threshold: -11.4590

Inference ID: Brandeis-26.1

Txt: British police officers concluded Howes had probably been killed soon after being captured.

Hyp: Howes did die . (Yes)

Howes
NNS		 did
VBD		 die
NN		 .
.
British:JJ 17.00 12.50 12.50 20.50
police_officers:NNS   8.55 15.00   8.95 20.00
concluded:VBD 15.50   6.25 12.62 19.24
Howes:NNP   0.00 15.50   9.45 20.50
had:VBD 13.55   7.32 12.61 20.00
probably:RB 15.46 17.95 12.44 18.53
been:VBN 14.84   6.07 13.07 20.00
killed:VBN 15.50   6.25   9.20 20.00
soon:RB 15.46 18.33 14.18 18.12
after:IN 20.50 20.00 20.00 20.00
being:VBG 13.61 10.00 13.95 20.00
captured:VBN 15.50   7.32 12.62 20.00
.:. 20.50 17.99 19.41   0.00
NO_WORD 10.00   1.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -13.9455
Features matched: Adjunct.dropPosCxt: text adjunct "British" of "police_officers" dropped on aligned hyp word "die"; Factive.unknownPassage: non factive text -- unknown: concluded-VBD; NullPunisher.aux: did; RootEntailment.poorlyAlignedRoot: "die" aligned badly to "police_officers"; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "concluded vs. hyp "." <-punct-- "die", which aligned to text "police_officers" args have different parents, different relations: text "Howes" <-nsubjpass-- "killed" vs. hyp "Howes" <-nsubj-- "die", which aligned to text "police_officers"
Hand-tuned score: -3.0500
Threshold: -11.4590

Inference ID: Brandeis-26.2

Txt: British police officers concluded Howes had probably been killed soon after being captured.

Hyp: Howes was killed soon after being captured . (Yes)

Howes
NNS		 was
VBD		 killed
VBN		 soon
RB		 after
IN		 being
VBG		 captured
VBN		 .
.
British:JJ 17.00 12.50 12.50 12.46 20.50 12.50 12.50 20.50
police_officers:NNS   8.55 14.34 15.00 14.96 20.00 13.11 15.00 20.00
concluded:VBD 15.50 10.00   6.25 18.32 20.00 10.00   7.62 19.24
Howes:NNP   0.00 14.84 15.50 15.46 20.50 13.61 15.50 20.50
had:VBD 13.55   9.34   6.55 19.96 20.00   8.11   6.84 20.00
probably:RB 15.46 19.96 19.96   7.43 20.00 19.96 19.85 18.53
been:VBN 14.84   0.50   8.33 19.96 20.00   0.50 10.00 20.00
killed:VBN 15.50 10.00   0.00 19.96 20.00 10.00   4.75 20.00
soon:RB 15.46 19.96 19.96   0.00 20.00 19.96 19.96 18.12
after:IN 20.50 20.00 20.00 20.00   0.00 20.00 20.00 20.00
being:VBG 13.61   0.50 10.00 19.96 20.00   0.00 10.00 20.00
captured:VBN 15.50 10.00   4.75 19.96 20.00 10.00   0.00 20.00
.:. 20.50 20.00 20.00 18.12 20.00 20.00 20.00   0.00
NO_WORD 10.00   1.00 10.00   9.00 10.00   1.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -2.5000
Features matched: Adjunct.dropPosCxt: text adjunct "probably" of "killed" dropped on aligned hyp word "killed"; Factive.unknownPassage: non factive text -- unknown: concluded-VBD; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "concluded vs. hyp "." <-punct-- "killed", which aligned to text "killed"
Hand-tuned score: -1.0000
Threshold: -11.4590

Inference ID: Brandeis-27

Txt: The police allowed the crowd to demonstrate on nearby streets.

Hyp: The crowd did demonstrate on nearby streets . (Yes)

The
DT		 crowd
NN		 did
VBD		 demonstrate
VB		 nearby
JJ		 streets
NNS		 .
.
The:DT   0.00 20.00 20.00 20.00 20.00 20.00 10.00
police:NN 20.00   7.81 14.29 15.00   8.92   4.46 20.00
allowed:VBD 20.00 13.07   7.62   7.27 11.15 15.00 18.49
the:DT   0.00 20.00 20.00 20.00 20.00 20.00 10.00
crowd:NN 20.00   0.00 12.45 12.80   8.44   5.86 19.27
to:TO 10.00 20.00 20.00 20.00 20.00 20.00 10.00
demonstrate:VB 20.00 12.80 10.00   0.00 11.96 13.47 19.33
nearby:JJ 20.00   8.44 11.52 11.96   0.00   7.19 18.62
streets:NNS 20.00   5.86 15.00 13.47   7.19   0.00 19.83
.:. 10.00 19.27 17.99 19.33 18.62 19.83   0.00
NO_WORD   1.00 10.00   1.00 10.00   9.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -5.0000
Features matched: Factive.inPositiveEmbedding: embedded positive text; NullPunisher.aux: did; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "allowed vs. hyp "." <-punct-- "demonstrate", which aligned to text "demonstrate" args have different parents, different relations: text "crowd" <-dobj-- "allowed" vs. hyp "crowd" <-nsubj-- "demonstrate", which aligned to text "demonstrate"
Hand-tuned score: -1.0500
Threshold: -11.4590

Inference ID: Brandeis-28.1

Txt: In Pakistan, the Taliban have forbidden women to work.

Hyp: The Taliban do want women to work in Pakistan . (MULTIPLE ANSWERS)

The
DT		 Taliban
NNP		 do
VBP		 want
VB		 women
NNS		 to
TO		 work
VB		 Pakistan
NNP		 .
.
Pakistan:NNP 20.50 13.65 15.50 14.84   9.84 20.50 13.02   0.00 20.50
,:, 10.00 20.50 19.52 19.41 18.95 10.00 19.61 20.50   5.73
the:DT   0.00 20.50 20.00 20.00 20.00 10.00 20.00 20.50 10.00
Taliban:NNP 20.50   0.00 15.50 15.50   8.63 20.50 15.50 13.65 20.50
have:VBP 20.00 15.50   6.02   6.13 13.05 20.00   6.51 14.84 20.00
forbidden:VBN 20.00 15.50   8.40   6.68 13.27 20.00   7.65 15.50 19.95
women:NNS 20.00   8.63 13.95 13.93   0.00 18.72 13.47   9.84 19.64
to:TO 10.00 20.50 20.00 20.00 18.72   0.00 20.00 20.50 10.00
work:VB 20.00 15.50   6.33   6.90 13.47 20.00   0.00 13.02 18.57
.:. 10.00 20.50 18.81 17.90 19.64 10.00 18.57 20.50   0.00
NO_WORD   1.00 10.00   1.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -11.6816
Features matched: NullPunisher.aux: do; RootEntailment.poorlyAlignedRoot: "want" aligned badly to "forbidden"
Hand-tuned score: -1.0500
Threshold: -11.4590

Inference ID: Brandeis-28.2

Txt: In Pakistan, the Taliban have forbidden women to work.

Hyp: Women do work in Pakistan . (No)

Women
NNP		 do
VBP		 work
NN		 Pakistan
NNP		 .
.
Pakistan:NNP   9.84 15.50   8.02   0.00 20.50
,:, 20.00 19.52 19.61 20.50   5.73
the:DT 20.00 20.00 20.00 20.50 10.00
Taliban:NNP   8.63 15.50 10.50 13.65 20.50
have:VBP 13.05   6.02 11.51 14.84 20.00
forbidden:VBN 15.00   8.40 12.65 15.50 19.95
women:NNS   0.00 13.95   8.47   9.84 19.64
to:TO 20.00 20.00 20.00 20.50 10.00
work:VB 13.95   6.33   0.00 13.02 18.57
.:. 20.00 18.81 18.57 20.50   0.00
NO_WORD 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -14.0228
Features matched: RootEntailment.poorlyAlignedRoot: "do" aligned badly to "have"; Structure.argsMismatch: args have different parents but same relations: text "." <-punct-- "forbidden vs. hyp "." <-punct-- "do", which aligned to text "have" args have different parents, different relations: text "women" <-dobj-- "forbidden" vs. hyp "Women" <-nsubj-- "do", which aligned to text "have" args have different parents, different relations: text "work" <-xcomp-- "forbidden" vs. hyp "work" <-dobj-- "do", which aligned to text "have"
Hand-tuned score: -4.0000
Threshold: -11.4590

Inference ID: Brandeis-29

Txt: The Administration is trying to decide whether Saddam Hussein has WMD.

Hyp: The administration is deciding whether Saddam_Hussein has WMD . (Yes)

The
DT		 administration
NN		 is
VBZ		 deciding
VBG		 whether
IN		 Saddam_Hussein
NNP		 has
VBZ		 WMD
NNP		 .
.
The:DT   0.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 10.00
Administration:NNP 20.50   0.50 15.50 15.50 20.50 10.50 14.19 10.50 20.50
is:VBZ 20.00 15.00   0.00 10.00 20.00 14.34   8.64 14.34 20.00
trying:VBG 20.00 12.72   8.07   7.29 20.00 15.00 10.00 15.00 18.06
to:TO 10.00 20.00 20.00 20.00 20.00 20.00 20.00 20.00 10.00
decide:VB 20.00 15.00 10.00   0.50 20.00 15.00 10.00 15.00 19.73
whether:IN 20.00 20.00 20.00 20.00   0.00 20.00 20.00 20.00 20.00
Saddam_Hussein:NNP 20.50   9.76 14.84 15.50 20.50   0.50 14.84   9.45 20.50
has:VBZ 20.00 13.69   8.64 10.00 20.00 14.34   0.00 14.34 20.00
WMD:NNP 20.00 10.00 14.34 15.00 20.00   8.95 14.34   0.00 20.00
.:. 10.00 19.88 20.00 18.94 20.00 20.00 20.00 20.00   0.00
NO_WORD   1.00 10.00   1.00 10.00   1.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -6.5000
Features matched: NullPunisher.aux: is; Structure.argsMismatch: args have different parents but same relations: text "Administration" <-nsubj-- "trying vs. hyp "administration" <-nsubj-- "deciding", which aligned to text "decide" args have different parents but same relations: text "." <-punct-- "trying vs. hyp "." <-punct-- "deciding", which aligned to text "decide"
Hand-tuned score: -2.0500
Threshold: -11.4590

Inference ID: Brandeis-30

Txt: Helicopters are trying to locate people stranded without food.

Hyp: Helicopters are locating people stranded without food . (Don't know)

Helicopters
NNPS		 are
VBP		 locating
VBG		 people
NNS		 stranded
VBN		 food
NN		 .
.
Helicopters:NNPS   0.00 15.00 15.00 10.00 13.03   9.34 20.00
are:VBP 15.00   0.00 10.00 15.00 10.00 15.00 20.00
trying:VBG 15.00 10.00   7.13 10.45   9.31 13.57 18.06
to:TO 20.00 20.00 20.00 20.00 20.00 20.00 10.00
locate:VB 15.00 10.00   0.50 12.61   8.15 13.80 19.11
people:NNS 10.00 15.00 13.67   0.00 13.38   9.96 17.46
stranded:VBN 13.03 10.00   9.93 13.38   0.00 12.34 20.00
food:NN   9.34 15.00 13.37   9.96 12.34   0.00 19.20
.:. 20.00 20.00 19.61 17.46 20.00 19.20   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -5.5000
Features matched: NullPunisher.aux: are; Structure.argsMismatch: args have different parents but same relations: text "Helicopters" <-nsubj-- "trying vs. hyp "Helicopters" <-nsubj-- "locating", which aligned to text "locate" args have different parents but same relations: text "people" <-nsubjpass-- "stranded vs. hyp "people" <-dobj-- "locating", which aligned to text "locate" args have different parents but same relations: text "." <-punct-- "trying vs. hyp "." <-punct-- "locating", which aligned to text "locate"
Hand-tuned score: -2.0500
Threshold: -11.4590

Inference ID: Brandeis-31

Txt: MCA agreed to wait on purchasing Cineplex.

Hyp: MCA is waiting to purchase Cineplex . (Yes)

MCA
NNP		 is
VBZ		 waiting
VBG		 to
TO		 purchase
VB		 Cineplex
NNP		 .
.
MCA:NNP   0.00 15.46 15.46 20.50 15.46   9.96 20.50
agreed:VBD 15.46   6.07   9.93 20.00   5.25 15.46 20.00
to:TO 20.50 20.00 20.00   0.00 20.00 20.50 10.00
wait:VB 15.46   8.07   0.50 20.00   8.35 15.46 19.70
purchasing:VBG 15.46 10.00   8.35 20.00   0.50 15.46 19.72
Cineplex:NNP   9.96 15.46 15.46 20.50 15.46   0.00 20.50
.:. 20.50 20.00 18.98 10.00 19.06 20.50   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00 10.00

Response: yes (CORRECT)
Justification:
Alignment score: -14.0000
Features matched: NullPunisher.aux: is; Structure.argsMismatch: args have different parents but same relations: text "MCA" <-nsubj-- "agreed vs. hyp "MCA" <-nsubj-- "waiting", which aligned to text "wait" args have different parents but same relations: text "." <-punct-- "agreed vs. hyp "." <-punct-- "waiting", which aligned to text "wait" args have different parents, different relations: text "purchasing" <-amod-- "Cineplex" vs. hyp "purchase" <-xcomp-- "waiting", which aligned to text "wait"
Hand-tuned score: -2.0500
Threshold: -11.4590

Inference ID: Brandeis-32

Txt: MGM Grand Inc. has agreed to pay $93 million to buy 117 acres of land in Las Vegas.

Hyp: MGM Grand Inc. has paid $ 93 million to buy 117 acres of land in Las_Vegas . (Don't know)

MGM_Grand_Inc.
NNP		 has
VBZ		 paid
VBN		 $
$		 93
CD		 million
CD		 to
TO		 buy
VB		 117
CD		 acres
NNS		 land
NN		 Las_Vegas
NNP		 .
.
MGM_Grand_Inc.:NNP   0.00 14.84 15.46 25.00 24.96 22.15 20.50 15.46 24.96 10.46   9.97   9.84 20.50
has:VBZ 14.84   0.00 10.00 20.50 20.46 19.19 20.00 10.00 20.46 15.00 12.52 12.52 20.00
agreed:VBN 15.46 10.00   5.31 18.12 20.46 20.50 20.00   4.78 20.46 13.20 13.83 15.00 20.00
to:TO 20.50 20.00 20.00 10.50 20.50 20.50   0.00 20.00 20.50 20.00 20.00 20.00 10.00
pay:VB 15.46 10.00   0.50 17.51 20.37 18.97 20.00   1.00 20.46 11.96 11.96 15.00 19.94
$:$ 25.00 20.50 17.45   0.00 18.62 16.63 10.50 19.51 21.95 19.22 20.50 20.50   9.91
93:CD 24.96 20.46 20.46 18.62   0.00   7.99 20.50 20.46   5.23 20.34 19.86 20.46 19.76
million:CD 22.15 19.19 18.35 16.63   7.99   0.00 20.50 20.12 13.06 17.97 18.81 19.84 19.92
to:TO 20.50 20.00 20.00 10.50 20.50 20.50   0.00 20.00 20.50 20.00 20.00 20.00 10.00
buy:VB 15.46 10.00   6.25 19.51 20.46 20.12 20.00   0.00 20.46 11.96 11.96 15.00 20.00
117:CD 24.96 20.46 20.46 21.95   5.23 13.06 20.50 20.46   0.00 18.74 20.46 20.46 18.39
acres:NNS 10.46 15.00 11.96 19.22 20.34 17.97 20.00 11.96 18.74   0.00   2.28 10.00 20.00
land:NN   9.97 12.52 11.96 20.50 19.86 18.81 20.00 11.96 20.46   0.00   0.00   6.33 18.77
Las_Vegas:NNP 14.34 13.02 15.50 25.00 24.96 24.34 20.50 15.50 24.96 10.50   6.83   0.50 20.50
.:. 20.50 20.00 19.61   9.91 19.76 19.92 10.00 20.00 18.39 20.00 18.77 20.00   0.00
NO_WORD 10.00   1.00 10.00 10.00 10.00 10.00 10.00 10.00 10.00 10.00 10.00 10.00 10.00

Response: yes (INCORRECT)
Justification:
Alignment score: -6.0000
Features matched: NullPunisher.aux: has; Structure.argsMismatch: args have different parents but same relations: text "MGM_Grand_Inc." <-nsubj-- "agreed vs. hyp "MGM_Grand_Inc." <-nsubj-- "paid", which aligned to text "pay" args have different parents but same relations: text "." <-punct-- "agreed vs. hyp "." <-punct-- "paid", which aligned to text "pay"
Hand-tuned score: -2.0500
Threshold: -11.4590

Word similarity table built on Thu Jul 06 11:49:16 PDT 2006 using command:
java edu.stanford.nlp.rte.WordSimilarityGenerator -info /u/nlp/rte/data/byformat/align/stochastic/brandeis_dev.pipeline.align.xml -output /u/nlp/rte/data/byformat/wordsim/stochastic/brandeis_dev.pipeline.wordsim.html -lex.BasicWN off