Personalizing rich media presentations based on user response to the presentation

What is claimed is:

1. A method for providing personalized media presentations over a network to user's terminal comprising:

delivering to the terminal hypermedia presentation streams of rich media material and a hierarchy of the hypermedia descriptions of hypermedia presentation streams for navigation by the user on the descriptions and presentation streams in at least three different modes of operation;

exposing a different set of selectable attributes in each the modes of operation;

allowing the user to switch between the modes and once in a particular mode select selectable attributes exposed in that mode;

gathering data about the user's interests expressed through selection events generated at the terminal by the user's navigation over the descriptions and presentation streams;

using the gathered data to generate a user model; and

dynamically changing what is delivered and rendered at the terminal in response to a new selection event in the user model based on the gathered selection events and reflected in the user model.

2. The method of claim 1 including dividing the hierarchy of hypermedia descriptions into a plurality of description levels in at least two of the modes where levels higher up in the hierarchial structure represent more abstract attributes of represented objects and levels lower in the hierarchial structure represent further refinements on those attributes.

3. The method of claim 2 including rendering the hierarchy of descriptions at the terminal so that the user first navigates over the hierarchial structure at a more abstract level and transitions to navigation over less and less abstract levels.

4. The method of claim 2 including first delivering to the terminal a hypermedia presentation stream at a level of the hierarchy with more abstract attributes of an item and thereafter delivering to the terminal at that level of the hierarchy in order of increasing refinement only those of the attributes in the hypermedia stream that reflect the user's interest and personality as expressed by navigation events.

5. The method of claim 2 including dynamically updating the content of certain of the abstract descriptive attributes in a hypermedia stream depending on the user's interest reflected by the user's navigation events while the remainder the hypermedia stream data itself is not replaced, to provide a more personalized description of what would be the same media content.

6. The method of claim 5 wherein three different modes of operation in a presentation includes;

a first navigation mode providing descriptive attributes for navigating a website including showing different categories, different members of the same category, and parent categories;

a second inspection mode providing descriptive attributes for examination of a selection in a category of goods including video and animation views of the selection panning and zooming in on the selection in a view and animating and rotating of the selection in a view; and

a third transaction mode providing descriptive attributes to facilitate acquisition of the selection including examining various options for the selection, comparison shopping for comparable items to the selection, calling for a sales representative, and completing the transaction.

7. The server of claim 6 including,

associating different descriptive attributes with different segments of an item in the hypermedia presentation involved in the client's navigation or interaction; and

expanding in later viewing by the user the descriptions of the segments of the item which the user has shown interest by the user's navigation events while leaving descriptions of other segments of the item unchanged.

8. The method of claim 1 including using explicit selection events and implicit selection events in generating the user model.

9. The method of claim 8 wherein an implicit selection event is one of the following events; repeated requests for more detail in the hypermedia presentation; dwell time on a rendered portion of the hypermedia presentation; and return parameters relative to a rendered portion of the hypermedia presentation.

10. The method of claim 8 including delivering an input to a user model based on a descriptive attribute associated with a portion of the hypermedia presentation involved in a selection event.

11. The method of claim 10 including providing entry input points for transactional commands whose parameters are determined by the descriptive attributes associated with the segment of the hypermedia involved in the navigation or interaction of the user.

12. The method of claim 11 including annunciation of a portion of the hypermedia presentation associated with an entry input point involved in the navigation or interaction by the user.

13. The method of claim 10 including varying descriptive attributes in a given portion of the hypermedia presentation depending on the mode of operation covered by that given portion.

14. A server for delivering personalizable media presentations to a user's terminal on a network comprising:

production logic for generating and continuing refining a user model of a user's requirements and temperament expressed through selection events generated by the user in navigation over a hypermedia presentation and a hierarchy of hypermedia descriptions of the presentation rendered at the user's terminal;

selection event capture software for transmission to the user's terminal for capturing user selection events and providing them back to the server; and

delivery logic responsive to the user model refined by one of the captured selection events, for delivering hypermedia streams containing the rendered presentation and the hierarchy hypermedia descriptions to the user's terminal, said delivery logic having rendering logic for dynamically changing the rendered presentations and hypermedia description based on said one of the selection events used to refine the user model.

15. The server of claim 14 wherein the hierarchy hypermedia description is represented in a multitude of description levels where levels higher up in the hierarchy represent more abstract attributes and levels lower in the hierarchy represent further refinements on the more abstract attributes.

16. The server of claim 14 wherein said selection event capture software includes code for providing inputs to user model to determine user's temperament based on implicit navigational actions of the user over the hypermedia.

17. The server of claim 16 wherein the implicit events include repeated request for details, dwell time and return parameters relative to portions of the hypermedia presentation.

18. A method for delivering personalizable media presentations to a user's terminal on a network comprising:

delivering rich media hypermedia presentation streams after first delivering and initially rendering in rich media a hierarchy of hypermedia descriptions of the hypermedia presentation stream for client navigation on the hypermedia descriptions;

loading the hypermedia presentation streams on the user's terminal based on user's selection events during navigation of the hypermedia descriptions;

constantly refining a user model based on the selection events;

varying selection and query inputs to points in the hypermedia descriptions in response to the selection events;

dynamically presenting to the user with the aid of the constantly refining user model, changes to the hypermedia presentation based on the user experiences first with the hypermedia description and then the hypermedia presentation;

dividing the presentation into three different modes of operation;

exposing a different set of selectable attributes in each of the modes of operation; and

allowing the user to switch between the modes of operation using one control action and once in a particular mode select selectable attributes in that mode using a different control action.

19. The method of claim 18 including representation by a hypermedia file is represented in a multitude of description levels where the description levels belong to a hierarchical structure where levels higher up in the hierarchial structure represent more abstract attributes and levels lower in the hierarchial structure represent further refinements on those attributes.

20. The method of claim 19 including presenting the hierarchy of descriptions on the client's terminal so that the client always first choose to navigate over representations at the most abstract level, of the descriptions, and transition to navigation over representations that correspond to less and less abstract descriptions.

21. The server of claim 18 wherein, said selection capture software includes counting software for determining the number times particular displayed material is visited and when a threshold level is exceeded providing user interest in that material back to the server.

22. The server of claim 21 wherein, said selection capture software includes timing software for timing user hovering on displayed material and when a time threshold is exceeded providing interest in that material back to the server.

23. The method of claim 18 wherein, the one and the different control actions are multiple and single mouse button clicks respectively.

24. The method of claim 18 wherein three different modes of operation in a presentation includes;

a first navigation mode providing descriptive attributes for navigating a website including showing different categories, different members of the same category, and parent categories;

a second inspection mode providing descriptive attributes for examination of a selection in a category of goods including video and animation views of the selection;

panning and zooming in on the selection in a view and animating and rotating of the selection in a view; and

a third transaction mode providing descriptive attributes facilitate acquisition of the selection including viewing various options for the selection, comparison shopping for comparable items to the selection, calling for a slates representative and completing the transaction.

25. The server of claim 24 including,

associating different descriptive attributes with different segments of an item in the hypermedia presentation involved in the client's navigation or interaction; and

expanding in future viewing by the user the descriptions of the segments of the item which the user has shown interest by the user's navigation events while leaving description of other segments of the item unchanged.

26. The method of claim 18 including:

dividing the presentation up into data segments and permitting the user to select between:

a manual mode where progress through the segments is by manual selection of the segments by the user;

an automatic mode where progress through the segments is from segment to segment at preset intervals; or

a semiautomatic mode where passing from one segment to another is a combination of manual and automatic actions.

27. The method of claim 26 permitting the user to select between rendering a portion or the whole of the data segments.

Description Submit all comments and votes

FIELD OF THE INVENTION

The present invention relates to providing media descriptions of objects on the internet and more particularly, to modifying those descriptions as they are presented to the user based on a user's reaction to the descriptions.

BACKGROUND OF THE INVENTION

The internet is capable of presenting rich media presentations, containing still images, animated images, video images and audio. One problem with these presentations, is the need to transmit large blocks of code and data to the client. This is a serious impediment for presenting products to electronic commerce customers who are typically connected over low-bandwidth dial-up inks. They cannot be expected to pre-load code and data or wait for very long before their shopping experiences commence. This problem is addressed in the above referenced copending U.S. patent application Ser. No. 09/100,418, entitled "Progressive Interleaved Delivery of Interactive Descriptions & Renderers for Electronic Publishing of Merchandise" filed on Jun. 20, 1988. This copending application is hereby incorporated by reference.

A second problem concerns optimized delivery of various rich media to easily and effectively address the right amount of information about the merchandise to each user. Each user has different interests, needs and expectations. The presentations must be able to attract and maintain the attention of users with a wide variety of interests and needs, different personalities and varying attention spans. Gathering and presenting information for these and other purposes needed by the user to express his or her selection preferences is difficult with presently available approaches.

Therefore, it is an object of the present invention to provide an improved method and apparatus for providing rich media experiences.

It is another object of the invention to provide a new method and apparatus to provide rich data experiences tailored to the individual user.

It is a further object of the invention to provide to the user improved selection capability for navigation in rich media.

BRIEF DESCRIPTION OF THE INVENTION

In accordance with the present invention, network presentations to be rendered at the user's terminal are programmed to exploit the advantages of rich media to expand user's ability to effectively provide information about selection preferences back on the network to the transmitting website as the presentation proceeds. In response to interrogation by a user for goods of services, the system presents a hypermedia presentation that starts out with delivery and presentation of simple thumbnail descriptions of the available goods and services, and proceeds in steps through delivery and presentation of more complex descriptions. The presentations are arranged in segments and permit multiple modes of operation. Each mode has a different set of attributes that permit progress through the presentation in a different manner. When a user first enters a website, he is placed in navigation mode which enables him to browse the website by moving from category to category of goods, having an initial look at the offerings of each category. The user can then move on to the inspection and/or transaction modes. In the inspection mode, the user can examine in more detail the various aspects of the goods he is interested in changing to a more detailed explanation, zoom in on various portions of the goods and obtain related information. When the user enters the transaction mode, he is apparently ready to purchase the goods and is given appropriate options for this purpose such as obtaining a quote, ask for human assistance, and trigger a transaction.

As the user navigates through the presentation in one or more of its modes of operation, his selection process is monitored not only for his overt acts concerning selected goods and services, but also for factors revealed by the user's selection process. Among factors that are analyzed are the time the user spends with various aspects of the presentation; the manner the user prefers to interact with the media (i.e., by mouse, voice or keyboard); and the number of times the user returns to a particular subject. In this way, the system accumulates information about the user and his preference and modifies the presentation on the fly to accommodate them. For instance, if the product to be sold is a vehicle, the system would quickly determine what type of vehicle the user was interested in, the features that he considered important in the car and the desired price range. The presentation would then be modified to emphasize vehicles of the preferred type within the price range the user showed interest in. The presentation would emphasize features the user has shown interest in by expanding the presentation of the features which the user showed interest in. Furthermore, the presentation is changed based on the personality of the user. If the data indicates the user is impatient or has short attention span, the presentation is made short and to the point. If the user is interested in going into details, the presentation is extended to include factors not normality conveyed in the usual presentation. If the user responds to audio, the verbal presentation is emphasized. If he has shown a preference for animation, video segments will be bypassed. Once the user's interests have been determined, they are presented on a transaction page offering options that can be added to the decided on goods.

BRIEF DESCRIPTION OF THE DRAWINGS

A better understanding of the invention can be had by the following description of one embodiment thereof while referring to the accompanying figures of which:

FIG. 1 is a schematic diagram which shows a typical selection process;

FIGS. 2A and 2B are a block diagram showing a network server in accordance with the present invention connected to a client terminal on the internet;

FIG. 3 is a view of the interaction between the client terminal user and network server of FIG. 1;

FIG. 4 is a schematic showing navigation through the hypermedia presentation;

FIG. 5 is a graphic representation of the possible modes of operation in the presentation;

FIGS. 6A and 6B are a diagram illustrating user selection of movement within and between the modes of operation by using a mouse;

FIG. 7 shows the attributes of the various modes of operation;

FIG. 8 is a three dimensional representation of the segments of a presentation;

FIG. 9 is a flow diagram of movement within the presentation of FIG. 7;

FIG. 10 is a flow diagram illustrating the accumulation of data on the customer's computer as a result of interaction with the server; and

FIGS. 11 and 12 are block diagramS of software in the user's terminal.

DETAILED DESCRIPTION OF THE INVENTION

FIG. 1 shows a hierarchical description of a portion of the articles being offered for sale at a website. Description level 0 shows the various goods available at the site. To begin with only level 0 information is loaded into a user's machine. While not loading level 1 and level 2 information until a user interaction sequence leads to such loading appears to be a lazy strategy, the aforementioned copending application provides for material to be loaded when rendering is anticipated during any stage of the presentation for smooth uninterrupted flow of the presentation.

As shown here, the selection of goods offered at a website includes vans 100, four door sedans 102 and accessories 104 for motor vehicles. The illustrated goods are only representative of the selections. The dashed vertical lines that separate the illustrated items from each other indicates that these level 0 items are not adjacent items in the browsable sequence, but are separated by segments of data for other items offered at the website. All the goods or the browsable sequence are linked by some semantic relationship. For instance, the van 100 is linked semantically to the sedan 102 as an item "similar to" it. While the accessories 104 are linked semantically to the sedan 102 by "benefits from" relationship. The user being presented with the level 0 choices on the screen selects what he or she is interested in: say, the purchase of a four door sedan as opposed to a van or an accessory. Where transitioning down from level 0 the level 1 in making a selection within the sedan category transitions are possible across the dashed horizontal lines following an "is a" semantic link. Thus, upon selection of the four-door sedan category, the user is presented with the next level of selections within that category. The level 1 descriptions cover a wide selection of four door sedans divided into price categories (here the range of possible choices is represented by an inexpensive Chevy 106, medium priced Toyota 10 and a expensive BMW 110). Let us assume the user selects the $25,000 Toyota. As provided for in the aforementioned copending application, as the user is proceeding through the process of selection, the material for the presentation is being loaded into the user's computer. Once the Toyota sedan has been selected, the computer is ready to present as a level 2 description either an animation presentation 112 or a video presentation 114 of the selected sedan.

An act of the user with respect to the presentation contains information the user intends to be conveyed by the act. However, other information not intended to be conveyed may be implied from the act. For instance, the choice of the Toyota by the user not only explicitly states his interest in that four door sedan, but can also indicate a price range that he or she is interested in. The selection of a $25,000 car could indicate that the user is interested in a mid-priced car and would not be interested in a more expensive car like a BMW or a less expensive one, such as a small Chevrolet. If the user loses interest in the Toyota sedan he or she is more likely to be interested in another car in the mid-price range than in a high priced luxury car. Therefore, anticipatory loading of presentations should be those of other mid-priced cars, and not those for luxury cars.

In addition to conveying information about the goods the user is interested in, the acts relative to the presentation may provide information about the user. For instance, the repeated choice of an animation or video presentation may indicate something about the user's temperament. By selecting one or the other, the individual may be indicating whether he is interested in details (by selecting the video track), or is impatient and wants information quickly (by selecting the animation track). The length of the presentation can be lengthened or shortened using this information. Also, repetitive acts such as examination of various segments of the animation and video tracks, signals aspects and features of the Camry the user is particularly interested in. As certain segments are repeated, the descriptions of the requested segments can be expanded over and above what would normally be shown. Information obtained with respect to the use's examination of the Camry presentation can also be used to deliver and render information about other automobiles. Information can also be stored for use when the user links up with the website at a later time looking for other goods or services.

Referring to FIG. 2, as the network server 200 interfaces with the user's terminal 202 over the internet 201, the user 203 interacts with the visual and audible presentations provided to the terminal 202 by the server 200, and software 204 at the user terminal monitors the user's actions. This software is provided by the network server 200 to the user's terminal 202 when product information is initially requested from the website.

The data obtained by the logic 204 is provided over the internet 201 to the server 200 through its network interface 205. The network interface provides this data to the delivery logic 206 in the modelling interpretation logic 207. If the request that can be responded to without interpretation (for instance, a change from the level 0 descriptions to level 1 descriptions), the request is sent by the delivery logic 206 to the presentation files 208 described in the previously mentioned co-pending application. The presentation files provide the requested information which is shipped out in packets 209 and 211 to the client terminal. The delivery logic may also obtain requested information through a media server interface 211 from outside media servers 212a to 212n.

If the request requires interpretation, the network interface provides the information to the modelling interpretation logic 207 which interfaces various software engines 214 to 217 through the non-media engine interface 213. Delivery data is then synthesized for delivery at the network server 200 by putting together in the delivery logic 206 segments from the presentation files 208 and/or media content segments from external media servers 212. The determined as to what the delivery logic 206 should deliver comes from both direct input from the client side through the network interface on new content request, and from the modeling and interpretation logic 207. The modeling and interpretation logic 207 receives all inputs from the user's terminal through the network interface 205, and processes them to provide even more sophisticated responses back to the user in the form of delivery data. To accomplish this, the modeling and interpretation logic not only builds and maintains its own simple session, state and user-model information, but through a non-media engine interface 213 also uses external user-model building, data base search, and statistics gathering engines 214, 215 and 216 respectively.

The modelling interpretation logic 207 takes the user interaction data and provides it to the statistics gathering engine 214 which collates the information provided by the interaction data. The interpretation logic then provides the collated information to the user model building engine 215. The user model building engine analyzes the collated information about the user's actions and builds a model which not only indicates what the user is interested in, but includes an analysis of the user's personal characteristics that will enable modification of the presentation to maintain the user's interest. In addition, the statistical engine 210 can gather user navigational statistics across a multitude of users and derive conclusions about the popularity of products and features. The user model generated by the user model building engine 215 is use