به سمت سامانه های توصیه‌گر
ترجمه نشده

به سمت سامانه های توصیه‌گر

عنوان فارسی مقاله: به سمت سامانه های توصیه‌گر مبتنی بر نوظهوری
عنوان انگلیسی مقاله: Towards novelty-driven recommender systems
مجله/کنفرانس: Comptes Rendus Physique
رشته های تحصیلی مرتبط: کامپیوتر
گرایش های تحصیلی مرتبط: مهندسی نرم افزار، هوش مصنوعی، معماری سیستم های کامپیوتری
کلمات کلیدی فارسی: سيستم هاي توصیه‌گر، نو‌ظهوری، Adjacent possible، دایره راحتی
کلمات کلیدی انگلیسی: Recommender systems، Novelties، Adjacent possible، Comfort zone
نوع نگارش مقاله: بررسی کوتاه (Mini Review)
نمایه: Scopus - Master Journals List - JCR
شناسه دیجیتال (DOI): https://doi.org/10.1016/j.crhy.2019.05.014
دانشگاه: Sony Computer Science Laboratories, Paris, 6, rue Amyot, 75005 Paris, France
صفحات مقاله انگلیسی: 9
ناشر: الزویر - Elsevier
نوع ارائه مقاله: ژورنال
نوع مقاله: ISI
سال انتشار مقاله: 2019
ایمپکت فاکتور: 3/367 در سال 2018
شاخص H_index: 60 در سال 2019
شاخص SJR: 1/208 در سال 2018
شناسه ISSN: 1631-0705
شاخص Quartile (چارک): Q1 در سال 2018
فرمت مقاله انگلیسی: PDF
وضعیت ترجمه: ترجمه نشده است
قیمت مقاله انگلیسی: رایگان
آیا این مقاله بیس است: بله
آیا این مقاله مدل مفهومی دارد: دارد
آیا این مقاله پرسشنامه دارد: ندارد
آیا این مقاله متغیر دارد: ندارد
کد محصول: E13217
رفرنس: دارای رفرنس در داخل متن و انتهای مقاله
فهرست مطالب (انگلیسی)

Abstract

1- Introduction

2- Recommender systems: space and user modeling

3- An agnostic assessment of recommender systems

4- A temporal non-trivial dynamics

5- Conclusion and perspectives: a new theoretical framework

References

بخشی از مقاله (انگلیسی)

Abstract

We get recommendations about everything and in a pervasive way. Recommender systems act like compasses for our journey in complex conceptual spaces and we more and more rely on recommendations to ground most of our decisions. Despite their extraordinary efficiency and reliability, recommender systems are far from being flawless. They display instead serious drawbacks that might seriously reduce our open-mindedness and our capacity of experiencing diversity and possibly conflicting views. In this paper, we carefully investigate the very foundations of recommendation algorithms in order to identify the determinants of what could be the next generation of recommender systems. We postulate that it is possible to overcome the limitations of current recommender systems, by getting inspiration from the way in which people seek for novelties and give value to new experiences. From this perspective, the notion of adjacent possible seems a relevant one to redesign recommender systems in a way that better aligns with the natural inclination of human beings towards new and pleasant experiences. We claim that this new generation of recommenders could help in overcoming the pitfalls of current technologies, namely the tendency towards a lack of diversity, polarization, the emergence of echo-chambers and misinformation.

Introduction

Recommender systems are ubiquitous in our everyday experience. We get books recommended, music recommended, food recommended, items to buy, hotels, trips. Even opinions and ideas. Recommendations are so entangled in our experience that perhaps we cannot even conceive our life without them. Still we base most of our decisions on the suggestions or tips we receive or seek for. Recommender systems act like compasses for our journey in complex conceptual spaces [25,1,2,9,13,22,14,13,22,14]. They exploit the knowledge about user behaviours and about the structure of the conceptual space itself, to suggest new directions to take, new experiences. Recommendations often concern something we may like because it is similar to something liked by someone else with a personal history similar to ours. This idea is brilliantly in place in almost all recommender systems with little variations taking into account personal histories, historical dependences, etc. And it turns out to be also very effective in many sectors thanks to the abundance of data about people and their choices. But all that glitters is not gold! Recently, a lot has been written about potential drawbacks of recommender systems in shaping our approach to information [25,1,2,9]. For instance, a controversy emerged about the impact of those systems on the diversity of contents experienced by users. While in some cases an enhancement of diversity has been reported [13,24,32], it is often claimed that personalization results in a loss of diversity [22,25,5]. This reduction could bring to a dangerous amplification of the human natural tendencies to homophily, to polarization, to the emergence of echo chambers and misinformation [25,1,2,9,10,7,29,11,8]. Is it possible to conceive recommender systems able to overcome these difficulties? This is a very pressing question that calls for a global rethinking of the way in which recommenders should accompany us in our everyday experience. Here we address this question by postulating the possibility of a new generation of recommender systems able to nudge us out of our “comfort zone,” experiencing instead novel, diverse, though still pleasant, experiences. New recommenders can be inspired by the way in which people experience novelties. The experience of the new is something very common in our lives, either at a personal level (e.g., reading a of a new book) or at the global level (e.g., writing a new book). In both cases, the experience of the new can be pictured as a path in a very special space, the space of the possible, the space of what could be. Francois Jacob highlighted the dichotomy between “actual” and “possible.” The “actual” is the set of things we experienced already. The “possible” is the set of things we might possibly experience in the future.