Interpretation of negative right contexts

[Daniel Stein]

I am not sure if I do understand the behavior of negative right contexts. Which interpretation for the following grammar is correct?

![<NAME>]<N>

a) All nouns that are not preceded by a name

b) All nouns that are not also names

After reading the manual I thought it has to be a) but my results seem to reflect that b) is the correct interpretation...

Thanks in advance

[Denis Maurel]

Hi Daniel,

It is b) All nouns that are not also names

to implement a), you have to add a box with a token only

![<NAME>]<TOKEN> <N>

Best regards,

Denis Maurel


____________________________________
Professor Denis Maurel
Université François Rabelais Tours
LI (Computer Science Research Laboratory)
EPU-DI
64 avenue Jean-Portalis
37200 Tours
France
Phone: 33-2.47.36.14.35
Fax: 33-2.47.36.14.22
mailto:denis....@univ-tours.fr

http://www.univ-tours.fr/maurel

http://www.li.univ-tours.fr
http://tln.li.univ-tours.fr/

---

--
You received this message because you are subscribed to the Google Groups "Unitex-GramLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unitex-gramla...@googlegroups.com.
To post to this group, send email to unitex-...@googlegroups.com.
Visit this group at http://groups.google.com/group/unitex-gramlab.
To view this discussion on the web visit https://groups.google.com/d/msgid/unitex-gramlab/4d8b53bf-80d6-47b4-a4f3-f03cc83a20b0%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Daniel Stein]

Dear Denis,

thank you very much, this is great.

The only change I made is that I inserted a left context after the token in order to get the wanted results:

![<NAME>]<TOKEN>*<N>

Kind regards
Daniel

[Daniel Stein]

Hello again,

after some tests I found out, that there is a shortcoming in this solution:

![<NAME>]<TOKEN>*<N>

<N> will not match as well when there is no token on the left side (e.g. at the beginning of a paragraph). Do you know if it is possible to exclude certain contexts while still allowing the non-existence of a context?

Thanks in advance

Daniel

[Denis Maurel]

Hi Daniel,

There is always a <TOKEN>, for instance {S}!

Except at the begining of the text...

May be you can use {S} * <N>?

Best regards,

Denis Maurel


____________________________________
Professor Denis Maurel
Université François Rabelais Tours
LI (Computer Science Research Laboratory)
EPU-DI
64 avenue Jean-Portalis
37200 Tours
France
Phone: 33-2.47.36.14.35
Fax: 33-2.47.36.14.22
mailto:denis....@univ-tours.fr

http://www.univ-tours.fr/maurel

http://www.li.univ-tours.fr
http://tln.li.univ-tours.fr/

---

Hello again,

[Daniel Stein]

Hi Denis,

you are right, the only situation with no preceding token is the beginning of the text, my fault. This reduces the impact of the problem but unfortunatley, {S}*<N> does not seem to be a solution.
I could insert a special token at the beginning of all my texts but I would prefer a solution that is not a workaround...

Any ideas?

Kind regards
Daniel

[Fanny Grandry]

Hi Daniel,

You could use this special token {^}, that represents the start of the text ({$} represents the end of the text), in your left context. 

Another solution would be to add a {S} during the preprocessing when this special token {^} is matched.

Hope it helps

Best regards,

Fanny

--

Fanny Grandry

gra...@evercontact.com

NLP Engineer ◆ Evercontact ◆ www.evercontact.com

+33 1 55 43 79 20 ◆ 15 rue Jean-Baptiste Berlier 75013 Paris

Use this 1-click link to easily save my contact details

--
You received this message because you are subscribed to the Google Groups "Unitex-GramLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unitex-gramla...@googlegroups.com.
To post to this group, send email to unitex-...@googlegroups.com.
Visit this group at http://groups.google.com/group/unitex-gramlab.

To view this discussion on the web visit https://groups.google.com/d/msgid/unitex-gramlab/487a22f4-0773-404a-8064-a5bdf161c47e%40googlegroups.com.

[Daniel Stein]

Hi Fanny,
thanks for your help! I already tried to use ^ but did forget { } - now it works! Perfect :-)
Kind regards
Daniel

Am Donnerstag, 13. März 2014 14:59:03 UTC+1 schrieb Fanny Grandry:

Hi Daniel,

You could use this special token {^}, that represents the start of the text ({$} represents the end of the text), in your left context. 

Another solution would be to add a {S} during the preprocessing when this special token {^} is matched.

Hope it helps

Best regards,

Fanny

[eric.laporte]

Dear Daniel,

As regards your initial question, the manual gives interpretation b (section 6.3.1, fig. 6.15).