GRAVITY R&D
BIG DATA IN ONLINE CLASSIFIEDS
Domonkos Tikk, CEO/CSO
23 May, 2014
What data is available in your application domain?
Page views
User data
Ad placements
Popular products
Number of visits
De...
What does BIG DATA mean for you?
Product details
10M item
meta-data
User behaviour
20M user
meta-data
1M items 10 paramete...
What does BIG DATA mean for you?
5/26/2014
Methods of collecting and distributing user data
COLLECT and REPORT aggregated
data of your visitors
USE a RECOMMENDATION ...
How can it be used for business purposes?
Insight into classified Big Data
Degreeofinsight
1st click 2nd click 3rd click 1...
How does personalization work?
𝐋 = (𝒖,π’Š)βˆˆπ‘»π’“π’‚π’Šπ’ 𝒓 𝒖,π’Š βˆ’ 𝒓 𝒖,π’Š
𝟐
+ 𝝀 𝑼 𝒖=𝟏
𝑺 𝑼
𝑷 𝒖
𝟐+𝝀 𝑰 π’Š=𝟏
𝑺 𝑰
π‘Έπ’Š
𝟐
Recommendation techniq...
1 4 3
4
4 4
4
2
1.4
-0.2
0.8
0.5
-1.3
-0.4 1.6
-0.1 0.5
0.3
1.2 -0.51.1 -0.4
1.2 0.9
0.4 -0.4
1.2 -0.3
1.3
-0.1
0.9
0.4
1....
1 4 3
4
4 4
4
2
1.5
-1.0
2.1
0.8
1.0
1.6 1.8
0.7 1.6
0.0
1.4 1.1
0.9 1.9
2.5 -0.3
P
Q
R
3.3 2.4
-0.5 3.5 1.5
1.14.9
What type of data can be used for recommendations?
COLLABORATIVE
FILTERING
CONTENT-BASED
FILTERING
CONTEXT
AWARENESS
SOCIA...
Personalized User Journeys – Understand your
users and exploit the potential in BIG DATA
β€’ Predicting not just the primary...
 More user action and better user experience
impact on your market position and revenue
 Generate from 3rd
additional pa...
Thank you for your attention!
DomonkosTikk, PhD
Founder, CEO, CSO
Email: domonkos.tikk@gravityrd.com
hu.linkedin.com/in/do...
Upcoming SlideShare
Loading in...5
×

Big Data in Online Classifieds

672

Published on

The slideshow was presented at ICMA Conference in Helsinki at the "How to Turn Big Data into Dollars" Workshop organized by Gravity R&D,
The presentation reviews the heterogeneity of data sources at classified media, shows the massive size of data available, and give some insights how to use those data for personalization in various scenarios.

Published in: Internet, Technology
0 Comments
2 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total Views
672
On Slideshare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
25
Comments
0
Likes
2
Embeds 0
No embeds

No notes for slide
  • Legyen apirosnak is felfutΓ‘sa
  • Big Data in Online Classifieds

    1. 1. GRAVITY R&D BIG DATA IN ONLINE CLASSIFIEDS Domonkos Tikk, CEO/CSO 23 May, 2014
    2. 2. What data is available in your application domain? Page views User data Ad placements Popular products Number of visits Device IP address Time of browsing Time spent on site User behavior Ad replies Featured ads Number of products Location ClickThrough Rates Purchase history
    3. 3. What does BIG DATA mean for you? Product details 10M item meta-data User behaviour 20M user meta-data 1M items 10 parameters per item x 2M unique visitors 10 parameters per visitor x Interactions popular categories geolocation User contextual data integration Item contextual data Catalogue extension
    4. 4. What does BIG DATA mean for you? 5/26/2014
    5. 5. Methods of collecting and distributing user data COLLECT and REPORT aggregated data of your visitors USE a RECOMMENDATION system TRACK each visitor individually
    6. 6. How can it be used for business purposes? Insight into classified Big Data Degreeofinsight 1st click 2nd click 3rd click 1 week 1 month 1 year Tracking & data collection Data analysis Adequate business response β€žTraditional” reactive marketing Real-time personalization Item-to-item reco Price range Context Device
    7. 7. How does personalization work? 𝐋 = (𝒖,π’Š)βˆˆπ‘»π’“π’‚π’Šπ’ 𝒓 𝒖,π’Š βˆ’ 𝒓 𝒖,π’Š 𝟐 + 𝝀 𝑼 𝒖=𝟏 𝑺 𝑼 𝑷 𝒖 𝟐+𝝀 𝑰 π’Š=𝟏 𝑺 𝑰 π‘Έπ’Š 𝟐 Recommendation techniques Content based filtering Collaborative filtering  Recommends products that are liked by users that have similar taste as the current user  Similarity between users is calculated using the transaction history of users  Domain independent Recommends additional products with similar properties
    8. 8. 1 4 3 4 4 4 4 2 1.4 -0.2 0.8 0.5 -1.3 -0.4 1.6 -0.1 0.5 0.3 1.2 -0.51.1 -0.4 1.2 0.9 0.4 -0.4 1.2 -0.3 1.3 -0.1 0.9 0.4 1.1 -0.2 1.5 0.0 1.1 0.8 -1.2 -0.3 1.2 0.9 1.6 0.11.5 0.0 0.5 -0.3 -1.1 -0.2 0.4 -0.20.5 -0.1 0.6 0.2 P Q R
    9. 9. 1 4 3 4 4 4 4 2 1.5 -1.0 2.1 0.8 1.0 1.6 1.8 0.7 1.6 0.0 1.4 1.1 0.9 1.9 2.5 -0.3 P Q R 3.3 2.4 -0.5 3.5 1.5 1.14.9
    10. 10. What type of data can be used for recommendations? COLLABORATIVE FILTERING CONTENT-BASED FILTERING CONTEXT AWARENESS SOCIAL RECOMMENDATIONS
    11. 11. Personalized User Journeys – Understand your users and exploit the potential in BIG DATA β€’ Predicting not just the primary, but the secondary, tertiary, etc. interests β€’ Apart from history and behavior, focusing on the current context Based on Interest Seasonality Ad Replies Holidays Searches Continuous Devices used Working hours Last activity peak Every 3 months, during weekends
    12. 12.  More user action and better user experience impact on your market position and revenue  Generate from 3rd additional party revenues placements  Optimize your marketing spending on ad networks by personalized banners and placements How can you monetize from recommendations?
    13. 13. Thank you for your attention! DomonkosTikk, PhD Founder, CEO, CSO Email: domonkos.tikk@gravityrd.com hu.linkedin.com/in/domonkostikk/ Q&A
    1. A particular slide catching your eye?

      Clipping is a handy way to collect important slides you want to go back to later.

    Γ—