I want to extract named entities from a text but I don't know whether that comes under classification or not.if it comes under classification then how to prepare classes for recognizing entities in a text.(does the collection of few org names and person names and using them as my training dataset makes the solution?)
$\begingroup$
$\endgroup$
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
Simply put, Named Entity Recognition (NER) is a multi-class structured prediction (classification) problem, so you have a sequence of words and you want to label each one most of the time with these labels ( start-of-a-person-name, continue-of-a-person-name, start-of-an-org-name, continue-of-an-org-name, start-of-a-location-name, continue-of-a-location-name, other). Note that these are not the only classes, classes with different granularity can also be used.
If you want to solve this problem with SVM you can use StructSVM or other variations of SVM for structure prediction. Though the common baseline for this task uses Maximum Entropy (Maxent)(log-linear) models.
$\begingroup$
$\endgroup$
Named entity recognition can be seen as a multi class classification problem. A large data set will be required to train a model(preferably bayesian) for recognising different named entities. You can use word embedding (like google word2vec) to for preparing your training set. Also, if you can try IBM Bluemix AlchemyAPI for named entity extraction if you want to get your job done.
