SEMAFOR v2.1 (Probables Bugs or Errors in semafor-semantic-parser)

[adrian...@gmail.com]

Dear doctors

My name is Adriano Proscia, a computer science student of the University of Bari (Italy).
I'm doing a thesis on Data Mining, and I'm using the tool SEMAFOR v2.1 (https://github.com/Noahs-ARK/semafor-semantic-parser) within my thesis system.

Excuse me, but I'm having problems with the tool SEMAFOR v2.1. Could you help me figure out what could be the solutions to these problems ? Where am I wrong ?

For example when i run semafor several times on the same input file, it behaves differently (sometimes it works fine and sometimes raises various objections in several parts). I would like to know if there is a maximum for the number of rows of the input file and I wonder if there is a maximum length of characters for each line of the input file. Sometimes raises exceptions in phase 'Sentence Processing' ( 'java.lang.OutOfMemoryError: GC overhead limit exceeded' or 'java.lang.OutOfMemoryError: Java heap space: failed reallocation of scalar Replaced objects') and other times in the phase 'Loading data' ( 'Loading data ....
Problem. Count of line 0 (73) not equal to zeroth line (312)'). Addition is then raised an exception in the final phase of Producing XML document ( 'java.lang.IndexOutOfBoundsException: Index: 41, Size: 41').

The consequence of all this is that the XML output file that contains the instances frame is empty. I need the XML file output to run my system.

My input files are created based on the 20 Newsgroups data set (http://qwone.com/~jason/20Newsgroups/).

I have a notebook asus N550JV (16 GB RAM DDR3, Intel® Core™ i7 4700HQ Processor) with ubuntu 16 LTS OS.

I am attaching (link DropBox) some results output of fnParserDriver.sh file together with the respective temporary files of semafor-semantic-parser and stackedParserServer and the respective input file.

I have done nine runs on nine several input files. I vacuum every run the temp folder of semafor-semantic-parser (in semafor-semantic-parser/release/temp).

In the compressed file RunsOfSemafor.7z you will find different folder called 'Run_n' and within each one you find the input file, the temporary files, the output of tool semafor and a file README.txt that shows info about errors or exception in that n-th run.

Link dropBox of file RunsOfSemafor.7z: https://www.dropbox.com/s/dcn6t4qfn4ewexd/RunsOfSemafor.7z?dl=0

Also, excuse me for my english.

I would greatly appreciate it if you kindly give me some feedback about these problems.
Trusting in your readiness, I offer you my cordial greetings.

Adriano Proscia

[Sam Thomson]

Hi Adriano,

It sounds like you're running out of memory. You can set the maximum memory allowed to a Java process with the -Xmx flag. Look for that flag in the release/fnParserDriver.sh script, and increase the value of its argument.

Cheers,

-Sam