Content recommendation apparatus and method using tag cloud

• US 20090132526A1
• Filed: 11/19/2008
• Published: 05/21/2009
• Est. Priority Date: 11/19/2007
• Status: Active Grant

• Current Assignee: Seoul National University Industry Foundation (Seoul National University), Synergy Innovation Co., Ltd.
• Original Assignee: Seoul National University Industry Foundation (Seoul National University), Synergy Innovation Co., Ltd.
• Inventors: Park, Jong-Hun
• Granted Patent: US 8,209,337 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually

Abstract

Content recommendation apparatus and methods using a tag cloud provide a content recommendation service via a network. The apparatus includes a content tag cloud generating module configured to generate a content tag cloud by analyzing a tag assigned to each content and accumulating frequencies per tag of each content. The apparatus also includes a user tag cloud generating module configured to generate a user tag cloud by accumulating frequencies per tag of contents used by a user. The apparatus further includes a similarity computing module and a recommending module. The similarity computing module is configured to compute a similarity between users using the user tag cloud, and the recommending module is configured to recommend content by computing a probability that a target user will use a specific content based on the computed similarity between users.

Claims

1. Content recommendation apparatus operative to provide a content recommendation service via a network, the apparatus comprising:

a content tag cloud generating module configured to generate a content tag cloud by analyzing a tag assigned to each content and accumulating frequencies per tag of each content;

a user tag cloud generating module configured to generate a user tag cloud by accumulating frequencies per tag of contents used by a user;

a similarity computing module configured to compute a similarity between users using the user tag cloud; and

a recommending module configured to recommend content by computing a probability that a target user will use a specific content based on the computed similarity between users.

2. The content recommendation apparatus of claim 1 wherein, in the accumulation of frequencies per tag of contents used by a user, the user tag cloud generating module normalizes a frequency per tag of a content used by the user and accumulates the normalized frequency per tag.

3. The content recommendation apparatus of claim 2 wherein the recommending module determines a user neighborhood similar to a target user and recommends a content based on a similarity between the target user and a user in the determined user neighborhood.

4. The content recommendation apparatus of claim 3 wherein the recommending module determines a user neighborhood based on users who have used at least two contents in a set of contents used by the target user.

5. The content recommendation apparatus of claim 3 wherein the recommending module determines N users from the user neighborhood most similar to the target user based on the similarity between users computed by the similarity computing module.

6. The content recommendation apparatus of claim 1 wherein the similarity computing module computes a cosine value of an angle between two vectors of tag clouds of two users.

7. The content recommendation apparatus of claim 1 further comprising a second recommending module configured to recommend a package of items or a web page having a large amount of content recommended by the recommending module.

8. A content recommendation method for providing a content recommendation service via a network;

the method comprising;

generating a content tag cloud by analyzing a tag assigned to each content and accumulating frequencies per tag of each content;

generating a user tag cloud by accumulating frequencies per tag of contents used by a user;

computing a similarity between users using the user tag cloud; and

recommending content by computing a probability that a target user will use a specific content based on the computed similarity between users.

9. The content recommendation method of claim 8 wherein, in accumulation of frequencies per tag of contents used by a user, generating the user tag cloud normalizes a frequency per tag of a content used by the user and accumulates the normalized frequency per tag.

10. The content recommendation method of claim 9 wherein the recommending a content comprises:

determining a user neighborhood similar to a target user; and

recommending a content based on a similarity between the target user and a user in the determined user neighborhood.

11. The content recommendation method of claim 10 wherein determining the similar user neighborhood comprises determining a neighborhood of users who have used at least two contents in a set of contents used by the target user.

12. The content recommendation method of claim 10 wherein determining the similar user neighborhood comprises determining N users from the user neighborhood most similar to the target user based on a similarity between users computed in the computing a similarity step.

13. The content recommendation method of claim 8 wherein computing the similarity comprises computing a cosine value of an angle between two vectors of tag clouds of two users.

14. The content recommendation method of claim 8, further comprising recommending a package of items or a web page having a large amount of content recommended in the recommending step.

15. A computer readable storage medium having a computer program stored thereon for implementing a content recommendation function using a tag cloud in a computer with a processor, the function comprising:

generating a content tag cloud by analyzing a tag assigned to each content and accumulating frequencies per tag of each content;

generating a user tag cloud by accumulating frequencies per tag of contents used by a user;

computing a similarity between users using the user tag cloud; and

recommending a content by computing a probability that a target user will use a specific content based on the computed similarity between users.