Query apparent anomalies in en/words/words-medical*
Paul McQuesten
after the numbered questions, along with the corresponding egrep.
1. None of the words in words-medical.adv.1 and words-medical.prep.1
have subscripts. Is this okay?
2. Some words in words-medical.v* have no subscript, although most do.
3. Most words in words-medical.v.4.[1234] have .v subscript.
4. words-medical.v.4.4 begins with a blank line.
5. Most words in words-medical.v.4.5 have .g subscript; 19 have none.
6. x_ray and X_ray differ only in initial capital. Does this impact
sentence
detection and/or entity recognition?
0. Are such matters appropriate for this group?
//
-------------------------------------------------------------------------------------
// 2. words-medical.v: some words with no subscript
en/words> egrep -c -v '[.][a-z]$' words-medical.v*
words-medical.v.4.1:11
words-medical.v.4.2:8
words-medical.v.4.3:21
words-medical.v.4.4:20
words-medical.v.4.5:19
//
-------------------------------------------------------------------------------------
// 3. words-medical.v.4.[1234] supposed to have .v subscript??
// words-medical.v.4.4 begins with a blank line; okay??
en/words> egrep -v '[.]v$' words-medical.v.4.[1234] >nov
words-medical.v.4.1:adenosine_diphosphate-ribosylate
words-medical.v.4.1:cross_dress
words-medical.v.4.1:cross_fertilize
words-medical.v.4.1:cross_match
words-medical.v.4.1:freeze_dry
words-medical.v.4.1:jack_knife
words-medical.v.4.1:over_weight
words-medical.v.4.1:portacaval_shunt
words-medical.v.4.1:scar_cicatrise
words-medical.v.4.1:scar_cicatrize
words-medical.v.4.1:tap_dance
words-medical.v.4.2:adenosine_diphosphate-ribosylates
words-medical.v.4.2:cross_fertilizes
words-medical.v.4.2:cross_matches
words-medical.v.4.2:freeze_dries
words-medical.v.4.2:jack_knifes
words-medical.v.4.2:over_weights
words-medical.v.4.2:portacaval_shunts
words-medical.v.4.2:scar_cicatrises
words-medical.v.4.3:adenosine_diphosphate-ribosylated
words-medical.v.4.3:bandpass_filtered
words-medical.v.4.3:bench_pressed
words-medical.v.4.3:catalase_tested
words-medical.v.4.3:C_banded
words-medical.v.4.3:cross_fertilized
words-medical.v.4.3:cross_matched
words-medical.v.4.3:cryo_sectioned
words-medical.v.4.3:field_tested
words-medical.v.4.3:free_grafted
words-medical.v.4.3:freeze_dried
words-medical.v.4.3:immuno_assayed
words-medical.v.4.3:jack_knifed
words-medical.v.4.3:Nissl_stained
words-medical.v.4.3:over_weighted
words-medical.v.4.3:portacaval_shunted
words-medical.v.4.3:scar_cicatrised
words-medical.v.4.3:skate_boarded
words-medical.v.4.3:ventricular_hypertrophied
words-medical.v.4.3:x_rayed
words-medical.v.4.3:X_rayed
words-medical.v.4.4:
words-medical.v.4.4:adenosine_diphosphate-ribosylating
words-medical.v.4.4:bandpass_filtering
words-medical.v.4.4:bench_pressing
words-medical.v.4.4:catalase_testing
words-medical.v.4.4:cross_fertilizing
words-medical.v.4.4:cryo_sectioning
words-medical.v.4.4:field_testing
words-medical.v.4.4:free_grafting
words-medical.v.4.4:freeze_drying
words-medical.v.4.4:immuno_assaying
words-medical.v.4.4:jack_knifing
words-medical.v.4.4:Nissl_staining
words-medical.v.4.4:over_weighting
words-medical.v.4.4:portacaval_shunting
words-medical.v.4.4:scar_cicatrising
words-medical.v.4.4:scar_cicatrizing
words-medical.v.4.4:ventricular_hypertrophying
words-medical.v.4.4:x_raying
words-medical.v.4.4:X_raying
//
-------------------------------------------------------------------------------------
// 5. words-medical.v.4.5 supposed to have .g subscript??
en/words> egrep -v '[.]g$' words-medical.v.4.5
adenosine_diphosphate-ribosylating
bandpass_filtering
bench_pressing
catalase_testing
cross_fertilizing
cryo_sectioning
field_testing
free_grafting
freeze_drying
immuno_assaying
jack_knifing
Nissl_staining
over_weighting
portacaval_shunting
scar_cicatrising
scar_cicatrizing
ventricular_hypertrophying
x_raying
X_raying
Linas Vepstas
> I have a few questions about words-medical. There is detailed data
> after the numbered questions, along with the corresponding egrep.
>
> 1. None of the words in words-medical.adv.1 and words-medical.prep.1
> have subscripts. Is this okay?
Yes. Subscripts aren't really needed; they are used for only one
reason: to distinguish multiple dictionary entries for the same word.
That they also happen to be kind-of-like part-of-speech tags is
a lucky accident for the user.
> 2. Some words in words-medical.v* have no subscript, although most do.
Feel free to send in patches, if you wish.
> 3. Most words in words-medical.v.4.[1234] have .v subscript.
Presumably, they're verbs. Although past-tense verbs
should be tagged with .v-d so that people who care about
past tense (e.g. the stanford-parser compat mode in relex)
get it right.
> 4. words-medical.v.4.4 begins with a blank line.
shouldn't matter.
> 5. Most words in words-medical.v.4.5 have .g subscript; 19 have none.
>
> 6. x_ray and X_ray differ only in initial capital. Does this impact
> sentence
> detection and/or entity recognition?
Possibly. I'd have to experiment & play with this. The capital-letter
logic is frustratingly delicate.
> 0. Are such matters appropriate for this group?
Yes.
> -------------------------------------------------------------------------------------
> // 3. words-medical.v.4.[1234] supposed to have .v subscript??
> // words-medical.v.4.4 begins with a blank line; okay??
> en/words> egrep -v '[.]v$' words-medical.v.4.[1234] >nov
> words-medical.v.4.1:adenosine_diphosphate-ribosylate
> words-medical.v.4.1:cross_dress
> words-medical.v.4.1:cross_fertilize
Nothing with an underscore can ever have a subscript;
this is due to a deep technical limitation.
--linas
Paul McQuesten
> > -------------------------------------------------------------------------------------
> > // 3. words-medical.v.4.[1234] supposed to have .v subscript??
> > // words-medical.v.4.4 begins with a blank line; okay??
> > en/words> egrep -v '[.]v$' words-medical.v.4.[1234] >nov
> > words-medical.v.4.1:adenosine_diphosphate-ribosylate
> > words-medical.v.4.1:cross_dress
> > words-medical.v.4.1:cross_fertilize
>
> Nothing with an underscore can ever have a subscript;
> this is due to a deep technical limitation.
> > 2. Some words in words-medical.v* have no subscript, although most do.
> Feel free to send in patches, if you wish.
above covers this case.
> > 5. Most words in words-medical.v.4.5 have .g subscript; 19 have none.
and v.4.5. Which rule do they get? I have been unable to devise a test
case, since I do not know how to use adenosine_diphosphate-
ribosylating in a sentence ;-)
Linas Vepstas
>> > 5. Most words in words-medical.v.4.5 have .g subscript; 19 have none.
> I should have pointed out that several of them appear in both v.4.4
> and v.4.5. Which rule do they get? I have been unable to devise a test
> case, since I do not know how to use adenosine_diphosphate-
> ribosylating in a sentence ;-)
That depends. The *v.*.4 words are gerunds, the v.*.5 are present
participles (or maybe the other way around) so which is used depends
on how its used.
--linas