analyzing organization e-mails in near real time using hadoop ecosystem tools by miguel romero &...

14

Upload: big-data-spain

Post on 11-Feb-2017

196 views

Category:

Business


2 download

TRANSCRIPT

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

Analyzing organization e-mails in Near Real Time using

Hadoop Ecosystem tools

Big Data Spain 2015

Miguel Romero, Hadoop Architect (@donkelito)

Alberto de Santos, Data Scientist (@adesantossierra)

Analytics and Data Management, Enterprise Services

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

Un trabajador feliz se ausenta un 28,4% menos

Un trabajador feliz es un 22% más productivo que

uno infeliz

$100,000

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

Hacerme su

amigo

¿Redes sociales

?

¡Correo de

empresa!

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

To/FromTo/From

AsuntoAsunto

ContenidoContenido

AdjuntosAdjuntos

¿Qué datos tengo?

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

¿Qué grupos he encontrado?

Entender quiénes envían/reciben más emailsUsuarios más influyentesDetectar grupos de redes y entender sus característicasAnalizar comunidades: cómo se forman, cómo se agregan integrantesDetectar evangelizadoresDetectar expertosMezcla de comunidadesHombres bisagra

Agregación multiescalaNormalized CutsMax-flow min-cut algorithm

To/FromTo/From

AsuntoAsunto

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

¿De qué hablan en esas comunidades?

To/FromTo/From

AsuntoAsunto

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

¿De qué hablan en esas comunidades?

To/FromTo/From

AsuntoAsunto

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

¿De qué hablan en esas comunidades?

To/FromTo/From

AsuntoAsunto

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

Online vs offline methods

Detección de grupos

¿de qué se habla en cada grupo?

Perfiles más influyentes

Modelos clasificación

Detección de anomalías

Evolución de conversacionesEvolución de las

comunidades

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

© Copyright 2014 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice.

Muchas gracias

Big Data Spain 2015

Miguel Romero (@donkelito)

Alberto de Santos (@adesantossierra)

Analytics and Data Management, Enterprise Services