e-book Persuasive Recommender Systems: Conceptual Background and Implications

Free download. Book file PDF easily for everyone and every device. You can download and read online Persuasive Recommender Systems: Conceptual Background and Implications file PDF Book only if you are registered here. And also you can download or read online all Book PDF file that related with Persuasive Recommender Systems: Conceptual Background and Implications book. Happy reading Persuasive Recommender Systems: Conceptual Background and Implications Bookeveryone. Download file Free Book PDF Persuasive Recommender Systems: Conceptual Background and Implications at Complete PDF Library. This Book have some digital formats such us :paperbook, ebook, kindle, epub, fb2 and another formats. Here is The CompletePDF Book Library. It's free to register here to get Book file PDF Persuasive Recommender Systems: Conceptual Background and Implications Pocket Guide.

Contents

  1. SESSION: Personalized Recommender Systems I
  2. The Advanced Guide to Emotional Persuasion
  3. Read Persuasive Recommender Systems: Conceptual Background and Implications (SpringerBriefs
  4. Conceptual Background and Implications

SESSION: Personalized Recommender Systems I

Whether users are likely to accept the recommendations provided by a recommender system is of utmost importance to system designers and the marketers who implement them. By conceptualizing the advice seeking and giving relationship as a fundamentally social process, important avenues for understanding the persuasiveness of recommender systems open up. Specifically, research regarding influential factors in advice seeking relationships, which is abundant in the context of human-human relationships, can provide an important framework for identifying potential influence factors in recommender system context.

This book reviews the existing literature on the factors in advice seeking relationships in the context of human-human, human-computer, and human-recommender system interactions. It concludes that many social cues that have been identified as influential in other contexts have yet to be implemented and tested with respect to recommender systems. Implications for recommender system research and design are discussed. JavaScript is currently disabled, this site works much better if you enable JavaScript in your browser. Computer Science Artificial Intelligence.

The Advanced Guide to Emotional Persuasion

Buy eBook. Buy Softcover. For all those concepts we present basic computational models directly applicable to the development of knowledge- based recommendation applications.

These components altogether form a framework consisting of modular components supporting the personalized configuration of product and service configurations. The remainder of this paper is organized as follows. In Section 2 we introduce the basic concepts of knowledge-based recommendation. Section 3 shows how utility- based approaches are used to rank product alternatives such that the most interesting items are presented first. In Section 4 we show how explanations can be ordered by taking into account serial position effects primacy and recency effects.

In Section 5 we present a model which supports the specific configuration of recommendation results sets taking into account context effects. Such a task can be defined as follows. Definition Recommendation Task. Example Recommendation Task. A simple example for such a recommendation task is the following. Note that P is a single constraint which specifies the available assortment in a disjunctive normal form i.

Definition Recommendation. On the basis of the given example, the two laptops r1, r2 can be recommended to the customer see Table 1.

About this book

This simple approach to identifying product recommendations can be extended by explicitly taking into account different types of decision phenomena. This issue is in the focus of the following sections.


  • International Classification of HRCT for Occupational and Environmental Respiratory Diseases;
  • Kundrecensioner.
  • The Advanced Guide to Emotional Persuasion.
  • Recommended for you.
  • Passar bra ihop!
  • Introduction To Meteorology.

Such a personalized ranking allows us to take into account primacy effects [5][8] which explain the fact that users pay more attention to items at the beginning of a list. Such rankings can increase the trust in recommendations since customers immediately see that the recommender system proposes interesting products in the first place. MAUT allows the evaluation of different product alternatives with regard to their utility for the current customer.

In this context, each product has to be evaluated according to a predefined set of interest dimensions. Economy and quality will be used as simple example interest dimensions throughout this paper. Tables reflect scoring rules describing the interrelationships between the given product attributes and the corres- ponding interest dimensions. For example, Table 3 denotes the fact that low-priced solutions have a higher contribution to the interest dimension economy whereas high- priced solutions have a higher contribution for the interest dimension quality.

On the basis of Tables , our set of example recommendations can be ranked as follows. The ranking of products productutility r is calculated using Formula 1, where con r,i denotes the contribution of product r to the interest dimension i and in i specifies the degree in which the customer in interested in dimension i.

Read Persuasive Recommender Systems: Conceptual Background and Implications (SpringerBriefs

Table 5 reflects the evaluation of the recommended products on the basis of the interests of a concrete customer. This customer has a higher interest in the dimension economy importance of 0,6 than in the dimension quality importance 0,4. Note that in real- world applications see, e. For reasons of simplicity, we assume that the importance weights have been directly specified by the customer. Such explanations can change or reinforce the beliefs and attitudes of customers by, e. An explanation can be seen as a specific type of argumentation [3] [15] which takes into account both, positive and negative argumentative intents [3].

In the recommendation context positive explanations are used for indicating the major benefits of a product, whereas negative explanations provide an indication for, e. Table 6 depicts a simple example for a set of explanations directly defined for the recommendation r1 only positive explanations are used in this context. After having defined our example set of explanations, we can now evaluate those explanations with regard to their utility for the customer. Formula 2 is used for this kind of utility calculation. The function exutility ea defines the utility of a specific explanation, con ea,i specifies the contribution of ea to the interest dimension i, and in i specifies the degree in which the current customer is interested in dimension i.

The offered system has a good price-performance ratio, e2 it includes lots of innovative technical features.

Movie Recommendation System with Collaborative Filtering

This recommendation represents a high-class e3 laptop in a lower price segment. The laptop includes proven components with extremely e4 low failure rates. The system represents the newest generation of the most e5 popular laptop series of company A. Table 6: Example set of explanations for r1. Given a set of arguments explanations related to a specific product, an important task is it to identify an ordering of those arguments which the customer finds interesting and convincing.

In our basic computational model for presenting explanations we assign to each argument position a corresponding utility value. The evaluation see Table 8 has been selected due to the results of our experiment regarding serial positioning effects in the presentation of product properties [5].

Conceptual Background and Implications

Related results indicated the existence of significant primacy and recency effects regarding the memorization of explanations regarding product properties. Consequently, our approach is to position the most valuable explanations at the beginning and at the end of a sequence. Out from those n! The formula used for determining the overall utility of a given set of ranked explanations is the following Formula 3 where exorderutility e Intuitively, this sequence corresponds to the fact that our example customer is more interested in the dimension economy that in the overall quality of a recommendation argumentations regarding price are included in the first and last position.

Note that in our simple example we do not take into account dependencies between explanations. However, in real-world settings such dependencies exist and must be taken into account. For example, we could require that explanation e3 is always positioned before e2 or that e3 is always on the first position. Furthermore, we could require that explanation sequences contain at most or exactly four items.

Consequently, explanation orderings have to take into account additional constraints typically imposed by marketing and sales experts. For calculating alternative orderings for a given set of explanations which take into account such constraints we introduce the following explanation ordering task. Definition Explanation Ordering Task.


  • “Hey, that could be me”: The role of similarity in narrative persuasion.
  • Life Lessons from Byron.
  • Newton and the Origin of Civilization.
  • Coffin Underground (John Coffin Mystery, Book 20);
  • Nuclear Forensic Analysis.

Example Explanation Ordering Task. A simple example for an explanation ordering task is the following. In this context, c1 imposes the restriction that all posi- tions have to differ and c2 requires that the explanation e3 has to be quoted before explanation e2. This is as well the upper bound since only five ordering variables have been defined.

In order to find the optimal solution, each of the identified sequences must be evaluated on the basis of Formula 3. A decision is taken depending on the context in which decision alternatives are presented [1].