I want to extract name(firstnames and lastnames) from a text using php.
Example:
From text below I want to extract names(in this case Aline Wright and Jesse Wright)
Aline Wright is a cancer survivor,
amputee and a newlywed. Wednesday
night she began to show signs she was
having a stroke.
"I started feeling some left arm
numbness and a facial droop," said
Aline.
"It appeared to me that I was probably
having a stroke."
That's when her husband of four days,
Jesse Wright, put her in the car and
rushed her to the Erlanger Medical
Center. Wright knows an emergency. He
is a nurse technician at Erlanger.
If you want to do named entity recognition (NER) using PHP, you'll need to either call out to an external NER package or make use of an online natural language processing API.
APIs
The three major NLP API are listed below. Complexity Intelligence and Alchemy API will probably be moderately easier to use for beginners than OpenCalais.
- Complexity Intelligence - see the PHP named entity tagging sample code here.
- Alchemy API - download their PHP SDK here.
- OpenCalais - see the PHP sample code here.
NER Packages
A few software packages you can use for NER are:
- Stanford CoreNLP (Java)
- LingPipe (Java)
- NLTK (Python)
- OpenNLP (Java)
- YamCha (C++)
Of these, Stanford CoreNLP is probably a good place to start. Similar to many NLP APIs, it provides a complete processing pipeline for common tasks such as NER.
I'd use a named entity recognizer. There are many of these up on CPAN where there is an active linguistics community.
Then, in PHP, do something like:
$result = \`perl named_entity_recogniser.pl "myText"\`;
I wrote this PHP extension but be prepared to compile from source.
https://github.com/rjjakes/MITIE-PHP
