WO2008043143A1

WO2008043143A1 - Personalised content generation

Info

Publication number: WO2008043143A1
Application number: PCT/AU2007/001537
Authority: WO
Inventors: Scott Mcauley Thomson; Danny Gruber; Yon Aran Rhee
Original assignee: Qdc Technologies Pty Ltd
Priority date: 2006-10-11
Filing date: 2007-10-11
Publication date: 2008-04-17

Abstract

A personalised content generation system, including a content generation engine for generating personalised content for a user based on social network data of at least one social network related to the user. The system includes a rules engine for executing content selection rules to process profile data for a user and social data for parties of a social network including the user to generate a personalised content template for the user. The template includes references for accessing selected content. The system uses a render engine for parsing the template and accessing the selected content to generate personalised content, in particular video content, for the user.

Description

PERSONALISED CONTENT GENERATION

FIELD

The present invention relates to a personalised content generation system.

BACKGROUND

Systems have been developed for dynamically adjusting and delivering content to consumers. For example, Internet based server systems are able to access profile data for a user when the user selects a URL on a web site, and a cookie is available for the user, or the user has previously logged onto a server hosting the site. The response returned to the user can include content selected or adjusted based on the profile data.

Also, targeted advertising on the Internet can include an advertisement that is selectively placed on a webpage (e.g. of a newspaper) or within a rich HTML content email. The advertisement is customised each time the webpage is selected (i.e. loaded, or viewed) by a visitor. This customisation may include inserting different items of content based on data that is known by the webpage host, or that is provided by the visitor.

The content delivered and the changes that can be made are, however, limited in scope. This is primarily because the existing systems are only able to roughly determine a person's interests and their susceptibility to be interested in content that can be delivered.

It is desired to address the above or at least to provide a useful alternative.

SUMMARY

In accordance with the present invention, there is provided a personalised content generation system, including a content generation engine for generating personalised content for a user based on social network data of at least one social network related to the user.

The present invention also provides a personalised content generation process, performed by a computer system, including: accessing social network data of at least one social network related to a user; and processing the social network data to select content elements of personalised content for the user.

The present invention also provides a content generation process, including: accessing profile data for a visitor to a site and a contributor to said site; processing content selection rules to select content elements, said processing including comparing said visitor and contributor profile data; generating content with said content elements for said visitor; and serving said content for said visitor.

The present invention also provides a personalised content generation system including: a rules engine for executing content selection rules to process profile data for a user and social data for parties of a social network including said user to generate a personalised content template for the user, said template including references for accessing selected content; and a render engine for parsing the template and accessing the selected content to generate personalised content for the user.

DESCRIPTION OF THE DRAWINGS

Preferred embodiments of the invention are hereinafter described, by way of non-limiting example only, with reference to the accompanying drawings in which:

Figure 1 is a schematic of a personalised content generation system in use; Figure 2 is a schematic of a hardware architecture of the personalised content generation system;

Figure 3 is a flow chart of a process performed by the personalised content generation system and the arrangement in Figure 1 ;

Figure 4 is a flow chart of a content selection process performed by the personalised content generation system; and

Figure 5 is a social networking base selection process performed by the personalised content generation system.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

A personalised content generation system 102, as shown in Figures 1 and 2, includes a content generation engine 104 for generating personalised content 108, and an agent 106 for triggering the generation of the personalised content 108. The agent 106 monitors activity 110 between a client 112 and a server 114, which hosts a site 1 16. When the client 112 accesses the site 116, the agent 106 triggers the content generation engine 104 to generate and deliver the personalised content 108 to the client 112. The content generation engine 104 generates the personalised content 108 based on visitor profile data 1 18, which is profile data of a user 120 of the client 112. The content generation engine 104 also generates the personalised content 108 based on social network profile data 122 for one or more parties 124 that are associated with the user 120. The content generation engine 104 may also generate the personalised content 108 based on host profile data 126 of a host 128, or author, of a page 130 on the site 116, and social network profile data 132 for one or more parties 134 associated with the host 128. The host 128 is, for example, an entity who generated at least some of the content on the site 116, i.e. a contributor. The contributor may host the site 116, or may have merely posted a comment on the site 1 16, e.g. a comment on a weblog / blog.

A page 130 on the site 116 may include:

(1) page data, associated with the characteristics of the page 130 (e.g. date created, date modified, page access statistics, or other metadata);

(2) host profile data 126, associated with the profile of the host 128, representing owner/creator of the page (e.g. name, date of birth, gender, ethnicity, body type, height, page purpose for dating / friends / networking / friendship, address, occupation, diary topics, interests, music, films, television, books, heroes, marital status, sexual orientation, home town, religion, smoking habits, drinking habits, children, education, income, places of education, places of work, areas of interest for networking, favourite songs / music, generics etc);

(4) link data relating to any hyperlinks on the page 130;

(5) back link data relating to links from other pages that link to the page 130 (e.g. data from the PageRank system of Google, Inc.);

(6) social network association data, indicating other parties 134 who are related to the host 128 (e.g. in a list of friends, as parties who have contributed to the page 130 via comments, or parties who have included the host 128 in their list of friends on their page); and

(7) any metadata stored on content within the hosted page including but not limited to HTML metadata information.

When the agent 106 detects the activity 110 between the client 112 and the server 114, the agent accesses profile data 126 associated with the page 130, including the profile data listed above.

The agent 106 accesses visitor profile data 118 associated with the user 120, including data extracted from the page 130, data extracted from the client 112 (e.g. Internet cookies), and other data from Customer Relationship Management (CRM) databases or provided by third-party enterprises. This data may include: information on past transactions or interactions with personalised content 108, subscriptions to newspapers, magazines, podcasts or email newsletters, or membership of online forums.

The agent 106 also accesses visitor social network profile data 122 for parties 124 associated with the user 120, and host social network profile data 132 for parties 134 associated with the host 128.

The visitor profile data 1 18 may be maintained on the server 1 14, and/or as data on the client 112 (e.g. cookies). The visitor profile data 118 may include search records (e.g. lists of visited web pages), and data representing physical location, income, demographic, name, date of birth, gender, ethnicity, body type, height, page purpose (e.g. for dating, friends, networking, friendship etc), address, occupation, diary topics, interests, music, films, television, books, heroes, marital status, sexual orientation, home town, religion, smoking habits, drinking habits, children, education, places of education, places of work, areas of interest for networking, and favourite songs/music. The visitor profile data 1 18 may be accessed directly from the client 112 or the server 1 14 or generated as part of an interactive communication with the user 120 (via the client) 112.

The visitor social network profile data 122 includes data from web pages that are associated with the parties 124 (e.g. home pages of the visitor's friends). Each associated page provides equivalent data to that described above in relation to the host page 130. The visitor social network profile data 122 also includes metadata about groups to which the user 120 and at least one of the parties 124 belong, for example both may be contributors to an online skiing forum. Other types of social network data may be stored and used, including data representing partner relationships, family relationships, friend relationships, shared housing relationships, school/alumni relationships, ethnic associations, and club or society memberships. The visitor social network profile data 122 may furthermore be used to generate profile data of third parties associated with the parties 124. These third parties and their connections provide access to a large collection of profile data; although this third party data may relate less closely to the interests of the user 120, the data may provide information about the interests of the user's wider community.

The host social network profile data 132 includes data from web pages that are associated with the parties 134 (e.g. home pages of the host's friends). Each associated page provides equivalent data to that described above in relation to the host page 130. As with the visitor social network profile data 122, the host social network profile data 122 may be used to generate profile data of third parties, i.e. parties associated with the parties 134. The host's third party connections again provides access to a large collection of profile data which provide information about the interests of the host's wider community. Accessing visitor profile data 118, host profile data 126 and/or social network profile data 122, 132 advantageously allows the content generation engine 104 to generate personalised content 108 that is very closely tailored to the interests of the user 120 visiting the site 1 16. This data relates to the interests of the user 120 and of the interests of parties 124 associated with the user 120 and furthermore of information associated with the page 130 being viewed including, for example, the interests of the host 128 and parties 132 associated with the host 128. The profile data 212 is compiled by accessing and analysing social networking sites, for example MySpace, Friendster, Orkut or Facebook, thus reducing or removing any need to gather detailed profile information from the user 120. The content generation engine 104 is advantageously able to adjust content in real time to fit with the interest of the user 120 viewing the page 130 on the site 1 16. A further advantage of the personalised content generation system 102 is that the profile data may be sourced from the client 112, or the server 114, or through analysis of the structure of the page 130 and/or linking or linked web pages. The agent 106 may also source profile data from interaction with the user 120 during delivery of the personalised content 108, and/or through data stored locally on the content generation engine 104.

The personalised content generation system 102 includes: the agent 106, in the form of a software module on the web page server 114; and the content generation engine 104 comprising a number of computer servers, including a rules engine 202, a render engine

204, a content server 206 and one or more network attached storage devices (NASs) or storage area networks (SANs) 208, shown in Figure 2. The agent 106 may be in the form of an XML Query interface, and can also be implemented using HTML, JavaScript and/or Macromedia Flash. The rules engine 202, the render engine 204, and the content server

206 may be combined on a single machine, or combined with the web page server 1 14.

Any of the software components of the agent 106, the rules engine 202 or the render engine 204, and any other software components described herein, can be replaced or substituted at least in part by hardware components, such as dedicated ASICs (Application Specific Integrated Circuits) or FPGAs (Field Programmable Gate Arrays), to perform at least part of the processes performed by the software components. The client 112 may be for example a personal computer (e.g. from Lenovo Group Ltd or Apple Corporation) attached to the Internet, a mobile telephone, a PDA, or any computer device capable of communicating with the server 114, and displaying the personalised content 108. The page 130, and the personalised content 108 associated therewith, are processed by a web browser running on the client 112, such that displays are generated within the web browser. Typically therefore no installation process is required on the client machine. The web browser may be any browser that supports Java and Flash, such as FireFox, Safari or Internet Explorer. The server 114 may include an Apache web server running on a Linux computer, together with Java Server Pages (JSP) written to trigger the agent 106 and deliver the personalised content 108. The content generation engine 104 communicates with the server 114 over a network 210, for example the Internet, so that the generation of the personalised content 108 can occur at a site remote from the server 1 14, and may furthermore operate to provide personalised content 108, and delivery of the personalised content 108 to a plurality of web page servers 1 14.

The rules engine 202, shown in Figure 2, analyses profile data 212 (including at least the visitor profile data 118, the host profile data 126 and/or the social network profile data 122, 132) to select content which is deemed appropriate for the user 120 based on determinations made from processing the profile data 212 in conjunction with selection rules. The rules engine 202, as described below, uses a weighting process that takes into account the user's interests, the host's interests and the interest of the user's and the host's friends, weighted by a distance of separation between each friend and the original user 120 in each social network. The distance of separation may be generated from the degrees of separation between each friend and the user 120. Based on the weighting process, the rules engine 202 generates content templates, i.e video project files, for each user for processing by the render engine 204. The templates include references to content elements which have been selected by the rules engine 202 on processing of the selection rules. The rules engine 202 may be implemented using Microsoft SQL Server commands running on a PC server, such as that provided by IBM. The render engine 204 may be implemented on render servers, e.g. G5 computers from Apple Corp with video rendering software such as Final Cut Pro, or PC servers running ActiveSWF by CPS Labs Ltd or Kinetic Fusion server software by Kinesis Software Ltd. ActiveSWF and Kinetic Fusion are able to generate SWF files from intermediary XML project file languages. The render servers each include render engine components for controlling the respective video rendering tool, e.g. File Cut Pro or ActiveSWF. The render engine 204 handles receipt of the templates from the rules engine 202 and automatically attends to rendering and production of the content data in a format appropriate for the required server 114 and client 1 12 (e.g. SWF, or video for a cell phone). The render engine 204 communicates with the content server 206 to access content. The render engine components may be written in Java computer program code.

The content server 206 provides content, stored in content storage 208, for use in the personalised content 108 that will be delivered to the client 112. The content server 206 includes a digital asset manager (DAM) which is based on the MySQL database. The content server may support other databases, for example Oracle, SQLServer or Ingres. The content server 206 may be an IBM PC server running MySQL on Linux, Microsoft Windows Server 2003 or Apple OS X. The DAM provides access to a plurality of file servers via the NASs 208. The DAM may store any data directly for the content server 206 or data merely associated with remote data that is required by the content server 206. The content may include digital moving or static images, audio files, movie images, sound, voice, text, markup languages (e.g. HTML or XML). The content server 206 can be linked to another party's DAM data source.

In use, as shown in Figure 3, the client 112, directed by the user 120, requests from the server 114, at step 302, the page 130 of the site 1 16 associated with the host 128. In response to the page request 302, the server 1 14 triggers the agent 106 that activity between the client 112 and the server 1 14 is occurring (step 304). When the agent 106 is triggered by the server 1 14, the agent 106 gathers available profile data, including visitor profile data 118, host profile data 126 and associated social network profile data 122, 132 of the visitor (user 120) and the host 128, at step 306. The profile data gathering step 306 may include accessing data stored on the client 112, e.g., cookies, accessing data associated with the page 130 and pages linking to the page 130 and links on the page 130, accessing data associated with the site 116, accessing Customer Relationship Management (CRM) data available on the server 114, and accessing data from third parties which may be in network communication with the server 114.

Once the agent 106 has accessed the profile data 212 in step 306, the agent 106 sends the profile data 212 to the rules engine 202 with a request to generate personalised content 108, on the basis of the profile data 212 (step 308). The rules engine 202 accordingly processes rules, including CRM rules, using the profile data 212 to generate a content template, i.e. video project file, specifically for the user and sends the template to the render engine 204 at step 310. The generated personal template or file is normally derived for an initial campaign or message template for all users that is accessed by the rules engine. The personal template is tagged with references and commands that refer to and enable access to the selected content files or elements. The render engine 204 parses the template at step 312 to request content from the content server 206 and/or the content store 208. The requested content is provided at step 314 by the content server 206 to the render engine 204, which collates and renders the requested content in accordance with the template to produce an item of content 108 at step 315. The agent 106 subsequently receives the item of content from the render engine 204, via the rules engine 202, at step 316. The agent 106 transmits a communications message with content 108 to the client 1 12 via the server 114 at step 318. The server 114 additionally serves other data associated with the page 130, and the site 116, to the client 1 12 at step 318. At step 320, the client views the page 130 and the content in the form of a marketing or communications message.

The user 120 may interact with the served page 130 and the content therein. The elements contained in the marketing/communications message may include still images, moving images, sound, voice, text, interactive agents, ecommerce agents and place holders or tags for local content. The local content already resides on the client 112, and is inserted dynamically as the client 112 displays the page 130 (step 322). The user 120, in interacting with the page 130 and the content therein at step 324, may enter further personal data (e.g., contact details) or commence an ecommerce transaction. If an interaction or transaction is commenced by the client 112, the agent 106 participates in the interaction or transaction (step 326) and subsequently generates updates to the profile data 212 regarding the proceedings and outcome of the interaction or transaction (step 328). After the marketing/communications message and the page 130 have been served and any interaction or transaction with the client 112 is complete, the agent 106 initiates an optional time delay at step 330 before repeating the process from step 304. This provides for an update of the marketing/communications content displayed in page 130; i.e., this may allow for a personalised advertisement to be updated after a period of time, or after some interaction or transaction in step 326.

The content selection process 310 performed by the rules engine 202, as shown in Figure 4, compares the host profile data 126 to the visitor profile data 118 in a step 402. If the profile data 118, 126 matches, social network content is selected for the marketing/communications template at step 404; if however the profiles 1 18, 126 do not match, the host profile data 126 is compared to the social network profile data 122 of the visitor (i.e. user 120) in step 406. If the host profile data 126 does not match the visitor's social network profile data 122, default content for the content template is selected at step 412. If, on the other hand, the host profile data 126 does match the visitor's social network profile data 122 in step 406, the rules engine 202 determines (in step 408) whether the profile data 212 indicates that the user 120 and/or the host 128 has imposed a social network limit that defines a subset of the parties 124, 134 associated with the host 128 and/or user 120. The social network limit restricts the profile data accessible by the rules engine to only the profile data of the parties within the limit. The limit may represent a degree of separation or define specific parties, relationships or data that can be accessed or that is restricted. For example a social network limit may restrict access to only the partner, family and alumni of the host 128. The limit may define access to data on a wedding registry that is only available to specific parties. If the social network data to be accessed does not exceed any imposed social network limits, content is selected based on social network data at step 410. If on the other hand a social network limit restricts access to data, the default content social network is selected for the marketing/communications template at step 412. The content selection process 310 may include further steps to compare profile data, for example comparing profile data between the visitor, i.e. visitor profile data 118, and the host's social network profile data 132, or comparing the visitor's social network profile data 122 and the host's social network profile data 132. The social network profile data 122, 132 includes layers of data linked by the social networks of each of the parties 124, 134 associated with the user 120 and the host 128.

An advantage of selecting content based on matching of the visitor profile data 1 18 and the host profile data 126 is that the marketing/communications content template, and associated content, may be selected to be highly relevant to the visitor/user 120. This is of interest to the user 120 and to the provider of the personalised content 108. The content selection process 310 also advantageously provides other content when permission is granted. Furthermore, the content selection process 310 also provides a default version of the personalised content 108 that has been determined a priori to be of maximum relevance when no matches in profile data are found; this is particularly valuable in situations where profile data is poorly developed, or the marketing/communications message does not have a clearly defined audience.

The social networking-based selection process 404, as shown in Figure 5, includes a comparison of the similarities between two profiles (e.g. the visitor profile data 1 18 and the host profile data 126) with predetermined weights applied to each category (step 502). The process 404 also compares differences between profiles represented by profile data and applies weights to the differences (step 504). The weights may include emphases preselected by a party wishing to deliver the personalised content 108, and/or weights related to the degrees of separation between the two profiles For example, there is one degree of separation between the user 120 and the host 128; there are two degrees of separation between the user 120 and a friend 134 of the host 128; a relationship between the user 120 and a friend of one of the parties 132 associated with the host 128 has three degrees of separation. Weights can also be assigned based on the strength of relationships, e.g. a partner or family relationship might be considered of higher weighting than a friend relationship. Weights may be dynamic: if a user repeatedly responds to communications related to content selected because of a specific relationship, then the weight for that relationship can be increased, and ignoring content may cause a weight to decrease. For example, repeatedly buying flowers for a spouse would increase the weight both for flowers and for the spouse's profile. Once the differences and similarities between two profiles have been compared (steps 502, 504), a content structure for the personalised content 108 is selected, at step 506, followed by the selection of content items to be included in the structure, at step 508. The content structure is defined by a communications template, such as a marked-up video project file written using a markup format such as XML. The personal template is generated from the communications using one or more selection procedure rules that define the processing performed. The content items are selected based on further procedural rules and/or file references referenced within the personal template. Content items include still images, moving images, sound, voice, text, markup language (e.g., HTML, XML), stand alone applications, and place holders for local content to be inserted. Similar steps are performed for the processes 410 and 412.

An advantage of comparing similarities and differences between the profile data of profiles using weights based on degrees of separation is that similar interests, displayed in profiles, may be considered to be of greater importance if there is a low degree of separation between the parties associated with the profiles. Other indications which may contribute to weights include the frequency and time spent by the user 120 viewing the page 130 of host 128, and the reciprocal event of a host 128 subsequently viewing a page hosted by a user 120. Other weighting considerations may include selection of a party in a "friends" list on the site 116, and posting of comments by the user 120 on the page 130 corresponding to the host 128, i.e. contributions by a contributor party.

Multiple remote agents 106 on different servers 1 14 may be used to compile and store the profile data 212. The agents can also work collaboratively on providing data for the rules engine 202 to generate content for a client 112. Many modifications will be apparent to those skilled in the art without departing from the scope of the present invention as herein described with reference to the accompanying drawings.

Claims

CLAIMS:

1. A personalised content generation system, including: a content generation engine for generating personalised content for a user based on social network data of at least one social network related to the user.

2. A system as claimed in claim 1 , wherein the social network data includes profile data of parties of a social network of the user, said profile data being maintained on at least one social networking site.

3. A system as claimed in claim 1 or 2, wherein the social network data includes profile data of a social network of a selected party, the selected party being associated with a website selected by the user.

4. A system as claimed in claim 3, wherein the selected party is a contributor to the selected website.

5. A system as claimed in claim 4, wherein the content generation engine generates said personalised content based on profile data of the user.

6. A system as claimed in claim 1 or 5, wherein the content generation engine includes a rules engine for processing the social network data for selecting content elements of the personalised content.

7. A system as claimed in claim 6, wherein said rules engine processes the social network using content selection rules to generate a personalised content template.

8. A system as claimed in claim 7, wherein the selection rules apply weights to the social network data.

9. A system as claimed in claim 8, wherein the weights are based on a degree of separation between the user and a party represented by the social network data.

10. A system as claimed in claim 9, wherein the content generation engine is triggered by activity between a client device of the user and a website.

11. A system as claimed in claim 10, including an agent for the client device for extracting social network data for the rules engine, and for monitoring activity between the client and the website for triggering the content generation.

12. A system as claimed in claim 11, including a plurality of agents for extracting social network data for use by the content generation engine.

13. A system as claimed in claim 11, including a render engine for parsing the template generated by the rules engine to generate personalised content for the user.

14. A system as claimed in claim 13, including a content server for accessing content elements in response to requests from the render engine when parsing the template.

15. A personalised content generation process, performed by a computer system, including: accessing social network data of at least one social network related to a user; and processing the social network data to select content elements of personalised content for the user.

16. A process as claimed in claim 15, wherein the social network data includes profile data of parties of a social network of the user, said profile data being maintained on at least one social networking site.

17. A process as claimed in claim 16 or 17, wherein the social network data includes profile data of a social network of a selected party, the selected party being associated with a website selected by the user.

18. A process as claimed in claim 17, wherein the selected party is a contributor to the selected website.

19. A process as claimed in claim 15, including generating said personalised content based on profile data of the user.

20. A process as claimed in claim 19, wherein said social network data is processed using content selection rules to generate a personalised content template.

21. A process as claimed in claim 20, wherein the selection rules apply weights to the social network data.

22. A process as claimed in claim 21, wherein the weights are based on a degree of separation between the user and a party represented by the social network data.

23. A process as claimed in claim 22, including triggering generation of said content in response to communication between a client device of the user and a website.

24. A process as claimed in claim 23, including extracting social network data from the client device for processing by the rules, and monitoring activity between the client and the website to trigger the content generation.

25. A process as claimed in claim 24, including parsing the template to generate the personalised content for the user.

26. A process as claimed in claim 25, including accessing content elements of the selected content when parsing the template in response to references associated with the content included in the template.

27. A process as claimed in claim 15, wherein said content is video content.

28. A content generation process, including: accessing profile data for a visitor to a site and a contributor to said site; processing content selection rules to select content elements, said processing including comparing said visitor and contributor profile data; generating content with said content elements for said visitor; and serving said content for said visitor.

29. A content generation process as claimed in claim 28, wherein the visitor profile data includes social network data representative of a party's profile in the visitor's social network; and the contributor's profile data includes social network data representative of a party's profile in the contributor's social network.

30. Computer program code stored on computer readable media for use in performing a process as claimed in any one of claims 15 to 29.

31. A personalised content generation system including: a rules engine for executing content selection rules to process profile data for a user and social data for parties of a social network including said user to generate a personalised content template for the user, said template including references for accessing selected content; and a render engine for parsing the template and accessing the selected content to generate personalised content for the user.

32. A system as claimed in claim 31, wherein the personalised template is generated from a general communications message template.

33. A system as claimed in claim 7, 31 or 32, wherein said personalised content template is a video project file and the content is video content.