underspecified / web-ka Goto Github PK

why in every seed the result always have score : 1 and Iteration: 0?
i think the score should be in decimals numbers
here is the mongodb result after running the code :

db.reverb_wikipedia_1000_2_promotes_esp_i_rel_nokeep.find()
{ "_id" : ObjectId("59a6c363c7a447369f600f2c"), "arg1" : "Probiotics", "arg2" : "overall good health", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f2d"), "arg1" : "Vitamin C", "arg2" : "healthy immune functions", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f2e"), "arg1" : "Vitamin C", "arg2" : "dietary iron", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f2f"), "arg1" : "Vitamin D", "arg2" : "Calcium", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f30"), "arg1" : "Sunlight", "arg2" : "Vitamin D", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f31"), "arg1" : "Vitamin A", "arg2" : "immunity", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f32"), "arg1" : "Sugary food and drinks", "arg2" : "tooth decay", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f33"), "arg1" : "Sugars", "arg2" : "tooth decay", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f34"), "arg1" : "HGF", "arg2" : "human MSCs", "score" : 1, "it" : 0 }
{ "_id" : ObjectId("59a6c364c7a447369f600f35"), "arg1" : "Vitamin E", "arg2" : "circulation", "score" : 1, "it" : 0 }

i already read and run all of this tables
we already run bootstrap_ex.sh &
start_espresso_tests.sh
we don't know which tables is the result
can u give me a clue step by step to run all of your code?
Thank u for your support

below is all tables in mongodb:
reverb_wikipedia_1000_2
reverb_wikipedia_1000_2_F_all
reverb_wikipedia_1000_2_F_i
reverb_wikipedia_1000_2_F_ip
reverb_wikipedia_1000_2_F_p
reverb_wikipedia_1000_2_inhibits_esp_i_pc_keep
reverb_wikipedia_1000_2_inhibits_esp_i_pc_nokeep
reverb_wikipedia_1000_2_inhibits_esp_i_rel_keep
reverb_wikipedia_1000_2_inhibits_esp_i_rel_nokeep
reverb_wikipedia_1000_2_inhibits_esp_p_pc_keep
reverb_wikipedia_1000_2_inhibits_esp_p_pc_nokeep
reverb_wikipedia_1000_2_inhibits_esp_p_rel_keep
reverb_wikipedia_1000_2_inhibits_esp_p_rel_nokeep
reverb_wikipedia_1000_2_max_pmi_ip
reverb_wikipedia_1000_2_necessary_esp_i_pc_keep
reverb_wikipedia_1000_2_necessary_esp_i_pc_nokeep
reverb_wikipedia_1000_2_necessary_esp_i_rel_keep
reverb_wikipedia_1000_2_necessary_esp_i_rel_nokeep
reverb_wikipedia_1000_2_necessary_esp_p_pc_keep
reverb_wikipedia_1000_2_necessary_esp_p_pc_nokeep
reverb_wikipedia_1000_2_necessary_esp_p_rel_keep
reverb_wikipedia_1000_2_necessary_esp_p_rel_nokeep
reverb_wikipedia_1000_2_part_of_esp_i_pc_keep
reverb_wikipedia_1000_2_part_of_esp_i_pc_nokeep
reverb_wikipedia_1000_2_part_of_esp_i_rel_keep
reverb_wikipedia_1000_2_part_of_esp_i_rel_nokeep
reverb_wikipedia_1000_2_part_of_esp_p_pc_keep
reverb_wikipedia_1000_2_part_of_esp_p_pc_nokeep
reverb_wikipedia_1000_2_part_of_esp_p_rel_keep
reverb_wikipedia_1000_2_part_of_esp_p_rel_nokeep
reverb_wikipedia_1000_2_pmi_ip
reverb_wikipedia_1000_2_promotes_esp_i_pc_keep
reverb_wikipedia_1000_2_promotes_esp_i_pc_nokeep
reverb_wikipedia_1000_2_promotes_esp_i_rel_keep
reverb_wikipedia_1000_2_promotes_esp_i_rel_nokeep
reverb_wikipedia_1000_2_promotes_esp_p_pc_keep
reverb_wikipedia_1000_2_promotes_esp_p_pc_nokeep
reverb_wikipedia_1000_2_promotes_esp_p_rel_keep
reverb_wikipedia_1000_2_promotes_esp_p_rel_nokeep
reverb_wikipedia_1000_2_source_of_esp_i_pc_keep
reverb_wikipedia_1000_2_source_of_esp_i_pc_nokeep
reverb_wikipedia_1000_2_source_of_esp_i_rel_keep
reverb_wikipedia_1000_2_source_of_esp_i_rel_nokeep
reverb_wikipedia_1000_2_source_of_esp_p_pc_keep
reverb_wikipedia_1000_2_source_of_esp_p_pc_nokeep
reverb_wikipedia_1000_2_source_of_esp_p_rel_keep
reverb_wikipedia_1000_2_source_of_esp_p_rel_nokeep

Hi, thank you for sharing the great implementation of Espresso!!

I have tried to run Espresso, but I couldn't.

My problems is occurred when executing cmd3 in bootstrap_ex.sh.

TypeError: 'NoneType' object has no attribute '__getitem__'

I confirmed that r in matrix2pmi.py was None, but I didn't understand why this was None.

Did I forget running some required preprocessing scripts?

Regards,

Hi, my name is Christian. I'm a student of Sekolah Tinggi Teknik Surabaya University in Indonesia. We're currently doing a research about relation extraction using bootstrapping algorithm. We've tried your espresso, but we couldn't run it well. Our plan is implementing this tools for Bahasa Indonesia. Can I ask you how to run espresso.py for our implementation?

Best regards,
Christian NP

bootstrap_ex.sh: 13: bootstrap_ex.sh: source: not found
bootstrap_ex.sh: 15: bootstrap_ex.sh: function: not found
.
.
.
2017-08-23 08:26:55,539 [INFO/Espresso:promotes] initializing mongodb connection: done
max_pmi:Traceback (most recent call last):
File "/home/admin_ps/data/Web-ka/tools/espresso.py", line 128, in
main()
File "/home/admin_ps/data/Web-ka/tools/espresso.py", line 124, in main
options.start)
File "/home/admin_ps/data/Web-ka/tools/espresso.py", line 87, in init
seeds, n, keep, reset, scorer, it
File "/home/admin_ps/data/Web-ka/tools/bootstrapper.py", line 28, in init
self.init_connection()
File "/home/admin_ps/data/Web-ka/tools/bootstrapper.py", line 53, in init_connection
self.db, self.matrix, self.boot_i, self.boot_p, self.logger
File "/home/admin_ps/data/Web-ka/tools/scorers.py", line 90, in init
self.max_pmi = self.pmi.max_pmi()
File "/home/admin_ps/data/Web-ka/tools/matrix2pmi.py", line 270, in max_pmi
print >>sys.stderr, 'max_pmi:', r['value']
TypeError: 'NoneType' object has no attribute 'getitem'
bootstrap_ex.sh: 30: bootstrap_ex.sh: Syntax error: "}" unexpected

Hi,

First, thanks for open sourcing this library!

I would like to get a better understanding of the Espresso algorithm. I would like to run your code against arbitrary text to see what happens. I don't understand and can't find how to turn raw text into something that I can feed to this library.

Could you explain what kind of preprocessing this library requires or how raw text needs to be fed to the cli commands?

Thanks a lot in advance,
Jeroen