python - Absolute position of leaves in NLTK tree -

-

i trying find span (start index, end index) of noun phrase in given sentence. following code extracting noun phrases

sent=nltk.word_tokenize(a) sent_pos=nltk.pos_tag(sent) grammar = r"""     nbar:         {<nn.*|jj>*<nn.*>}  # nouns , adjectives, terminated nouns      np:         {<nbar>}         {<nbar><in><nbar>}  # above, connected in/of/etc...     vp:         {<vbd><pp>?}         {<vbz><pp>?}         {<vb><pp>?}         {<vbn><pp>?}         {<vbg><pp>?}         {<vbp><pp>?} """  cp = nltk.regexpparser(grammar) result = cp.parse(sent_pos) nounphrases = [] subtree in result.subtrees(filter=lambda t: t.label() == 'np'):   np = ''   x in subtree.leaves():     np = np + ' ' + x[0]   nounphrases.append(np.strip())

for a = "the american civil war, known war between states or civil war, civil war fought 1861 1865 in united states after several southern slave states declared secession , formed confederate states of america.", noun phrases extracted are

['american civil war', 'war', 'states', 'civil war', 'civil war fought', 'united states', 'several southern', 'states', 'secession', 'confederate states', 'america'].

now need find span (start position , end position of phrase) of noun phrases. example, span of above noun phrases

[(1,3), (9,9), (12, 12), (16, 17), (21, 23), ....].

i'm new nltk , i've looked http://www.nltk.org/_modules/nltk/tree.html. tried use tree.treepositions() couldn't manage extract absolute positions using these indices. appreciated. thank you!

there isn't implicit function returns offsets of strings/tokens highlighted https://github.com/nltk/nltk/issues/1214

but can use ngram searcher used ribes score https://github.com/nltk/nltk/blob/develop/nltk/translate/ribes_score.py#l123 

>>> nltk import word_tokenize >>> nltk.translate.ribes_score import position_of_ngram >>> s = word_tokenize("the american civil war, known war between states or civil war, civil war fought 1861 1865 in united states after several southern slave states declared secession , formed confederate states of america.") >>> position_of_ngram(tuple('american civil war'.split()), s) 1 >>> position_of_ngram(tuple('confederate states of america'.split()), s) 43

(it returns starting position of query ngram)


Comments

Post a Comment

Popular posts from this blog

Load Balancing in Bluemix using custom domain and DNS SRV records -

-
i'm trying load balancing of application in bluemix using custom domain , srv record in dns manager. created own doamin ccbluemix.dynu.com , added custom domain bluemix. created application in region , created route custom domain myapp.ccbluemix.dynu.com deployed same application in uk region , created route custom domain myapp.ccbluemix.dynu.com there well. now whenever user hits route want load balancing between , uk region alternate requests served , uk regions. for created srv records like _http._tcp.ccbluemix.dynu.com srv eu-gb.mybluemix.net [priority: 0] [weight: 50] 10 _http._tcp.ccbluemix.dynu.com srv mybluemix.net [priority: 0] [weight: 50] 10 but nothing happening , i'm not able access application using custom route @ all. if create 'a' record cc-load-bal-app.ccbluemix.dynu.com 75.126.81.68 i can access app using custom route ends hitting server. how can achieve load balancing using srv records? wrong creating srv records here? i read ne...
Read more

oracle - pls-00402 alias required in select list of cursor to avoid duplicate column names -

-
my cursor : select xzis.msisdn msisdn, xzis.in_account_number in_account_number xxobw_zte_in_sync xzis not exists (select 1 hz_contact_points hcp, hz_cust_accounts hca hcp.owner_table_id = hca.party_id , hcp.owner_table_name = 'hz_parties' , hcp.phone_number = substr (xzis.msisdn, -8) , nvl (hca.attribute12, 'nil') = xzis.in_account_number); i have given alias columns, still facing issue. suggestions? it's cursor. 1 has no such problems. problem is, when you're joining 2 tables, or apply wrong aliases in subselect, cursor may @ end return 2 columns same name/alias. the problem not need alias, no 2 columns can have same alias (the name counting alias, when no alias given). apparently has happened, far can tell, it's ...
Read more

python - Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>] error -

-
Image
i getting same error when python manage.py runserver. there no such error when have ubuntu 15.10. got started when upgraded ubuntu 16.04. question might duplicate have tried solution provided question have applied command dpkg --configure -a, have done apt-get update,upgrade,dist-upgrade, clean, -f install. have reinstalled python2.7 , python3 no success though. here screenshot while trying run django application i got after re-installing python please me? bashrc file # ~/.bashrc: executed bash(1) non-login shells. # see /usr/share/doc/bash/examples/startup-files (in package bash-doc) # examples # if not running interactively, don't case $- in *i*) ;; *) return;; esac export pythonhome="/usr/bin/python" echo $virtualenvwrapper_python echo 'tushant' # don't put duplicate lines or lines starting space in history. # see bash(1) more options histcontrol=ignoreboth # append history file, don't overwrite shopt -s histappend # setting h...
Read more