Where communities thrive

  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
Repo info
  • May 14 18:06
    rspeer commented #303
  • Apr 25 12:49
    gsittyz opened #303
  • Mar 18 18:53
    rspeer closed #302
  • Mar 18 18:53
    rspeer commented #302
  • Mar 18 00:22
    lalchand-pandia opened #302
  • Mar 08 11:42
    senisioi commented #300
  • Jan 05 15:38
    JohannaOm opened #301
  • Dec 18 2020 10:20
    annargrs opened #300
  • Nov 19 2020 15:43
    jlowryduda closed #299
  • Nov 19 2020 15:43
    jlowryduda commented #299
  • Nov 19 2020 13:10
    kartik7511 commented #299
  • Nov 19 2020 12:51
    henningsmith commented #299
  • Nov 19 2020 10:52
    kartik7511 edited #299
  • Nov 19 2020 10:49
    kartik7511 edited #299
  • Nov 19 2020 10:49
    kartik7511 opened #299
  • Nov 19 2020 03:07
    ghamzak closed #298
  • Nov 18 2020 17:52
    ghamzak opened #298
  • Nov 01 2020 09:08
    ChenTao98 opened #297
  • Oct 28 2020 14:02
    sdaitzman opened #296
  • Oct 13 2020 14:26
    felix9 opened #295
Alexandros Vassiliades
Can anybody answer me how much will the servers be down?? its 3 days now!!
Alexandros Vassiliades
Its fixed, its fixed :) :) :) :')
It is somewhat off-topic, but I figured that I would find some power users of search engines around here! I am working on a survey about search engines users and uses. I put together some questions at https://bimestriel.framapad.org/p/6phzfohmhs-9iot?lang=en
Hey guys! I have a question, I'm trying to find ConceptNet 5.6.0 image 2 in Community AMIs and connot found, why is happnes?

It is somewhat off-topic, but I figured that I would find some power users of search engines around here! I am working on a survey about search engines users and uses. I put together some questions at https://bimestriel.framapad.org/p/6phzfohmhs-9iot?lang=en

Please add relevant questions / topics you would like to see answered :)

I am working on a project where I am using ConceptNet database. I have a question. In the database, near the word in the brackets, there is a pos and a term. For example, Boston (n, location). Where can I find this relation (Boston is a location) in the ConceptNet database? I have looked at IsA relation and it is not the same.
That's a little bit different what I mean. For many words there are descriptions like that: armchair(n,artifact), tiger(n,animal) etc.
I was asking about this relation between the term and its area where it belongs in the brackets next to it.
ok, I get it. You can infer from A -> AtLocation -> B that B is a location.
it is not perfec, tho
You will have better data using wikidata.
@anastasiia_kai_twitter ^
Robyn Speer
@MaximDobrolinsky The most likely reason you'd fail to find the AMI is that you're looking in a region besides us-east-1
but on that note: the AMI for ConceptNet 5.8.1 is published, with some updated directions! See https://github.com/commonsense/conceptnet5/wiki/Running-your-own-copy
The build instructions (https://github.com/commonsense/conceptnet5/wiki/Build-process) are also updated, and one alternative is to use the new conceptnet-deployment (https://github.com/commonsense/conceptnet-deployment ) repository, which sets up ConceptNet using Puppet 5
Stanislav Shelemekh
Screenshot 2020-09-17 at 13.09.33.png
Hey! On the conceptnet.io you can find such UI component
Where does the "communication" part come from?
Robyn Speer
When sources like WordNet and Wiktionary give us disambiguating information, we attempt to preserve it
a custom script in building ConceptNet decides how to assign WordNet senses to human-readable categories like "communication". It comes from the WordNet properties called "domain_category" and "lexical_domain"
Stanislav Shelemekh
Okay, thank you so much, what about the big CSV file in the Downloads ( https://github.com/commonsense/conceptnet5/wiki/Downloads )? Does it have any "domain_category" or "lexical_domain" info? I was trying to find any such cases but didn't succeed :(
I do not know the answer, but a way to tell where communication come from is to look at the python code of the given page
Pranav Pawar

Hi, I am new to conceptnet; Have set up the Amazon AMI, following instructions from here https://github.com/commonsense/conceptnet5/wiki/Running-your-own-copy

And as mentioned here, there is a timeout error issue with using AMI; though I tried out the solution for this problem, as mentioned in the documentation

Here's a command to warm up the entire disk, by accessing every byte of data on the disk:

sudo dd if=/dev/xvda of=/dev/null bs=16M

But it doesn't seem to work, facing issue of missing /dev/xvda directory

Can somebody help me out resolving this? Or any other solution for the timeout issue is welcome!

Zhang Cheng
Hi, everyone! I'm new to ConceptNet. And I want to know how to calculate the relatedness based on the NumberBtach. Is it just cosine similarity or dot product?
Zhang Cheng
s = a 1 · a 2 + b 1 · b 2 + w 1 (b 2 − a 2 ) · (b 1 − a 1 ) + w 2 (b 2 − b 1 ) · (a 2 − a 1 )
Is this the formula for relatedness? While w 1 = 0.2 and w 2 = 0.6 is for NumberBatch 16.09, the weight for latest NumberBatch(19.08) is what?
Hi, Wondering if anyone would be so kind to assist in installing ConceptNet with the database on a machine. I've spent a week on it using virtual machines, configuring puppet, resetting my machine, etc. and pulling my hair out...super frustrated. :-(
@christat13 Same thing all over again, I put online a simple way to load conceptnet in a database that has a REST interface, but nobody gave any feedback, so I do not know how to proceed. See https://github.com/amirouche/easy-conceptnet
Does conceptnet APIs accept spelling mistakes?
Anna Rogers
@christat13 if a local database would suffice, here's a Python library that makes things simple: https://pypi.org/project/conceptnet-lite/

Error in rule convert_opensubtitles_ft:
jobid: 0
output: data/vectors/fasttext-opensubtitles.h5

CalledProcessError in line 674 of /home/conceptnet5/Snakefile:
Command 'set -euo pipefail; CONCEPTNET_DATA=data cn5-vectors convert_fasttext -n 2000000 data/raw/vectors/ft-opensubtitles.vec.gz data/vectors/fasttext-opensubtitles.h5' returned non-zero exit status 137.
File "/home/conceptnet5/Snakefile", line 674, in __rule_convert_opensubtitles_ft
File "/usr/lib/python3.6/concurrent/futures/thread.py", line 56, in run
Exiting because a job execution failed. Look above for error message
[Fri Nov 27 21:43:20 2020]
Finished job 183.
11 of 94 steps (12%) done
Shutting down, this might take some time.
Exiting because a job execution failed. Look above for error message
Complete log: /home/conceptnet5/.snakemake/log/2020-11-27T213538.909921.snakemake.log

getting this error??
it would be great if anyone can help
@amirouche Hi... when trying 'make database-download' step facing 'make: * No rule to make target 'database-download'. Stop.
' issue. Can you help?
@subhadeepdas7593-prog I just figured that I commited the wrong thing
It will take a few days to fix
I need to go though my backups
Hallo ,
Is ConceptNet version 5 from 2012 available to download? Unfortunately i couldn't find a download source for this version.
Alistair Nottle (Personal)
Hello! I am trying to install a local instance of ConceptNet, using the Puppet method. It initially executes fine, but towards the end I'm getting a consistent error. Anybody got any hints?! thank you.
[Tue Dec 22 15:02:54 2020]
rule join_propagate:
    input: data/vectors/numberbatch-biased.h5.shard0, data/vectors/numberbatch-biased.h5.shard1, data/vectors/numberbatch-biased.h5.shard2, data/v
ectors/numberbatch-biased.h5.shard3, data/vectors/numberbatch-biased.h5.shard4, data/vectors/numberbatch-biased.h5.shard5
    output: data/vectors/numberbatch-biased.h5
    jobid: 18
    resources: ram=24

Unable to open/create file 'data/vectors/numberbatch-biased.h5.shard5'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "/home/conceptnet/env/bin/cn5-vectors", line 33, in <module>
    sys.exit(load_entry_point('ConceptNet', 'console_scripts', 'cn5-vectors')())
  File "/usr/lib/python3/dist-packages/click/core.py", line 764, in __call__
    return self.main(*args, **kwargs)
  File "/usr/lib/python3/dist-packages/click/core.py", line 717, in main
    rv = self.invoke(ctx)
  File "/usr/lib/python3/dist-packages/click/core.py", line 1137, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "/usr/lib/python3/dist-packages/click/core.py", line 956, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "/usr/lib/python3/dist-packages/click/core.py", line 555, in invoke
    return callback(*args, **kwargs)
  File "/home/conceptnet/conceptnet5/conceptnet5/vectors/cli.py", line 339, in run_join_shard_files
    join_shards(filename, nshards, sort=sort)
  File "/home/conceptnet/conceptnet5/conceptnet5/vectors/retrofit.py", line 58, in join_shards
    shard = load_hdf(output_filename + '.shard%d' % i)
  File "/home/conceptnet/conceptnet5/conceptnet5/vectors/formats.py", line 21, in load_hdf
    return pd.read_hdf(filename, 'mat', encoding='utf-8')
  File "/home/conceptnet/env/lib/python3.8/site-packages/pandas/io/pytables.py", line 389, in read_hdf
    store = HDFStore(path_or_buf, mode=mode, errors=errors, **kwargs)
  File "/home/conceptnet/env/lib/python3.8/site-packages/pandas/io/pytables.py", line 553, in __init__
    self.open(mode=mode, **kwargs)
  File "/home/conceptnet/env/lib/python3.8/site-packages/pandas/io/pytables.py", line 729, in open
    raise IOError(str(err)) from err
OSError: HDF5 error back trace

  File "H5F.c", line 509, in H5Fopen
    unable to open file
  File "H5Fint.c", line 1400, in H5F__open
    unable to open file
  File "H5Fint.c", line 1709, in H5F_open
    unable to read root group
  File "H5Groot.c", line 239, in H5G_mkroot
    can't check if symbol table message exists
  File "H5Omessage.c", line 883, in H5O_msg_exists
    unable to protect object header
  File "H5Oint.c", line 1100, in H5O_protect
    unable to load object header
  File "H5AC.c", line 1625, in H5AC_protect
    H5C_protect() failed
  File "H5C.c", line 2362, in H5C_protect
    can't load entry
 File "H5C.c", line 6712, in H5C_load_entry
    incorrect metadatda checksum after all read attempts
  File "H5Ocache.c", line 219, in H5O__cache_get_final_load_size
    can't deserialize object header prefix
  File "H5Ocache.c", line 1231, in H5O__prefix_deserialize
    bad object header version number

End of HDF5 error back trace

Unable to open/create file 'data/vectors/numberbatch-biased.h5.shard5'
Is there a lightweight, English-only version of Conceptnet for integrating into other apps?
5 replies
Wellington Franco

Anybody getting this error??
Please, it would be great if anyone can help

return _wf_cache[args]
KeyError: ('##', 'ja', 'best', 0.0)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "/root/anaconda3/bin/cn5-vectors", line 11, in <module>
load_entry_point('ConceptNet', 'console_scripts', 'cn5-vectors')()
File "/root/anaconda3/lib/python3.7/site-packages/click/core.py", line 722, in call
return self.main(args, kwargs)
File "/root/anaconda3/lib/python3.7/site-packages/click/core.py", line 697, in main
rv = self.invoke(ctx)
File "/root/anaconda3/lib/python3.7/site-packages/click/core.py", line 1066, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/root/anaconda3/lib/python3.7/site-packages/click/core.py", line 895, in invoke
return ctx.invoke(self.callback,
File "/root/anaconda3/lib/python3.7/site-packages/click/core.py", line 535, in invoke
return callback(
args, *kwargs)
File "/dados/conceptnet5/conceptnet5/vectors/cli.py", line 288, in run_miniaturize
mini = miniaturize(frame, other_vocab=othervocab, k=k)
File "/dados/conceptnet5/conceptnet5/vectors/miniaturize.py", line 51, in miniaturize
term for term in frame.index if '
' not in term and termfreq(term) >= 1e-8
File "/dados/conceptnet5/conceptnet5/vectors/miniaturize.py", line 51, in <listcomp>
term for term in frame.index if '
' not in term and term_freq(term) >= 1e-8
File "/dados/conceptnet5/conceptnet5/vectors/miniaturize.py", line 25, in term_freq
return wordfreq.word_frequency(term, lang)
File "/root/anaconda3/lib/python3.7/site-packages/wordfreq/init.py", line 301, in word_frequency
_wf_cache[args] = _word_frequency(
File "/root/anaconda3/lib/python3.7/site-packages/wordfreq/init.py", line 244, in _word_frequency
tokens = lossy_tokenize(word, lang)
File "/root/anaconda3/lib/python3.7/site-packages/wordfreq/tokens.py", line 313, in lossy_tokenize
tokens = tokenize(text, lang, include_punctuation, external_wordlist)

hello everyone! i apologize if this has been asked before, but is there a way to report incorrect connections on conceptnet? when i searched "robot", it listed that robots are capable of "[crying] salt water" and "[playing] the ocarina". neither of these seem accurate to me, but then again, i'm not a robot expert. any advice would be appreciated :)
Hanna L Tischer
Hey I was wondering what the best way to parse the csv file for conceptnet is? Also is there a file that only has English words?
Rejnald Lleshi
Hi everyone! Whenever I query the REST API like so:'http://api.conceptnet.io/c/en/' + word + '?limit=2000' I get a variable result even though I am setting a limit=2000. Any idea why this is happening?
@hannalt hi hanna, I wrote a simple python python script that filteres english assertions from the raw assertions.csv, you can take a look here: