nlp - Identifying the context of word in sentence -

nlp - Identifying the context of word in sentence -

i created classifier classy class of nouns,adjectives, named entities in given sentence. have used big wikipedia dataset classification.

like :

where abraham lincoln born?

so classifier give short of result - word - class

where - question abraham lincoln - person, movie, book (because classifier find abraham lincoln in there categories) born - time

when titanic released?

when - question titanic - song, movie, vehicle, game (titanic classified in these categories)

is there way identify exact context word?

please see :

word sense disambiguation not help here. because there might not near word in sentence can help

lesk algorithm wordnet or sysnet not help. because suppose word bank lesk algo behave this

======== testing simple_lesk ===========

testing simple_lesk() ...

context: went bank deposit money

sense: synset('depository_financial_institution.n.01')

definition: financial establishment accepts deposits , channels money lending activities

testing simple_lesk() pos ...

context: river bank total of dead fishes

sense: synset('bank.n.01')

definition: sloping land (especially slope beside body of water)

here word bank suggested financial institute , slopping land. while in case getting such prediction titanic can movie or game.

i want know there other approach apart lesk algo, baseline algo, traditional word sense disambiguation can help me identify class right particular keyword?

titanic -

thanks using pywsd examples. regards wsd, there many other variants , i'm coding them myself during free time. if want see improve bring together me in coding open source tool =)

meanwhile, find next technologies more relevant task, such as:

knowledge base of operations population (http://www.nist.gov/tac/2014/kbp/) tokens/segments of text assigned entity , task link them or solve simplified question , reply task.

knowledge representation (http://groups.csail.mit.edu/medg/ftp/psz/k-rep.html)

knowledge extraction (https://en.wikipedia.org/wiki/knowledge_extraction)

the above technologies includes several sub-tasks such as:

wikification (http://nlp.cs.rpi.edu/kbp/2014/elreading.html) entity linking slot filling (http://surdeanu.info/kbp2014/def.php)

essentially you're asking tool np-complete ai scheme language/text processing, don't think such tool exists of yet. maybe it's ibm watson.

if you're looking field into, field out there if you're looking @ tools, wikification tools closest might need. (http://nlp.cs.rpi.edu/paper/wikificationproposal.pdf)

nlp data-mining nltk semantics