City Mood Based on Twitter Conversation

Mood Parameter Research Based on Twitter Conversation

Research by:For:

Method of Analysis

•Track twitter conversation that indicates mood of user,

exclude Retweet

* From 9 cities in Indonesia

1. Jabodetabek

2. Bandung

3. Surabaya

4. Yogyakarta

5. Semarang

6. Malang

7. Denpasar

8. Palembang

9. Medan

• Research Period 05/12/2011 – 29/01/2012. Devide into

4 time interval a day

Period 1 = 00:00:00 – 06:00:00 (real 00:00:00 – 05:59:59)

Period 2 = 06:00:00 – 12:00:00 (real 06:00:00 – 11:59:59)

Period 3 = 12:00:00 – 18:00:00 (real 12:00:00 – 17:59:59)

Period 4 = 18:00:00 – 00:00:00 (real 18:00:00 – 23:59:59)

* Analyze qualilatively from the social media data.

Research by:

General Scheme

Note:

Example of Bad Mood : Bete, Sebel, Kesel, Kecewa, Marah, Gak Semangat, Gak hepi etc

Example of Good Mood : Seneng, Gembira, Hepi, Gak bete, Semangat, Bahagia etc

Twitter

Good Mood

Crawler

Facebook

Good Mood

Crawler

Twitter

Bad Mood

Crawler

Facebook

Bad Mood

Crawler

Injected

Bad Mood & Good Mood

Keywords

Good Mood Data

Storage

Bad Mood Data

Storage

Aggregator

APIfor

external access

XML

JSON

Exte

rnal

Enti

tie

s(M

izo

ne

Serv

er )

GET

GET

OR

Internal Entities

(MediaWave Server)

Data Crawl & Data Feed Scheme

Twitter

Good Mood

Crawler

Facebook

Good Mood

Crawler

Twitter

Bad Mood

Crawler

Facebook

Bad Mood

Crawler

Good Mood Data

Storage

Bad Mood Data

Storage

Aggregator

Twit

ter

and

Fac

eb

oo

kA

PI

1.

For

Twit

ter

usi

ng

Rea

ltim

eSt

ream

AP

I2

.Fo

r Fa

ceb

oo

ku

sin

g Se

arch

AP

I

to API for

external

access

City Detection

Method (Twitter)

Good Mood Data

Storage

Bad Mood Data

Storage

Aggregator

Extr

act

Cit

y

"id_str": "276483979",

"default_profile": false,

"follow_request_sent": null,

"verified": false,

"profile_link_color": "0084B4",

"location": "UT: -8.547824,115.172388",

"id": 276483979,

"utc_offset": null

In these period there are 2.856.382 conversations about bad mood and 2.293.067 people who have bad mood.

Otherwise there are 2.350.052 conversations about good mood and 1.770.030 people who have good mood

susah

sekolah

males

pagi

jelek

Palembang

sakit

telat

males

susah

lama

galau

tai

libur

sekolah

bolos

Top 10 Bad Mood

Palembang

ujian

beraktifitas

sukses

kemudahan

usek

cantik

membuat

semoga

senin

kelancaran

Top 10 Good Mood

selamat pagi

semoga

semangat

amin

cantik

Most bad mood day is Wednesday with 539.104 conversation (18,87%) from all bad mood

conversations), followed by Thursday with 508.699 conversation (17,81%) from all badmood conversations).

Most bad mood period is beetween 6 AM - 12 AM (36,43%), followed by 12 AM - 6 PM(35,44%) from all bad mood conversation

Top 10 bad mood words are Parah, Males, Cape, Bete, lemot, nungguin, lama, ga enak, Galau, Payah

Visualize bad mood relation

Social media has changed the communication betweenbrand and consumer. Nowadays communications are twoways communication and horizontally. The consumer'sperception over your product is determined by otherconsumer's experience sharing about the product, notdetermined by your promotion material and advertisinganymore. MediaWave is the first Social Media Monitoring &Analytics Platform in Indonesia. MediaWave helps you toobserve and measures the consumer perception upon yourbrand in Social Media.We provide you the most valuable insight directly fromconsumer voice

Find more about us at www.mediawave.biz

Contact us :

- [email protected] / @mediawave_id- [email protected] / @yoseazka- [email protected] / @erikpalupi- [email protected] / @dwiwahyono