Flickr Interestingness Algorithm

Interestingness is a media (images, videos, etc) ranking algorithm to provide as an additional metric for search results. The algorithm is based upon an algorithm created by the Flickr team, and is further enhanced by metadata knowledge resulted from the Del.icio.us development. Yahoo acquired both companies before the patent filing, and as such the invention is assigned to Yahoo.

The algorithm was first unveiled publicly on Flickr on 2005-08-01.1

See also Algorithm, Google, Search, Search Algorithm, Yahoo.

Factors related to the algorithm

SML.SML: Interestingness = f(views, faves, comments, tags, time, user, network relationships);

The ranking is based on one or more of the following factors:

  • the quantity of user-entered metadata2 (i.e. tags)
  • the number of users who have assigned metadata3
  • the number of favorites assigned to the photo4
  • relationship between the person who uploaded the photo and the people who are commenting5
  • access patterns related to the media object6 (i.e. where the clickthroughs are coming from; who comments on it and when)7
  • a lapse of time related to the media object8 (i.e. velocity of metadata acquisition)

It is important to note that the ranking fluctuates over time as additional activities accrued on the media objects.

SML Analysis

Note: the first draft of SML Analysis is posted under Flickr Analytics, before SML was aware of the US Patent Filing.

  • Background information
    • Text search - status - 2007: While text search algorithms are largely explored, search companies have by and large struggled to create image search algorithms that would yield satisfactory results.
    • Image Search - status - 2007: Google tackled early image search engines by using the filename of the images in addition to the context (surrounding text) by which these images appear.

United States Patents

Applications

  • 2006-10-26: Interestingness ranking of media objects (Patent ID 20060242139)
    • Abstract: Media objects, such as images or soundtracks, may be ranked according to a new class of metrics known as "interestingness." These rankings may be based at least in part on the quantity of user-entered metadata concerning the media object, the number of users who have assigned metadata to the media object, access patterns related to the media object, and/or a lapse of time related to the media object.
    • Inventors
    • Assignee: Yahoo
    • Filed: 2006-02-08
  • 2006-10-26: Media object metadata association and ranking (Patent ID 20060242178)
    • Abstract: Metadata may be associated with media objects by providing media objects for display, and accepting input concerning the media objects, where the input may include at least two different types of metadata. For example, metadata may be in the form of tags, comments, annotations or favorites. The media objects may be searched according to metadata, and ranked in a variety of ways.
    • Inventors
    • Assignee: Yahoo
    • Filed: 2006-02-08

SML Wiki: Algorithm

Algorithm
FoldUnfold Table of Contents SML Blogs: Algorithm Del.icio.us Popular Latest Flickr Interesting Backlinks Social SML Blogs: Algorithm Del.icio.us Popular http://del.icio.us/popular/algorithm...

Eric Fabricant
Search Ask.com: Eric Fabricant Clusty: Eric Fabricant Google: Eric Fabricant Live Search: Eric Fabricant SML Search: Eric Fabricant Yahoo Search: Eric Fabricant FoldUnfold Table of Contents ...

Flickr Analytics
Update: Because of newly discovered US Patent Filing of the Interestingness algorithm, attention has been shifted towards the analysis of this algorithm. Objective Analyze the Flickr Interestingness...

Image Resize
When scaling down images, apply small doses of USM afterwards to prevent blurriness. FoldUnfold Table of Contents Algorithms Down-sampling examples Image Resizing for Web and Email ImageMagick...

Interestingness
Interestingness is a media (images, videos, etc) ranking algorithm to provide as an additional metric for search results. The algorithm is based upon an algorithm created by the Flickr team, and is...

Popularity Algorithm
Variables at play Number of votes Record Age Trustworthiness of user who submitted vote Trustworthiness of user who voted Velocity of vote gain over time1 One-mindedness penalized2 Also See ...

Search Algorithm
Algorithm Search engines can utilize GMAT score honing technique to tweak relevant results. Keyword density (SEO tip from Simon Metz) United States Patent Applications Flickr Interestingness...

Steering Behaviors
Libraries AS3 Looking for one. Flickr Groups: Generative Visuals: iso (Java to AS3 porting tools) OR (AS3 Steering Behavior Lib) C++ MetaAgent OpenSteer Java SteeringBehaviors.de / Authors:...

Search

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License