Personalized Hypermedia Final Plus3

Personalized hypermedia presentation techniques for improving online customer relationships Alfred Kobsa, Jurgen Koenemann, Wolfgang Pohl Presented by: Satyajit Das, Pornpat Nikamanon INF 231 – Human-Computer Interaction November 29, 2007

Outline Introduction Input data and acquisition methods User data Usage data Environment data Representation and secondary inferences Adaptation production Conclusions and prospects

I. Introduction E-commerce Hypermedia Personalisation techniques Personalisation

E-commerce Scenario in today's competitive business environment . Pre-sales phase Corporate Identity, draw attention to products & services Sales phase Online storefronts, ordering/purchasing facilities Post-sales phase Reassurance, product support, loyalty program

Costumer Relationship Software Facilitates collection of information about large number of customers (interests, purchase behavior, support needs) Dynamic nature - updates of contents & presentation to react to new opportunities & challenges Global around-the-clock presence independent of its locality Dynamic creation of content & presentation for personalised information delivery.

Popular e-commerce website offer personalisation

Hypermedia presentation techniques Hypermedia system is an interactive system that allows users to navigate a network of linked hypermedia objects. Hypermedia objects contain a set of related content-bearing elements of different media types such as text, images, video-clips, audio- clips, small applications (applets) and interaction elements (e.g., menus, buttons or checkboxes). Knowledge based techniques allow to incorporate experiences with regard to customers and their characteristics, customer types, products and their characteristics, and sales strategies.

Personalised Hypermedia Application A system which adapts the content, structure and/or presentation of the networked hypermedia objects to each individual users characteristics , usage behavior and usage environment . There are different basic types of adaptation depending upon the amount of control user has over the adaptation. Adaptive system - system performs all steps autonomously. User controlled adaptivity - the system lets the user make selection and performs the adaptation. User initiated adaptivity - user request and lets the system decide the best option.

Adaptive system This is a result of Adaptivity , which means that the system recognizes that the user is interested in churches, thus it highlights the corresponding options. AVANTI - Tourist information system Personalised Hypermedia Application

II. Input data and acquisition methods Different types of the data and methods of acquiring the information about the user's characteristics, computer usage behavior and usage environment which are required in adapting the system to the user's needs User data Usage data Environment data

Input data and acquisition methods User data Personal characteristics of the user. Demographic data User knowledge User skills and capability User interests and preferences User goals and plans

User data User Knowledge data Assumptions about users’ knowledge about concepts, relationships between concepts and facts and rules with regard to the domain of the application system are the most important sources for personalisation. Examples:- Restricting or increasing the explanatory pages to be presented to the user depending on his or her expertise. Generating expertise dependent product description. Intelligent tutoring systems.

Conditional text Presentation Generates interest-tailored descriptions of objects which are tuned to different user interests. ILEX – adaptive learning system

User data User Skills and capabilities Knowledge of the ability and skills of the users plays an important role in adapting systems to users needs. The system may also distinguish between the actions a user is familiar with and the actions he or she is actually able to perform. It is possible that a user knows how to do something but is not able to perform the action due to lack of required permissions or to some physical handicap. Example:- Adaptive help systems

Tourist information system AVANTI takes the needs of different kinds of disabled people (wheelchair-bound, motor-impaired & vision-impaired) into account, therefore only recommends actions that these users are actually able to perform. User data

User data User interests and preferences Interests among users of the same application often vary considerably. Example:- Promotion of cars to different audiences, conflicting sets of attributes (speed, sex-appeal, safety, family- friendliness) must be emphasised

User data User interests and preferences User interest is a central notion for the Recommendation systems. The items recommended may be products, services, documents, news and so on.

User interests and preferences Recommendations can also be made by asking users to rate items with which they are already familiar. User data

User data User goals and plans Typical goals may be to find information on a certain topic or to shop for some kind of product. A system that supports users in achieving their goals facilitates and speeds up interaction considerably since the system has expectations about the next user actions and can therefore interpret them in a more flexible way.

User data User goals and plans Plan-Recognition Aims at identifying the goal (or intention) of the user based on the actions they perform in an environment Narrows the number of possible goals by observing the actions the user performs.

User data User goals and plans Plan Recognition – Inputs and outputs Inputs: a set of goals the user carries out in the domain, a set of plans describing the way in which the user can reach each goal, an user-action observed by the system. Output: foretelling the user's goal, and determining how the observed action contributes to reach it

User model acquisition methods User supplied information Active acquisition User data is acquired through questions asked by the system, typically in an initial phase of system usage. Caution: - Self-assessment is error-prone since users are often not correctly aware of things like their own capabilities. Some systems therefore present controlled queries, tests, exercises that are aimed at a more objective assessment of the user.

User model acquisition methods User supplied information Downside of active acquisition - Paradox of the Active User Users are motivated to get started and are in a hurry to get their immediate task done. In cases of competing information sources, users may simply refuse to visit the site if they have to respond to an interview ﬁrst. The paradox is information acquired will be helpful in adapting the system and making in more user-friendly in the long run. The acquisition phase should therefore be minimised and ideally be administered only after the user has already obtained some impression about the beneﬁts the site has to offer.

User model acquisition methods Stereotype reasoning A simple method for making a first assessment of others is to classify them into categories and to then make predictions about them based on a stereotype that is associated with each category. Main components of stereotype are:- a body, which contains information that is typically true of users to whom the stereotype applies a set of activation conditions (“triggers”) for applying the stereotype to a user.

Usage data Observable usage Selective actions Temporal viewing behavior Ratings Purchases and purchase-related actions Other confirmatory and disconfirmatory actions Usage regularities

Usage data - Observable usage Selective actions User makes a choice if competitive links are available on the current page Actions Clicking link Scrolling and enlarging operation Document expansion operation Movie and audio operation Other actions at user interfaces Indicators Interest Unfamiliarity Preferences iGoogle MY YAHOO!

Usage data - Observable usage Temporal viewing behavior Viewing time Difficulty of measurement User not present in front of the computer Window is covered by other windows Item is outside the visible window area Negative evidence – not interesting to user Presentation time less than certain threshold Abort download Presses the back button shortly after the page download commenced Streamed data (video/audio) User reaction shortly after termination User interest in this streaming Further research Micro interaction level Usage of eye-tracking YouTube

Usage data - Observable usage Ratings Users are required to explicitly rate objects How relevant/interesting to user How relevant/interesting to other user Rating type Binary scale (I don’t like it / I like it) Numeric scale (I hate it / I don’t like it / It’s OK / I like it / I love it) Problem Relevance is always relative (changing) ‏ User not rate Pandora Amazon

Usage data - Observable usage Purchases and purchase-related actions A purchase is a strong indicator of user interest in some of the features of the purchased product React adaptively by suggesting similar or related items No one-to-one mapping between purchases and interests Purchase for other people (gift) ‏ Already own on available item Amazon

Usage data - Observable usage Other confirmatory and disconfirmatory actions Strengthen an assumption in concert with a preceding selection Examples Saving Printing Bookmarking Forwarding by email Amazon ZDNet

Usage data - Usage regularities Further processing of usage data to acquire information about users’ preferences, habits, and levels of expertise User frequency Categorize events and count their frequencies Examples Microsoft word - Adaptive icon toolbar AVANTI – Introduce shortcut links Situation-action correlations Interface agents / personal assistants Suggestions based on statistics correlations between previous situations and action Action sequences Recommend the generation of macros Predict future user actions Recommend actions

Environment data Constraints Software environment Hardware environment Locale Mapping model Single-user machine Multi-user machine

Environment data Software environment Browser version and platform Availability of plug-ins Java and JavaScript Header of HTTP requests User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1)

Environment data Hardware environment Bandwidth Processing speed Display devices Input devices

Environment data Locale Users’ current location Characteristics of usage locale noise level, brightness of the surrounding, etc.

III. Representation and secondary inferences Most common representation approaches and the inference techniques Deductive reasoning - from the more general to the more specific Logic-based representation and inference Representation and reasoning with uncertainty Inductive reasoning - from specific cases to the general case Learning about the user Analogical reasoning - from similar cases to the present case Clique-based filtering Clustering user profiles A hybrid approach : user profiles as learning results

IV. Adaptation production Adaptation Production explains the possible types of adaptation to the user, usage and environment data of an individual user. Adaptations in hypermedia systems can take place at different levels that we will discuss in the following sections: Adaptation of content Adaptation of presentation and modality Adaptation of structure

Adaptation of content Adaptation of Content describes methods for personalising the content of hypermedia objects and pages in accordance with user, usage and environment data. Personalization functions of content adaptation Techniques for content adaptation

Adaptation of content – Personalisation functions

Adaptation of content - Techniques Techniques for content adaptation A number of techniques on different levels of granularity and localisation have been developed so far for adapting hypermedia content to different user needs. Page variants Fragment variants Fragment coloring Adaptive stretchtext Adaptive natural-language generation

Adaptation of content - Techniques Page variants Authoring different versions of all pages in which adaptation occurs Adaptation at runtime is confined to adaptive page selection. Cumbersome since a completely new page has to be written for each variation of local adaptations that may occur on a page Inflexible since many pages have to be modified if a single local adaptation is changed iGoogle

Adaptation of content - Techniques Fragrant variants Authored for each adaptive page fragment At runtime, the appropriate fragments are included into a static page frame. It requires the dynamic generation of web pages at runtime. The granularity of a page fragment A paragraph of text An image A table A video iGoogle

Adaptation of content - Techniques Fragrant colouring The content of the hypermedia remains unchanged across all users. For each individual user certain elements of the hypermedia presentation are marked out, e.g. as being important, irrelevant or too demanding for him or her. Fragment colouring can only be applied in such areas where content can be presented in the same formulation to all users, and where the variability of adaptation across all users is relatively low.

Adaptation of content - Techniques Adaptive Stretchtext Stretchtext is “elastic” text that the user can extend or collapse by clicking on it with the mouse. Instead of retrieving a new page, clicking on an active link or hotword results in additional text being displayed in a pop-up window. In personalised hypermedia systems, stretchtext can additionally be automatically expanded and collapsed by the system, taking the user model into account. Advantage: - Users can adapt the page content manually if the adaptation that was automatically performed by the system is not appropriate or desired.

Adaptation of content - Techniques

Adaptation of content - Techniques Adaptive natural-language generation Natural language generation techniques to create alternative text descriptions for different users. A simple approach are text templates with slots that can be ﬁlled with descriptions of different complexity based on the user’s level of expertise Natural-language generation also seems to be a promising complement to stretchtext.

Generates interest-tailored descriptions of objects which are tuned to different user interests. Adaptation of content - Techniques

Adaptation of presentation and modality Adaptations of presentation are adaptations where the information content of the hypermedia objects ideally stays the same while the format and layout of the objects change. Change of modality - images to text, from text to audio, or from video to still images. Adaptation in hypermedia systems concerning multimedia presentations are often based on User’s preferences User’s physical abilities System performance

Modality “image” has been changed to “text” based on the selection of different modalities by the user’s physical abilities. Adaptation of presentation and modality

Adaptation of presentation and modality

Adaptation of structure Adaptation of structure refers to changes in the way in which the link structure of hypermedia documents or its presentation to users is changed. Techniques for structure adaptation: - Adaptive link sorting Adaptive link annotation Adaptive link hiding and “unhiding” Adaptive link disabling and enabling Adaptive link removal/addition

Adaptation of structure - Techniques Adaptive link sorting Primarily employed for ranking the target web pages based on their relevance to users’ interests and goals, and their appropriateness for the user based on the user’s background knowledge. example: - ranked lists of recommended items, such as movies ranked lists of recommended items as technical equipment frequency of use, e.g. in personalised views and spaces The sorting of link lists can be used for non-contextual links only. Downside: - Caution should be exercised, however, since automatic item sorting in menus based on usage frequency has been found to potentially confuse users

Adaptive link annotation Links that have already been visited change their colours. This is non-adaptive link annotation is well known from all major web browsers There are up to six different annotations with the following meanings: suggested , ready , inferred , known , done and not-ready . Adaptive hypermedia systems use different colours and symbol codes to annotate links in a personalised manner. Adaptation of structure - Techniques

Adaptive link hiding/unhiding Adaptive link hiding removes the “ visible cue ” of a link in such a way that the link anchor looks like normal text or a normal icon. The idea is to visually reduce the hyperspace to support users’ navigation, and to guide users to those pages which the system assumes to be the currently most relevant ones or that are probably comprehensible to the user given his or her presumed level of knowledge. Adaptation of structure - Techniques

Adaptive link disabling and enabling Link disabling removes the functionality of a link but leaves its visual appearance nearly untouched (i.e. the link anchor still looks like a link, but nothing happens when one clicks on it). Disabled Link Downside: - This behaviour of a link considerably violates the principle of expectation conformance in human-computer interaction, link disabling and enabling is currently used together with link hiding and unhiding only. Adaptation of structure - Techniques

Adaptive link removal/addition Adaptive link removal deletes the link anchors completely. Its is an effective way to support users’ navigation by reducing the number of navigation steps to achieve a certain goal and by reducing the user’s cognitive overload. This technique can only be applied to non-contextual links. example: - removal of links to pages which are not yet appropriate for a learner removing links to irrelevant subtasks removing items in a product listing that are probably of no interest to the user Downside:- if a stable listing of links is used frequently, removal of individual links would also violate the constancy principle of human- computer interaction and should therefore be used with caution. Adaptation of structure - Techniques

(HIPS – Mobile web-based Museum guide) Link addition - system automatically introduces links to nearby paintings and also links to more distant paintings based on user interest in their topic, painter and time periods of paintings that the user visited before. Adaptation of structure - Techniques

Adaptation of structure – Personalisation functions Some personalisation functions of structure adaptation Adaptive recommendations. Adaptive orientation and guidance Personal views and spaces

Adaptive recommendations Recommendations concerning products Lists of links to “products” and services are filtered or ranked based on user data, and presented to the user. Amazon.com, moviefinder.com. Recommendation concerning information Lists of links to documents or other pieces of information are ranked based on user and usage data Recommended news - Google news. Navigation recommendations Links to hypermedia pages (usually at the same site) are filtered or ranked based on user, usage and environment data. Systems customised for different user profiles. Adaptation of structure – Personalisation functions

Adaptive orientation and guidance Overview maps Personalised overview maps mark those pages that users visited or bookmarked in the past. Guided site tours Personalised guided tours can take user data into account and modify the tour so that it caters better to users’ presumable interests. It helps in familiarising ﬁrst-time users with the basic offerings of a website. Personalised next buttons This is a very ﬂexible method for the presentation of learning material because the destination node of the next button is not directly connected to the current node but can be dynamically computed at runtime, taking even the very last actions of the user into account. Adaptation of structure – Personalisation functions

Personalised view bookmarking facilities that are integrated into most current web browsers provide users with personalised access to web resources generate shortcuts for frequently followed links Adaptation of structure – Personalisation functions

Personalised Space view histories of their past actions (e.g. purchases and reservations) set markers (e.g. for books to buy in the future) deﬁne shortcuts to site-speciﬁc resources they frequently access specify information they want to have forwarded to them automatically (e.g. quotes for certain stocks, news from certain categories) ‏ save documents and news in a personal repository. Adaptation of structure – Personalisation functions

Personalised Space iGoogle Netvibes Adaptation of structure – Personalisation functions

V. Conclusion and prospects Public websites Information kiosk Ubiquitous access opportunities for customers “ Universal access”, “Design for all”, “User interfaces for all” with respect to computer systems

Public websites Objectives are to keep visitors at the site, to turn them into customers and to make them come back. Different types of support must be given to ﬁrst-time visitors, returning visitors, and infrequent and frequent users of the website.

Information kiosks Fairs, exhibitions, showrooms, and public places. System needs to support “walk up and use” by ﬁrst-time users or infrequent users. An integrated approach should be pursued where the adaptation of content and presentation format is also based on the different hardware and software environments and the different locales.

Ubiquitous access opportunities for customers Web-capable appliances and web-capable mobile devices like car-mounted displays and portable digital assistants integrated with telephone functionality

Ubiquitous access opportunities for customers Embedded devices that have access to data from their physical environments can provide the basis for adapting information to the current needs of individual customers, such as providing a list of dealers closest to their current location. Wigglestick, Mobile Technologies Group, Georgia Tech

Universal access “ Universal access”, “Design for all”, “User interfaces for all” with respect to computer systems - (Shneiderman, 2000; Stephanidis, 2001) Software systems should be designed in such a way that they pose no access barriers to people with special needs, such as users with disabilities, elderly users and users with different cultural background. Specialized systems (such as screen readers or web browsers for the blind) vs. Generic applications (adapt to the needs of these different special audiences) ‏

Personalized Hypermedia Final Plus3

More Related Content

Similar to Personalized Hypermedia Final Plus3 (20)

More from Satyajit Das (9)

Recently uploaded (20)

Personalized Hypermedia Final Plus3