Selection of associated content for content items

• US 8,332,409 B2
• Filed: 08/24/2009
• Issued: 12/11/2012
• Est. Priority Date: 09/19/2008
• Status: Active Grant

• Current Assignee: Google Technology Holdings LLC (Alphabet Inc.)
• Original Assignee: Motorola Mobility LLC (Lenovo Group Ltd.)
• Inventors: Waddington, Simon, Menemenis, Fotis, Papadopoulos, Symeon, Bratu, Ben M., Kompatsiaris, Ioannis
• Primary Examiner(s): THAI, HANH B
• Application Number:
  US13/060,288
• Publication Number:
  US 20110179084A1
• Time in Patent Office: 1,205 Days
• Field of Search: 707736-798, 704/9
• US Class Current: 707/737
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A content device may select associated content, such as adverts, for a user selected content item based on textual characterizing data for the associated content and the user selected content item. A term set characterizing the user selected content item is expanded using semantic graphs and similarity values between the expanded term set and term sets describing associated content is calculated. A specific associated content item is then selected based on the similarity values. The semantic graph based term set expansion may allow improved accuracy in selecting appropriate associated content while providing a process that is suitable for resource constrained scenarios. In particular, communication resource, memory resource, and computational resource usage may be kept low.

Claims

1. A method of selecting a content item, the method comprising:

providing a plurality of associated term sets, each associated term set of the plurality of associated term sets comprising a set of terms characterizing a user selected content item of a group of provided content items from a content source;

providing a set of semantic graphs, each semantic graph of the set of semantic graphs comprising nodes corresponding to terms and edges corresponding to a relationship between terms of nodes connected by the edge;

providing a search term set comprising terms characterizing a first content item;

expanding the search term set in response to a first semantic graph of the set of semantic graphs to generate an expanded set of terms;

determining similarity measures between the expanded set of terms and each of the plurality of associated term sets; and

selecting an associated content item comprising an advert for the first content item from the group of provided content items in response to the similarity measures.

2. The method of claim 1 wherein expanding the search term set comprises:

selecting a subset of nodes of the first semantic graph having terms corresponding to terms of the search term set;

selecting a set of neighboring nodes for nodes of the subset of nodes; and

including at least some terms of the neighboring nodes in the expanded set of terms.

3. The method of claim 2 wherein selecting the subset of nodes comprises further selecting the subset of nodes in response to a co-occurrence degree of terms of the search term set.

4. The method of claim 2 wherein selecting the subset of nodes comprises further selecting the subset of nodes in response to at least one of a node value and an edge value.

5. The method of claim 2 wherein selecting the set of neighboring nodes comprises selecting the set of neighboring nodes in response to at least one of a node value and an edge value for neighbor nodes.

6. The method of claim 2 further comprising determining edge connections in the first semantic graph between nodes belonging to a group comprising the subset of neighboring nodes and nodes corresponding to terms of the search term set;

and

selecting terms of the neighboring nodes to be included in the expanded set of terms in response to the edge connections.

7. The method of claim 1 further comprising:

determining a term similarity between each semantic graph of the set of semantic graphs and the search term set, the term similarity for a semantic graph reflecting a correlation between terms of the semantic graph and terms of the search term set; and

selecting the first semantic graph in response to the term similarities.

8. The method of claim 1 wherein each semantic graph of the set of semantic graphs is associated with a topic and the method further comprises:

determining a search topic associated with the search term set; and

selecting the first semantic graph in response to the search topic.

9. The method of claim 1 wherein the first content item is a text content item and the method further comprises generating the search term set to comprise text terms of the text content item.

10. A content device comprising:

an associated term processor for providing a plurality of associated term sets, each associated term set of the plurality of associated term sets comprising a set of terms characterizing a user selected content item of a group of provided content items from a content source;

a receiver for receiving a set of semantic graphs from a remote server, each semantic graph of the set of semantic graphs comprising nodes corresponding to terms and edges corresponding to a relationship between terms of nodes connected by the edge;

a search term processor for providing a search term set comprising terms characterizing a first content item;

an expansion processor for expanding the search term set in response to a first semantic graph of the set of semantic graphs to generate an expanded set of terms;

a similarity processor for determining similarity measures between the expanded set of terms and each of the plurality of associated term sets; and

a selection processor for selecting an associated content item comprising an advert for the first content item from the group of provided content items in response to the similarity measures.