RM95 Chunking Benchmark Task

The RM95 Chunking Benchmark Task is a Benchmark Task and Benchmark Dataset proposed in (Ramshaw & Marcus, 1995).

• Context:
  • A Benchmark Task for Shallow Parsing: RM95 Chunking Benchmark Task.
  • The Data can be found at ftp://ftp.cis.upenn.edu/pub/chunker/.
• Example(s):
  • One sentence:

He	PRP	I	I	O	B
reckons	VBZ	O	O	I	B
the	DT	I	I	O	B
current	JJ	I	I	O	B
account	NN	I	I	O	B
deficit	NN	I	I	O	B
will	MD	O	O	I	B
narrow	VB	O	O	I	B
to	TO	O	O	I	B
only	RB	I	O	I	B
#	#	I	I	O	B
1.8	CD	I	I	O	B
billion	CD	I	I	O	B
in	IN	O	O	I	B
September	NNP	I	I	O	B
.	.	O	O	I	B

• See: CoNLL-1999 Shared Task.

---

---

References

• An implementation of their NP Chunker: http://www.dcs.shef.ac.uk/~mark/phd/software/chunker.html

2001

• (Kudo & Matsumoto, 2001) ⇒ Taku Kudo, and Yuji Matsumoto. (2001). “Chunking with Support Vector Machines.” In: Proceedings of NAACL 2001.
  • Had achieved the highest performance by 2002 on the RM95Chunking Benchmark Task.
    • P=94.15%, R=94.29%, F=94.22

1995

• (Ramshaw & Marcus, 1995) ⇒ Lance A. Ramshaw, and Mitch P. Marcus. (1995). “Text Chunking Using Transformation-based Learning.” In: Proceedings of the Third ACL Workshop on Very Large Corpora (WVLC 1995).

1989

• (Abney, 1989) ⇒ Steven P. Abney. (1989). “Parsing By Chunks." The MIT Parsing Volume, 1988-89. Center for Cognitive Science, MIT.

---