logo-polimi
Loading...
Risorse bibliografiche
Risorsa bibliografica obbligatoria
Risorsa bibliografica facoltativa
Scheda Riassuntiva
Anno Accademico 2014/2015
Scuola Scuola di Ingegneria Industriale e dell'Informazione
Insegnamento 094748 - ADVANCED USER INTERFACES
  • 094747 - ADVANCED USER INTERFACES: TECHNOLOGY
Docente Cremonesi Paolo
Cfu 5.00 Tipo insegnamento Modulo Di Corso Strutturato

Corso di Studi Codice Piano di Studio preventivamente approvato Da (compreso) A (escluso) Insegnamento
Ing Ind - Inf (Mag.)(ord. 270) - MI (434) INGEGNERIA INFORMATICA* AZZZZ094746 - ADVANCED USER INTERFACES: TECHNOLOGY
094748 - ADVANCED USER INTERFACES
Ing Ind - Inf (Mag.)(ord. 270) - MI (474) TELECOMMUNICATION ENGINEERING - INGEGNERIA DELLE TELECOMUNICAZIONI* AZZZZ094748 - ADVANCED USER INTERFACES
Ing Ind - Inf (Mag.)(ord. 270) - MI (481) COMPUTER SCIENCE AND ENGINEERING - INGEGNERIA INFORMATICA* AZZZZ094748 - ADVANCED USER INTERFACES
094746 - ADVANCED USER INTERFACES: TECHNOLOGY

Programma dettagliato e risultati di apprendimento attesi

ADVANCED USER INTERFACES: TECHNOLOGY

Personalized interfaces based on recommender systems play a central role in the solution of big data problems. Recommender systems adopt data-mining and artificial-intelligence techniques to facilitate search in large amounts of digital information, and help users to identify the content that are likely to be more attractive or useful for them. Recommender systems infer such recommendations on the basis of different elements: popularity, demographic information about users, individual’s or community’s past preferences and choices, explicit ratings or comments, social networks, context of use. During the last years, recommender systems have seen an increasing adoption in various interactive services. The most famous success story of a recommender system is Netflix. The Netflix company launched in 2008 a competition (www.netflixprize.com) offering a 1 million dollar prize to anyone able to create a better recommendation system than the one adopted by Netflix itself for its video-streaming service.

This course explores the persuasiveness of recommender systems and presents several recommender algorithms, including the Netflix winner algorithm. For each algorithm, we investigate its quality in terms of accuracy and novelty of recommendations.

 

Topics

  1. Introduction to recommender systems
  2. Collaborative-based techniques
  3. Content-based techniques
  4. Advanced recommender systems: hybrid, context-aware, and cross-domains algorithms
  5. Evaluation of recommender systems
  6. Human-Computer interfaces for recommender systems: elicitation and explanation

 


Note Sulla Modalità di valutazione

There are four different ways for a student to pass the exam.

 

1. Homeworks (for students attending most of the lessons)

Homeworks consist of three activities:

  • Presentation and public defence of one paper: 12 points
  • Review and public attack of one paper: 12 points
  • Algorithm competition: 8 points (optional)

Defence and attack (24 points).

Two public mini-conferences will be held during regular class hours, the first one at the end of November, the second one at the end of January.

In preparation of the first mini-conference, students will be divided into two groups: defenders and attackers. Each student in the defender group will be assigned with one paper on recommender systems. The same paper will be assigned to a student in the attacker group. Papers will be assigned two weeks in advance. Papers will be 4-12 pages long. During the mini-conference, each defender will have 15 minutes to present his/her paper. The attacker will have 5 minutes to attack the paper, by pointing out weaknesses in its content or in the presentation. The defender will have to respond to the critiques.

For the second mini-conference, students will swap roles and new papers will be assigned.

Competition (8 points) - optional

At the beginning of the semester, a recommender system competition similar to the Netflix prize will be organised. The competition will close at the end of the first semester. Participation to the competition is optional.

Students participating to the competition can group in teams of no more than three students. Teams will have to write a recommender algorithm. The goal of the algorithm and the methodology used to evaluate its quality will be detailed at the opening of the competition.  Teams will be able to submit their solution once a week, till the closing of the competition. Each week, a score will be computed for each submission, based on the quality of the algorithm. A scoreboard will be published with the ranked list of submissions. At the closing of the competition, each team will submit a paper of 4 pages describing the algorithm implemented. Points will be assigned to each team based on the final position on the scoreboard and on the description of the algorithm. Best algorithms will be submitted for oral presentation to international conferences in the area; in case of acceptance, students will be given a limited grant to attend the conference.

 

2. Project

At the beginning of the semester, a very limited number of projects will be published. Projects consist in the implementation and evaluation of one or more recommender algorithms. Each project can be taken by a team of no more than three students. Projects will be assigned on a first-come-first-served basis. Projects have deadlines. Failing to complete the project within the deadline means failing the project. Best projects will be submitted for oral presentation to international conferences in the area; in case of acceptance, students will be given a limited grant to attend the conference. Projects can be extended to master thesis.

 

3. Integrated projects (for students taking the Technology + Interaction modules).

At the beginning of the semester, a very limited number of projects will be published, reserved for students taking both the modules of the AIU course. Each project can be taken by a team of no more than three students. Projects will be assigned on a first-come-first-served basis. Projects have deadlines. Failing to complete the project within the deadline means failing the project. Best projects will be submitted for oral presentation to international conferences in the area; in case of acceptance, students will be given a limited grant to attend the conference. Projects can be extended to master thesis.

 

3. Regular exam

Regular exam sessions will be organised as a mini-conference (defence and attack). Students willing to take the exam during regular sessions will have to contact the professor or the assistants two week in advance. Students will be assigned two papers: one paper to present and defend, one paper to attack. During the exam, the student will present the first paper (defence paper). The professor will attack the paper pointing out weaknesses in its content or in the presentation. The student will have to respond to the critiques. After the defence, the student will present the second paper (attack paper) pointing out its weaknesses.


Bibliografia
Risorsa bibliografica obbligatoriaCourse website http://recsys.paolocremonesi.org
Risorsa bibliografica facoltativaJannach, D., Zanker, M., Felfernig, A., & Friedrich, G, Recommender systems: an introduction, Editore: Cambridge University Press., Anno edizione: 2010 http://www.amazon.it/Recommender-Systems-Introduction-Dietmar-Jannach/dp/0521493366
Note:

Basic concepts

Risorsa bibliografica facoltativaKantor, Paul B., Lior Rokach, Francesco Ricci, and Bracha Shapira, Recommender systems handbook, Editore: Springer, Anno edizione: 2011 http://www.amazon.it/Recommender-Systems-Handbook-Francesco-Ricci-ebook/dp/B00D8D1Y10/ref=sr_1_1?s=english-books&ie=UTF8&qid=1413366876&sr=1-1&keywords=Recommender+systems+handbook
Note:

Advanced book


Mix Forme Didattiche
Tipo Forma Didattica Ore didattiche
lezione
30.0
esercitazione
20.0
laboratorio informatico
0.0
laboratorio sperimentale
0.0
progetto
0.0
laboratorio di progetto
0.0

Informazioni in lingua inglese a supporto dell'internazionalizzazione
Insegnamento erogato in lingua --
schedaincarico v. 1.6.1 / 1.6.1
Area Servizi ICT
22/11/2019