Support for users of CCG software closed 7-27-20

[Sourajit Basak <sourajit.basak AT gmail.com>]

hello,

I am trying to extract features as mentioned in the following examples. The attached file contains the sentence (question) and its POS annotated representation. I have the following queries.

1)  Feature: One word each around the target word
Is this rule correct ? -1 loc: w(x=Java) [-1,1]; -1 loc: w(x=XML-RPC) [-1,1] I do not see the lexicon and example populated; guess that the rule is not fired. This fired "4 loc: w [-1, -1]" whereas this "Java loc: w [-1, 1]" failed.

2) What will be the feature rule for POS tags around the target ?

3) What will be the feature for non overlapping phrases around the target ?

I am afraid I didn't fully comprehend the FeX manual. A brief explanation for the above will be extremely helpful.

rgds,
Sourajit


Preprocessing steps:-
#1. Example: How to call an XML-RPC method from Java code ?

#2. POS tagger and chunker output
$$java -Xmx512m edu.illinois.cs.cogcomp.lbj.chunk.ChunksAndPOSTags 
<plain_text_file>
POS Tagged and chunked: [ADVP (WRB How) ] [VP (TO to) (VB call) ] [NP (DT an) 
(NNP XML-RPC) (NN method) ] [PP (IN from) ] [NP (NNP Java) (NNP code) ] (. ?) 

#3. Columnar output. Used as the actual corpus.
$$  perl chunks_to_cols.pl <pos_tagged_question>

O       0       0       B-ADVP  WRB     How     x       0       0
O       0       1       B-VP    TO      to      x       0       0
O       0       2       I-VP    VB      call    x       0       0
B-NP    0       3       B-NP    DT      an      x       0       0
I-NP    0       4       I-NP    NNP     XML-RPC x       0       0
O       0       5       B-PP    IN      from    x       0       0
B-NP    0       6       B-NP    NNP     Java    x       0       0
I-NP    0       7       I-NP    NNP     Code    x       0       0
O       0       8       O       .       ?       x       0       0