def getPapersAbstract():
from collections import defaultdict
from dcclient import DataCenterClient
f = open("E:\\ids.txt")
f.next()
ids = []
import codecs
f_out = codecs.open("E:\\abstracts_1.txt","w",encoding="utf-8")
from bs4 import UnicodeDammit
for line in f:
x = line.split("\n")
ids.append(int(x[0]))
print len(ids)
c = DataCenterClient("tcp://10.1.1.211:32011")
for i in range(len(ids)/1000):
print "DUMP %s"%(i*1000)
x = c.getPublicationsById(ids[i*1000:(i+1)*1000])
id_set = set(ids)
count = 0
abs = {}
conf = {}
authors = {}
title = {}
year = {}
for p in x.publications:
abs[p.id] = p.abs.replace("\n"," ").replace("\t"," ")
conf[p.id] = p.jconf_name
authors[p.id] = ",".join([str(a) for a in p.author_ids])
title[p.id] = p.title
year[p.id] = p.year
for p in abs:
if len(abs[p]) > 2:
f_out.write("%s\n%s\n%s\n%s\n%s\n%s\n"%(p,year[p],conf[p],authors[p],title[p],UnicodeDammit(abs[p]).markup))
def getPapersCitation():
from collections import defaultdict
from dcclient import DataCenterClient
f = open("E:\\vis.txt")
f.next()
ids = []
for line in f:
x = line.split("\t")
ids.append(int(x[0]))
c = DataCenterClient("tcp://10.1.1.211:32011")
x = c.getPublicationsById(ids)
id_set = set(ids)
count = 0
citation = defaultdict(set)
for p in x.publications:
for y in p.cite_pubs:
if y in id_set:
print count
count += 1
citation[p.id].add(y)
for y in p.cited_by_pubs:
if y in id_set:
print count
count += 1
citation[y].add(p.id)
f_out = open("E:\\citation.txt", "w")
for p in citation:
for q in citation[p]:
f_out.write("%s\t%s\n" % (p, q))
def getPapersCitation():
from collections import defaultdict
from dcclient import DataCenterClient
f = open("E:\\vis.txt")
f.next()
ids = []
for line in f:
x = line.split("\t")
ids.append(int(x[0]))
c = DataCenterClient("tcp://10.1.1.211:32011")
x = c.getPublicationsById(ids)
id_set = set(ids)
count = 0
citation = defaultdict(set)
for p in x.publications:
for y in p.cite_pubs:
if y in id_set:
print count
count += 1
citation[p.id].add(y)
for y in p.cited_by_pubs:
if y in id_set:
print count
count += 1
citation[y].add(p.id)
f_out = open("E:\\citation.txt","w")
for p in citation:
for q in citation[p]:
f_out.write("%s\t%s\n"%(p,q))
def getPapersAbstract():
from collections import defaultdict
from dcclient import DataCenterClient
f = open("E:\\ids.txt")
f.next()
ids = []
import codecs
f_out = codecs.open("E:\\abstracts_1.txt", "w", encoding="utf-8")
from bs4 import UnicodeDammit
for line in f:
x = line.split("\n")
ids.append(int(x[0]))
print len(ids)
c = DataCenterClient("tcp://10.1.1.211:32011")
for i in range(len(ids) / 1000):
print "DUMP %s" % (i * 1000)
x = c.getPublicationsById(ids[i * 1000:(i + 1) * 1000])
id_set = set(ids)
count = 0
abs = {}
conf = {}
authors = {}
title = {}
year = {}
for p in x.publications:
abs[p.id] = p.abs.replace("\n", " ").replace("\t", " ")
conf[p.id] = p.jconf_name
authors[p.id] = ",".join([str(a) for a in p.author_ids])
title[p.id] = p.title
year[p.id] = p.year
for p in abs:
if len(abs[p]) > 2:
f_out.write("%s\n%s\n%s\n%s\n%s\n%s\n" %
(p, year[p], conf[p], authors[p], title[p],
UnicodeDammit(abs[p]).markup))
def getCitationNetwork():
import time
import datetime
from collections import defaultdict
c = DataCenterClient("tcp://10.1.1.211:32011")
x = c.searchPublications("deep learning")
data_fields = [
"id", "mid", "uid", "parent", "type", "t", "user_created_at",
"followers_count", "statuses_count", "friends_count", "username",
"text", "words", "verified", "emotion"
]
items = []
cite_pubs = []
key_terms = defaultdict(int)
year_terms = defaultdict(lambda: defaultdict(int))
for p in x.publications:
if p.year <= 1970:
continue
item, children, parents, kt = extractPublication(p)
if len(children) > 0:
items.append(item)
cite_pubs.extend(children)
cite_pubs.extend(parents)
for k in kt:
key_terms[k.lower()] += 1
year_terms[p.year][k.lower()] += 1
cite_pubs = list(set(cite_pubs))
x = c.getPublicationsById(cite_pubs)
for p in x.publications:
if p.year <= 1970:
continue
item, children, parents, kt = extractPublication(p)
if len(children) > 0 and len(children) > 0:
items.append(item)
cite_pubs.extend(children)
for k in kt:
key_terms[k.lower()] += 1
year_terms[p.year][k.lower()] += 1
sorted_key_terms = sorted(key_terms.items(),
key=lambda x: x[1],
reverse=True)
import json
dump = open("pubs_dump.json", "w")
d = json.dumps(items)
dump.write(d)
dump.close()
def getPapersAbstractYearConf():
from collections import defaultdict
from dcclient import DataCenterClient
import codecs
from bs4 import UnicodeDammit
import os
f = open("E:\\ids.txt")
f.next()
ids = []
for line in f:
x = line.split("\n")
ids.append(int(x[0]))
c = DataCenterClient("tcp://10.1.1.211:32011")
def createFile(year, conf):
if not os.path.exists(str(year)):
os.makedirs(str(year))
return codecs.open(os.path.join(str(year), conf),
"w",
encoding="utf-8")
#files = defaultdict(dict)
files = {}
for i in range(len(ids) / 10000):
print "DUMP %s" % (i * 10000)
x = c.getPublicationsById(ids[i * 10000:(i + 1) * 10000])
id_set = set(ids)
count = 0
abs = {}
conf = {}
title = {}
year = {}
for p in x.publications:
abs[p.id] = p.abs.replace("\n", " ").replace("\t", " ")
conf[p.id] = p.jconf_name #.replace("/"," ").replace("*"," ")
title[p.id] = p.title
year[p.id] = p.year
for p in abs:
if len(abs[p]) > 2 and len(conf[p]) > 1:
#if not files[year[p]].has_key(conf[p]):
if not files.has_key(year[p]):
files[year[p]] = codecs.open(str(year[p]),
"w",
encoding="utf-8")
#files[year[p]][conf[p]] = createFile(year[p], conf[p])
file = files[year[p]]
file.write(
"%s\n%s\n%s\n%s\n" %
(p, conf[p], title[p], UnicodeDammit(abs[p]).markup))
def getCitationNetwork():
import time
import datetime
from collections import defaultdict
c = DataCenterClient("tcp://10.1.1.211:32011")
x = c.searchPublications("deep learning")
data_fields = ["id", "mid", "uid",
"parent", "type", "t",
"user_created_at", "followers_count", "statuses_count",
"friends_count", "username", "text", "words", "verified", "emotion"];
items = []
cite_pubs = []
key_terms = defaultdict(int)
year_terms = defaultdict(lambda: defaultdict(int))
for p in x.publications:
if p.year <= 1970:
continue
item, children, parents, kt = extractPublication(p)
if len(children) > 0:
items.append(item)
cite_pubs.extend(children)
cite_pubs.extend(parents)
for k in kt:
key_terms[k.lower()] += 1
year_terms[p.year][k.lower()] += 1
cite_pubs = list(set(cite_pubs))
x = c.getPublicationsById(cite_pubs)
for p in x.publications:
if p.year <= 1970:
continue
item, children, parents, kt = extractPublication(p)
if len(children) > 0 and len(children) > 0:
items.append(item)
cite_pubs.extend(children)
for k in kt:
key_terms[k.lower()] += 1
year_terms[p.year][k.lower()] += 1
sorted_key_terms = sorted(key_terms.items(), key=lambda x: x[1], reverse=True)
import json
dump = open("pubs_dump.json", "w")
d = json.dumps(items)
dump.write(d)
dump.close()
def getPapersAbstractYearConf():
from collections import defaultdict
from dcclient import DataCenterClient
import codecs
from bs4 import UnicodeDammit
import os
f = open("E:\\ids.txt")
f.next()
ids = []
for line in f:
x = line.split("\n")
ids.append(int(x[0]))
c = DataCenterClient("tcp://10.1.1.211:32011")
def createFile(year, conf):
if not os.path.exists(str(year)):
os.makedirs(str(year))
return codecs.open(os.path.join(str(year), conf), "w", encoding="utf-8")
#files = defaultdict(dict)
files = {}
for i in range(len(ids) / 10000):
print "DUMP %s" % (i * 10000)
x = c.getPublicationsById(ids[i * 10000:(i + 1) * 10000])
id_set = set(ids)
count = 0
abs = {}
conf = {}
title = {}
year = {}
for p in x.publications:
abs[p.id] = p.abs.replace("\n", " ").replace("\t", " ")
conf[p.id] = p.jconf_name#.replace("/"," ").replace("*"," ")
title[p.id] = p.title
year[p.id] = p.year
for p in abs:
if len(abs[p]) > 2 and len(conf[p]) > 1:
#if not files[year[p]].has_key(conf[p]):
if not files.has_key(year[p]):
files[year[p]] = codecs.open(str(year[p]), "w", encoding="utf-8")
#files[year[p]][conf[p]] = createFile(year[p], conf[p])
file = files[year[p]]
file.write("%s\n%s\n%s\n%s\n" % (p, conf[p], title[p], UnicodeDammit(abs[p]).markup))