advertisement

Real Time Recommendation System using Kiji

100 %
0 %
advertisement
Information about Real Time Recommendation System using Kiji
Technology

Published on March 6, 2014

Author: daqingzhao

Source: slideshare.net

Description

The January Kiji User Group Meetup

Macys.com uses the Kiji environment and its portfolio of tools to power a real-time site recommendation platform that serves the right customers with the right recommendations at the right time, improving the relevance of each message and customer experience. In this talk, Macys.com team will present how Macys.com uses the Kiji tools for real-time marketing activities.
advertisement

Real Time Recommender System with Jan 22, 2014 Daqing Zhao, Director of Advanced Analytics Macy’s.com

Agenda  Big data analytics versus traditional BI  Macy’s Advanced Analytics Team  Our analytics projects  Example: site recommendations using Kiji  High level architecture  Kiji Schema table structure  Model deployment using Kiji  Key benefits of Kiji and WibiData team 1

Traditional BI process Knowledge Discovery Segmentation and Predictive Modeling Most companies Stay in this area Multidimensional Report Standard Report Schema definition, ETL into RDMS Baseline Consulting  Data can be accessed and analyzed only after ETL  Schema definition may not be optimal 2

Hadoop/NoSQL: paradigm shift Decisions Insights Models Decision Agent Segmentation and Predictive Modeling Multi dimensional Report Reports Standard Report Hive, Mahout, Cascading, Scalding, Kiji, … MapReduce Raw data Volume Velocity Variety Write Append Read Distributed storage Computation near data Hadoop, HBase, avro, …  We can access raw data and analyze using MapReduce  With pros and cons 3

Macy.com’s Advanced Analytics Group  We are at the frontiers of Big Data science: • Using Big Data technology • Machine learning and Statistical algorithms  We have predictive modeling, experimental design and data science teams  Our team members have very strong background in • Quantitative fields, math, stat, physics, bioinformatics, decision sciences, and cs • We collaborate with systems and IT teams internally as well as 3rd party vendors like WibiData, SAS Research, IBM Research…  We use a wide range of tools • Hadoop, SAS, R, Mahout, and others, as well as Kiji Models  We are data scientists with keen focus on domain problems 4

Customer acquisition and retention  Targeting the right message to the right customer at the right time • Build predictive models of purchase behavior and identify drivers  Site recommendation algorithms • Recommend products based on items that are added to bag for cross- and up-sell • We also look at market basket analysis • Most work is in batch mode, expanding slowly into real time  Rapid-prototyping and testing of algorithms and policies • All done in short development cycles  Output of the team’s work support other marketing teams to identify, and reach best customers • Search, display, social network, affiliates, retention, customer services, … 5

Some other projects  Data organization or data munging • • • • • Data collections, individual and event level, 360 degrees, … Segmentation of customers Customer value, revenue, costs Multiple channel attribution of marketing contacts Product attributes  Experimentation platform • Success of online marketing depends highly on testing, learning and optimization • Both for site layout as well as contents and recommendations  Forecast and optimization • Prediction, simulation, and search and optimize  Big data refinement and scalability • New data sources, more efficient ways of accessing data, and organizing and processing data 6

Example: similar and complementary products 7

Example: customer segmentation Demographic Socio-economic Behavioral Values and styles Channels Modality 8

Example: product social network Demographic Style Size Brand Price range Season 9

Example: site product recommendation  Customer Adds to Bag one or more products  We recommend in real time similar/complementary products • Based on product associations and customer profile  We use various machine learning algorithms • • • • • • Association rules Collaborative filtering Predictive modeling Business rules And others, … Models built offline  Real time data, real time model scoring and real time decision  Champion/challenger tests, models evolve quickly in time  Frequent model updates, add new data 10

Architecture Real Time Data access, Scoring Decisions Others data mining Kiji Express environment data mining Mahout environment data mining R environment SAS Environment products Kiji Model Kiji Kiji Scoring Scoring Kiji Kiji Rest Rest Kiji Kiji Rest Rest Hadoop HBase 11

Kiji Schema table structure Customer table entity id customer email metadata order Product table entity id product category metadata inventory Schema have column names and types, compared to bits stored in HBase Group column families are structured, while Map column families are flexible Accessible as collections from Kiji Express Scala code focuses on model and business logic Scalding underneath takes care of generating MapReduce jobs 12

Model Build and Deployment Model Model building Model building Model building Model building building Kiji Express Kiji Scoring Kiji PMML Kiji MR Deployment Kiji Schema HBase Hadoop Offline Kiji Modeling R, SAS, Mahout, … Real time data update Real time scoring Real time decisions 13

Key benefits of partnership with WibiData  Open source, Kiji suite, abstracted with focus in modeling • Kiji Schema, KijiMR, Kiji Model, Kiji Scoring, Kiji Express, Kiji REST • Allow quick development cycle  Package popular open source projects • Hadoop, HBase, Avro, Cascading, Scalding, Scala  Better organization • Create tables, query by field name, flexibility, …, more DB like than HBase  WibiData professional services team help develop, integrate, maintain, train in-house team, consult,… • Competence, knowledge • Support infrastructure, so that we can focus on the science  Real time model deployment environment and scalable • Interactive • In milliseconds 14

Acknowledgement  Macy’s teams  Analytics team: Kerem Tomak, Albert Zhai  Infrastructure team: Winslow Holmes, Rakesh Sharma, Cherry Peng  WibiData team  Professional Services team: Adam, Christophe, Renuka, Lynn 15

Add a comment

Related presentations

Presentación que realice en el Evento Nacional de Gobierno Abierto, realizado los ...

In this presentation we will describe our experience developing with a highly dyna...

Presentation to the LITA Forum 7th November 2014 Albuquerque, NM

Un recorrido por los cambios que nos generará el wearabletech en el futuro

Um paralelo entre as novidades & mercado em Wearable Computing e Tecnologias Assis...

Microsoft finally joins the smartwatch and fitness tracker game by introducing the...

Related pages

Building a Real-time, Personalized Recommendation System ...

Jon Natkins explains in this article how to create a personalized recommendation system fed with large amounts of real-time data using Kiji ...
Read more

Kiji Project · GitHub - GitHub · Where software is built

kiji. The Kiji project suite Updated Jun 18, 2015. ... and inspecting schemas on HBase using KijiSchema Updated Sep 26, 2014. Java 5 18 kiji ...
Read more

Recommender system - Wikipedia, the free encyclopedia

Recommender systems or recommendation systems ... The system generates recommendations using only information about ... Embedded systems; Real-time ...
Read more

kijiproject/kiji-bento · GitHub

kijiproject / kiji-bento. ... and server that supports the real-time per-row calculations on kiji ... new music recommendations. kiji-express ...
Read more

Real-time Recommendation Systems using Apache Storm ...

Real-time Recommendation Systems using Apache Storm. Home; ... Pranab Ghosh discussed the real time recommendations feature of Sifarish, ...
Read more

Real-time mobile recipe recommendation system using food ...

Real-time mobile recipe recommendation system using food ingredient recognition on ... We propose a real-time object recognition method for a ...
Read more

Real-time Recommendation Systems using Apache Storm ...

... Home » Amazon EC2 » Real-time Recommendation Systems using Apache Storm. ... Pranab Ghosh discussed the real time recommendations feature of Sifarish
Read more

Real-time Recommendation Systems using Apache Storm ...

Real-time Recommendation Systems using Apache Storm File; Brochure of Real-time Recommendation Systems: Please enter your contact information to receive ...
Read more