Method and arrangements for analysing multimedia content

• US 8,898,689 B2
• Filed: 07/08/2009
• Issued: 11/25/2014
• Est. Priority Date: 07/08/2009
• Status: Active Grant

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: Georgakis, Apostolos
• Primary Examiner(s): Dubasky, Gigi L
• Application Number:
  US13/382,559
• Publication Number:
  US 20120124607A1
• Time in Patent Office: 1,966 Days
• Field of Search: 725 9- 14, 725 19- 20, 725/22
• US Class Current: 725/20
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: H04H 60/45 for identifying users
• CPC Class Codes: H04H 60/66 for using the result on distributors' side
• CPC Class Codes: H04N 21/23424 involving splicing one content stream with another content stream, e.g. for inserting or substituting an advertisement
• CPC Class Codes: H04N 21/251 Learning process for intelligent management, e.g. learning user preferences for recommending movies details of learning user preferences for the retrieval of video data in a video database G06F16/739; computer systems using learning methods G06N3/08
• CPC Class Codes: H04N 21/25883 being end-user demographical data, e.g. age, family status or address arrangements for identifying locations of users in broadcast systems H04H60/52
• CPC Class Codes: H04N 21/25891 being end-user preferences retrieval of video data in a video database based on user preferences G06F16/739; arrangements for recognizing users' preferences H04H60/46; user profiles in network data switching protocols H04L67/306; processing of user preferences or user profiles in wireless networks H04W8/18
• CPC Class Codes: H04N 21/2668 Creating a channel for a dedicated end-user group, e.g. insertion of targeted commercials based on end-user profiles information retrieval from the Internet by querying with filtering and personalisation G06F16/9535; arrangements for replacing or switching information during the broadcast H04H20/10; push services over packet-switching network H04L12/1859; adaptation of message content in packet-switching networks H04L51/063
• CPC Class Codes: H04N 21/4663 involving probabilistic networks, e.g. Bayesian networks

Abstract

A method and arrangements at a multimedia distribution network for estimating a present user constellation of a household are suggested. A sequence of textual descriptors that are associated with a multimedia sequence that is presently provided to the household is monitored. By acquiring a set of parameters of a statistical model that is based on textual descriptors that have previously been provided to the household, and by estimating a present user constellation by correlating the sequence of textual descriptors with the parameters of the statistical model, the estimated information may be used for customization of content that is provided to the household.

Claims

1. A method at a multimedia distribution network of estimating a present user constellation of a household, the method comprising:

monitoring a sequence of textual descriptors that are associated with a multimedia sequence that is being distributed to the household, wherein the sequence of textual descriptors is determined from subtitle information and is distributed to the household in a stream along with a stream of associated multimedia content;

acquiring, from a storing unit, a set of probabilistic parameters of a statistical model that is based on a plurality of sequences of textual descriptors that have previously been provided to the household; and

estimating a present user constellation by correlating the monitored sequence of textual descriptors being distributed with the probabilistic parameters of the statistical model,

wherein the statistical model is determined by executing the following steps;

acquiring the plurality of sequences of textual descriptors that have previously been provided, monitored and stored, each sequence of textual descriptors being associated with and determined from subtitle information of a multimedia sequence that has previously been provided to the household and representing a sequence of observations;

performing a statistical analysis on the acquired plurality of sequences of textual descriptors that have previously been provided, such that the statistical model, defined by a finite number, N, of most likely user constellations representing a sequence of hidden states, and the set of probabilistic parameters associated with state transitions of the most likely user constellations, is generated; and

storing the probabilistic parameters that define the statistical model.

2. The method according to claim 1, wherein the estimating step comprises:

estimating the present user constellation by executing on the fly Viterbi decoding on the stream of the sequence of textual descriptors being distributed and the set of probabilistic parameters.

3. The method according to claim 1, further comprising:

acquiring user preference information on the estimated present user constellation; and

providing the user preference information to a content provider, thereby enabling the content provider to selectively customize multimedia content provided to the household on the basis of the provided user preference information.

4. The method according to claim 3, wherein the customized multimedia content comprises any of:

video, audio, TV programs, or advertisements.

5. The method according to claim 1, wherein the statistical analysis is further based on additional, user specific information, comprising at least one of:

possible number of users at the household, socioeconomic information or user preferences associated with the household.

6. The method according to claim 1, wherein the performing step is executed by using a Baum-Welch algorithm on the acquired plurality of sequences of textual descriptors that have previously been provided.

7. The method according to claim 1, wherein the statistical model is generated by a Dynamic Bayesian Network.

8. The method according to claim 7, wherein the Dynamic Bayesian Network is a Hidden Markov Model.

9. The method according to claim 7, wherein the Dynamic Bayesian Network is a Multi-Resolution Hidden Markov model.

10. The method according to claim 7, wherein the Dynamic Bayesian Network is a Hidden Semi-Markov Model.

11. The method according to claim 1, wherein the monitoring step comprises monitoring the sequence of textual descriptors being distributed from metadata, subtitle information or tag information, that is being distributed parallel to its associated multimedia sequence.

12. The method according to claim 1, wherein the distributed multimedia sequence comprises an audio or a video stream.

13. An arrangement in a multimedia distribution network for estimating a present user constellation of a household, the arrangement comprising:

a monitor adapted to monitor a sequence of textual descriptors that is associated with a multimedia sequence that is being distributed to the household, wherein the sequence of textual descriptors is determined from subtitle information and is distributed to the household in a stream along with a stream of associated multimedia content; and

a correlator adapted to acquire a set of probabilistic parameters of a statistical model from a storing unit, the statistical model being based on a plurality of sequences of textual descriptors that have previously been provided to the household, and to estimate a present user constellation by correlating the monitored sequence of textual descriptors being distributed with the probabilistic parameters of the statistical model,

wherein the statistical model is determined by executing the following steps;

acquiring the plurality of sequences of textual descriptors that have previously been provided, monitored and stored, each sequence of textual descriptors being associated with and determined from subtitle information of a multimedia sequence that has previously been provided to the household and representing a sequence of observations;

performing a statistical analysis on the acquired plurality of sequences of textual descriptors that have previously been provided, such that the statistical model, defined by a finite number, N, of most likely user constellations representing a sequence of hidden states, and the set of probabilistic parameters associated with state transitions of the most likely user constellations, is generated; and

storing the probabilistic parameters that define the statistical model.

14. The arrangement of claim 13, wherein the correlator is further adapted to estimate a most probable present user constellation by executing on the fly Viterbi decoding on the sequence of textual descriptors being distributed and the acquired set of probabilistic parameters.

15. The arrangement of claim 13, further comprising:

a communication unit adapted to acquire user preference information on the present user constellation and to provide the user preference information to a content provider, thereby enabling the content provider to customize multimedia content provided to the household on the basis of the provided user preference information.

16. multimedia distribution network, comprising an arrangement according to claim 13.