User:Natasa Siencnik/workshop1

Time for Revolution

World Clock Installation measuring Political Debate on Twitter

Sketches and Ideas

Project Description

Inspired by both the current political struggles in North Africa and the supposed lack of public participation in Western societies, the installation of parallel clocks is monitoring the political activity according to Twitter in a selection of countries. Conventional world clocks are typically found in loci such as hotel lobbies or stock market halls, representing stand-alone heterotopias, yet displaying their undeniable dependencies and power relations. But instead of progressing in time simultaneously, the incoming tweets about revolution are the engine for driving the clock hand. While time is running faster in some countries, others seem to be in a static mode, providing both stability and stagnancy. Presented simultaneous and equal in their physical manifestation, some clocks indicate it is time for change in an infinite loop.

Data Flow

1 — Searching Twitter for keywords related to political debate for a selection of countries
2 — Entries related to each country are the engine for moving the clock hand
3 — Displaying current political time for each country

Scripting Progress

Python : Script 2011-02-15

#!//usr/bin/python

import urllib2, json, sys
#from apikey import apikey

#api = twitter.Api(username=username, password=password)
url = 'http://search.twitter.com/search.json?q=%23revolution'
f = urllib2.urlopen(url)
data = json.load(f)
 
for r in data['results']:
#    print r['created_at'] + r['from_user'] + ' : "' + r['text'] + '"'
    sys.stdout.write r['created_at'] + r['from_user'] + ' : "' + r['text'] + '"'

Python : Script 2011-03-10

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import os, urllib2, json, sys, codecs

enc = codecs.getencoder("ascii")

def twittersearch(tag): 
    url = enc(u'http://search.twitter.com/search.json?q=%%23%s' % tag, "xmlcharrefreplace")[0]
    print("Query-URL: " + url)
    f = urllib2.urlopen(url)
    data = json.load(f)

    out = []
 
    for entry in data['results']:
        d = {}
        for key in ('created_at', 'id', 'text', 'iso_language_code') :
            d[key] = entry[key]
        out.append(d)
 
    return out

list_of_languages = [
u"Rewolusie",
u"Revolucion",
#u"الثورة",
#u"Հեղափոխություն",
u"Inqilab",
u"Iraultza",
#u"Рэвалюцыі",
#u"Революция",
u"Revolució",
#u"革命",
u"Revolucija",
u"Revoluce",
u"Revolution",
u"Revolutie",
u"Revolutsioon",
u"Rebolusyon",
u"Vallankumous",
u"Révolution",
u"Revolución",
#u"რევოლუცია",
#u"Επανάσταση",
u"Revolisyon",
#u"מהפכה",
#u"क्रांति",
u"Forradalom",
u"Revolusi",
u"Réabhlóid",
u"Rivoluzione",
#u"革命",
#u"רעוואָלוציע",
#u"혁명",
u"Revolūcija",
u"Revoliucija",
#u"Револуција",
u"Rivoluzzjoni",
#u"انقلاب",
u"Rewolucji",
u"Revolução",
u"Revoluţie",
#u"Революции",
#u"Револуција",
u"Revolúcia",
u"Mapinduzi",
#u"การปฏิวัติ",
u"Devrim",
#u"Революції",
u"Cách mạng",
u"Chwyldro",
]
 
out = []
 
for tag in list_of_languages:
    out.extend(twittersearch(tag))

ids = []

for entry in out:
    if entry['id'] in ids:
        print('non-unique id found\n')
    else:
        ids.append(entry['id'])

    print("%(created_at)s - %(id)d : %(text)s\nIn %(iso_language_code)s" % entry)	
 
sys.exit()

Python : Script 2011-03-24

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import os, urllib2, json, sys, codecs, datetime

enc = codecs.getencoder("ascii")

def twittersearch(query): 
    url = enc(u'http://search.twitter.com/search.json?q=%s' % query, "xmlcharrefreplace")[0]
    print("Query-URL: " + url)
    f = urllib2.urlopen(url)
    data = json.load(f)

    out = []
 
    for entry in data['results']:
        d = {}
        for key in ('created_at', 'id', 'text', 'iso_language_code') :
            d[key] = entry[key]
        out.append(d)
 
    return out

list_of_queries = [
    {"country":"Egypt","tags": ["revolution"]},
    {"country":"Libya","tags": ["revolution"]}
]
 
out = {}
 
for qd in list_of_queries:
    out[qd["country"]] = []
    for tag in qd["tags"]:
        query = u"%%23%s+%%23%s" % (tag, qd["country"])
        out[qd["country"]].extend(twittersearch(query))

for (country, entries) in out.items():
    ids = []
    print(country)
    for entry in entries:
        if entry['id'] in ids:
            print('non-unique id found\n')
        else:
            ids.append(entry['id'])

        print("%(created_at)s - %(id)d : %(text)s\nIn %(iso_language_code)s" % entry)	
     
sys.exit()

Python : Compare Date and Time

import datetime

In [4]: dt = datetime.datetime.strptime("24 Mar 2011 10:48:33", "%d %b %Y %H:%M:%S")

In [5]: dt
Out[5]: datetime.datetime(2011, 3, 24, 10, 48, 33)

In [6]: dt2 = datetime.datetime.strptime("24 Mar 2011 10:48:35", "%d %b %Y %H:%M:%S")

In [7]: dt2
Out[7]: datetime.datetime(2011, 3, 24, 10, 48, 35)

In [8]: dt2 > dt
Out[8]: True

In [9]: dt2 == dt
Out[9]: False

Python : Script 2011-03-28

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import os, urllib2, json, sys, codecs, datetime, MySQLdb

enc = codecs.getencoder("ascii")
search = "revolution"

def twittersearch(query): 
    url = enc(u'http://search.twitter.com/search.json?q=%s' % query, "xmlcharrefreplace")[0]
    print("Query-URL: " + url)
    f = urllib2.urlopen(url)
    data = json.load(f)

    out = []
 
    for entry in data['results']:
        d = {}
        for key in ('created_at', 'id', 'text', 'iso_language_code') :
            d[key] = entry[key]
        out.append(d)
 
    return out
    
def getDateTime(timestring):
    #return datetime suitable for mysql
    return 0

conn = MySQLdb.connect (host = "localhost", user = "natasa", passwd = "********", db = "revolution")

list_of_queries = [
    {"country":"Egypt","tags": [search]},
    {"country":"Libya","tags": [search]}
]

out = {}
 
for qd in list_of_queries:
    out[qd["country"]] = []
    for tag in qd["tags"]:
        query = u"%%23%s+%%23%s" % (tag, qd["country"])
        out[qd["country"]].extend(twittersearch(query))

for (country, entries) in out.items():
    ids = []
    print(country)
    for entry in entries:
        if entry['id'] in ids:
            print('non-unique id found\n')
        else:
            ids.append(entry['id'])

        print("%(created_at)s - %(id)d : %(text)s\nIn %(iso_language_code)s" % entry)
        
        time = 0;#getDateTime(created_at)
        
        try: 
            query = "insert into tweets (tweet_id, tweet_time, tweet_search, tweet_text) values('%s', %d, '%s', '%s')" % (entry['id'], time, search, entry['text'])
            print('query %s' % query)
            cursor = conn.cursor ()
            cursor.execute (query)
            #update tags/tweet table
            #update current time country
        
        except ValueError: 
            print "tweet already in database"
        
conn.close()     
sys.exit()

Python : Script 2011-04-05

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import os, urllib2, json, sys, codecs, datetime, time

#url = BIG_URL+"q="+search
#data = json.load(urllib2.urlopen().read(url))

search = codecs.open("twitterFeed_json", encoding="utf-8")
data = json.load(search)

tweetTimes = []

import dateutil, dateutil.parser

#for d in data["results"]:
#    tweetTimes.append((d["created_at"], d["text"]))

for d in reversed(data["results"]):
     createdTime = dateutil.parser.parse(d["created_at"])
     tweetTimes.append((createdTime, d["text"]))

# Perform the Tweets!
for i, (createdTime, text) in enumerate(tweetTimes):
	print text
	print
	if i+1 < len(tweetTimes):
		(nextTime, nextText) = tweetTimes[i+1] 
		print createdTime, nextTime
		diff = nextTime - createdTime
		diffsecs = diff.seconds + (diff.days * 60 * 60 * 24)
		time.sleep(diffsecs * 0.0001 )
#		os.system("mplayer foo.wav")

Database

Here are my first attempts in creating a SQL database.

List of Countries

LY - Libya - Tripoli
EG - Egypt - Cairo
IQ - Iraq - Baghdad
RU - Russia - Moskow
JP - Japan - Tokyo
IN - India - New Delhi
CN - China - Beijing
KP - North Korea - Pyongyang
IL - Israel - Jerusalem
PS - Palestine - Ramallah
US - United States - New York
UK - United Kingdom - London
NL - Netherlands - Amsterdam

List of Keywords

Biopolitics
Capitalism
Capitalist
Communism
Democracy
Democratic
Dictator
Election
Free
Freedom
Government
Governmentality
Ideology
Left-Wing
Liberalism
Lobby
Nomination
Nominee
Police
Political
Politics
President
Reactionary
Rebells
Revolution
Revolutionary
Right-Wing
Election

List of Languages

Link: http://www.explore2india.com/worldcapital.htm Link: http://en.wikipedia.org/wiki/List_of_countries_and_capitals_in_native_languages

Afriaans: Rewolusie
Albanian: Revolucion
Arabic: الثورة
Armenian: Հեղափոխություն
Azerbaijjani: Inqilab
Basque: Iraultza
Belarusian: Рэвалюцыі
Bulgarian: Революция
Catalan: Revolució
Chinese: 革命
Croatian: Revolucija
Czech: Revoluce
Danish: Revolution
Dutch: Revolutie
English: Revolution
Estonian: Revolutsioon
Filipino: Rebolusyon
Finnish: Vallankumous
French: Révolution
Galician: Revolución
Georgian: რევოლუცია
German: Revolution
GreeK: Επανάσταση
Haitian: Revolisyon
Hebrew: מהפכה
Hindi: क्रांति
Hungarian: Forradalom
Icelanding: Revolution
Indonasian: Revolusi
Irish: Réabhlóid
Italian: Rivoluzione
Japanese: 革命
Jiddish: רעוואָלוציע
Korean: 혁명
Latvian: Revolūcija
Lithuanian: Revoliucija
Macedonian: Револуција
Malay: Revolusi
Maltese: Rivoluzzjoni
Norvegian: Revolution
Persian: انقلاب
Polish: Rewolucji
Portuguese: Revolução
Romanian: Revoluţie
Russian: Революции
Serbian: Револуција
Slovak: Revolúcia
Slovenian: Revolucija
Spanish: Revolución
Swahili: Mapinduzi
Swedish: Revolution
Tagalog: Rebolusyon
Thai: การปฏิวัติ
Turkish: Devrim
Ukrainian: Революції
Urdu: انقلاب
Vietnamese: Cách mạng
Welsh: Chwyldro

Questions and Tasks

1 - What are the next steps codewise
2 - Talk to Stock about building a physical installation
3 - Buy clocks at HEMA

Thanks to Stock and Evo for the technical support and Stefan for his helpful input and feedback.