Ofertes de projectes

Most of the research on group recommendation investigated the core algorithms used for recommendation generation.  Two different strategies have been mostly used for generating group recommendations: aggregating individual predictions into group predictions or aggregating individual models into group models. Differences among these strategies differ in the timing of data aggregation step. 

In fact, the role of a group recommender system is to make suggestions that reflect the preferences of the group as a whole, while offering reasonable and acceptable options to individual group members. An important issue to be addressed in this kind of recommenders is how to reach consensus among members during and at the end of the recommendation process.

In this proposal, the focus will be on generating a group recommender framework that will be based on the well-known LibRec library. The main steps will be: 

1. Adapt Librec to be a GroupLibRec library 
2. Include in GroupLibRec the standard methods for generating group recommendations 
3. Propose new methods for generating group recommendations 
4. Analyze the proposals against well-known state-of-the-art methods

Note: Students interested in the proposal contact with the supervisor to an interview 

Boltzmann Machines are probabilistic models developed in 1985 by
D.H. Ackley, G.E. Hinton and T.J. Sejnowski. In 2006, Restricted
Boltzmann Machines (RBMs) were used in the pre-training step of
several successful deep learning models, leading to a new renaissance
of neural networks and artificial intelligence.

In spite of their nice mathematical formulation, there are a number of
issues that are hard to compute:

1. The computation of the partition function is NP-hard, involving an  exponential sum of terms
2. The exact computation of the derivative of the log-likelihood is  also NP-hard, since it contains the derivative of the partition function

Therefore, in practice we have to approximate both the computation of
the probabilities and several components of the learning process
itself. These drawbacks have prevented RBMs to show their real
potential as truly probabilistic models.

Currently, we are working on trying to improve several of the unsolved
issues related to RBMs:

1. - Mechanisms to control the learning process: www.lsi.upc.edu/~eromero/Publications/Downloads/2018-tnnls-stopcritRBM.pdf
2. Better approximations of the derivative of the log-likelihood: http://www.lsi.upc.edu/~eromero/Publications/Downloads/2019-nn-weightedCD.pdf
3. Efficient approximation of the partition function (work in progress)

These works have opened new lines of research, some of which can be
the topic of a Master's Thesis. The scope and degree of depth of the
work can be adapted to the estimated times to complete the Thesis. For
further details, contact Enrique Romero (eromero@cs.upc.edu).

This is precisely why companies such as Google have invested a lot of resources on exploring applications of GNN, some of which are as mediatic as Alpha Fold (recently solving the Protein-Folding problem), or used by Goolge Maps to make predictions of the expected traffic of a road. In this regard, the goal of this thesis is to continue investigating about the potential of GNNs when applied to the field of Computer Networks. Our goal is to provide GNN-based solutions that make computer networks run more autonomously and in a more efficient way.

Current regulation and ethical principles are requiring our models to preserve properties such as their explainability (i.e., the capability of a model to let the end-users know why a certain result was produced) and fairness (i.e., the capability to generate results that do not discriminate against certain categories of users, usually characterized by certain sensitive attributes). While the literature is rapidly advancing in both areas, the interplay between these two properties remains underexplored.

In this thesis, the student will engage with modern Python-based implementations of recommender systems, to inject in the optimization function both explainability and fairness guarantees.

The medical records of each patient contain textual information about the clinical evolution of the patient (including drugs, chemicals, diseases, symptoms and body parts). This corpus has already been represented in a structured format as a set of semantic graphs, using Neo4J graph database.

This thesis would aim at the acquisition of patterns of clinical behavior from these graphs. These patterns would be specifically devoted to provide help in diagnosis to the medical community in primary care. For example, we can get to automatically infer that a certain drug has a previously unknown side effect, or that patients suffering from a certain disease develop certain symptoms in a certain period of time. A simple example of pattern might be "patient has fever and is prescribed ibuprofen -(after x days) - patient has fever and soreness - (after y days) - patient has fever and breathing difficulty -> patient diagnosed with COVID".

The thesis would be a continuation of a previous master thesis in which we have built a baseline system that is able to extract simple rules whose elements are generalized codes. However, we want to extend this system to deal with sequence patterns as the ones mentioned before, eventually composed by sentences in Natural Language.

Since we have an annotated corpus of medical records, the project might use supervised and semi-supervised machine learning techniques as well as more standard data mining techniques.

Current regulation and ethical principles are requiring our models to preserve properties such as ser's privacy (i.e., the capability of a model to avoid disclosing sensitive information of the users) and fairness (i.e., the capability to generate results that do not discriminate against certain categories of users, usually characterized by certain sensitive attributes). While the literature is rapidly advancing in both areas, the interplay between these two properties remains underexplored.

In this thesis, the student will engage with modern Python-based implementations of recommender systems, to assess if privacy-preserving recommender systems, by hiding sensitive attributes of the users, might lead to unfair results. The final goal of the thesis is to produce novel approaches able to provide both privacy and fairness guarantees.

The goal of this thesis is to enable the application of Deep Reinforcement Learning techniques for optimization problems in very large and complex scenarios. Inspired by the last related work from OpenAI (https://openai.com/blog/evolution-strategies/), we will explore time-efficient alternatives to train a Graph Neural Network based DRL agent on large optimization scenarios in a reasonable amount of time.

A more detailed description of the project can be found in

https://www.cs.upc.edu/~eromero/Downloads/Retina-TFM-Project-01.pdf

The project is proposed in collaboration with Javier Zarranz Ventura
(Institut Clínic d'Oftalmologia, ICOF, Hospital Clínic de Barcelona, and
Institut d'Investigacions Biomèdiques August Pi I Sunyer, IDIBAPS),
which would provide a large annotated database to develop the project. For further information, please contact Alfredo Vellido (avellido@cs.
upc.edu) or Enrique Romero (eromero@cs.upc.edu).

Chatbots have proven to be an effective tool for information elicitation, to the point where they are considered better at engaging users and obtaining quality information than traditional surveys [1]. Recent efforts in chatbots have been directed towards studying how ethical questions should be tackled.

A first approach in the literature focuses on how chatbots should be designed to ensure ethical values such as transparency or privacy are taken into account.

An alternative is to study how human values can be integrated to improve interaction, by providing an actual representation of these [2]. In this second approach, the field of Natural Language Processing has made advances to make Conversational Agents seem as human-like as possible, studying how emotional aspects or values such as empathy can be considered in order to provide a more natural interaction [3].

These ethical matters are also a topic of discussion for other computer science areas. With research in the field of Machine Ethics recently orienting towards how Reinforcement Learning can be used for the value alignment of agents [4], with different approaches, such as the use of norm systems, to model moral values or encode moral reasoning.

The objective of this TFM is to endow a conversational agent with a value-aligned behaviour, combining recent Conversational Agent advances with the value-aligned approaches used in Machine Ethics. Specifically, we aim to align the chatbot with a given moral value, such as respect, through Reinforcement Learning.  Thus, we will study how the agent can learn to interact with the user in an ethical manner, and the value will be used to assess the agent's actions. The development is to be carried in the context of a serious game, with the chatbot being responsible for conducting a questionnaire with the user to evaluate the game.

 This offer has been pre-assigned to the student Eric Roselló Marin.

[1]  X. Han, M. Zhou, M. J. Turner, and T. Yeh, "Designing effective interview chatbots:  Automatic chatbot profiling and design suggestion generation for chatbot debugging," in Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems, 2021, pp. 1¿15.

[2]  J. Casas-Roma and J. Conesa, "Towards the design of ethically-aware pedagogical conversational agents," in International Conference on P2P, Parallel,Grid, Cloud and Internet Computing.    Springer, 2020, pp. 188¿198.

[3]  V. Scotti, R. Tedesco, and L. Sbattella, "A modular data-driven architecture for empathetic conversational agents," in2021 IEEE International Conference on Big Data and Smart Computing (BigComp).IEEE,  2021,  pp.365¿368.

[4]  M. Rodriguez-Soto, M. Lopez-Sanchez, and J. A. Rodriguez-Aguilar, "Guaranteeing the learning of ethical behaviour through multi-objective reinforcement learning.

Online educational platforms are promising to play a primary role in mediating the success of individuals' careers. Hence, while building overlying content recommendation services, it becomes essential to ensure that learners are provided with equal learning opportunities, according to the platform values, context, and pedagogy. Even though the importance of creating equality of learning opportunities has been well investigated in traditional institutions, how it can be operationalized scalably in online learning ecosystems through recommender systems is still under-explored.

The goal of this project is to analyze state-of-the-art metrics[1] that combines them to quantify the equality of learning opportunities among learners and mitigate the inequalities generated by the recommender systems as post-processed approach that balances personalization and learning opportunity equality in recommendations. To do it, we will consider state-of-the-art recommender systems, covering both model- and memory -based approaches and point-wise and pair-wise algorithms.

[1] http://www.mirkomarras.com/publication/ijaied2020/

Note: Students interested in the proposal contact with the supervisor to an interview 

The main objective of this project is to apply the most recent advances in Deep Learning, including Deep Reinforcement Learning and Graph Neural Networks, to uncover the particular methods used to track Internet users and collect PI.

This project will be useful for both Internet users and the research community, and will produce open source tools, real data sets, and publications revealing most privacy attempting practices.

Some preliminary results of our work in this area were recently published in Proceedings of the IEEE (IF: 9.237) and featured in a Wall Street Journal article.

More info at:

http://personals.ac.upc.edu/pbarlet/papers/web-tracking.survey2015.pdf

http://blogs.wsj.com/digits/2015/08/04/7-ways-youre-being-tracked-online-and-how-to-stop-it/

Recommender systems analyze the behavior of the users and their preferences, to learn patterns and understand what might be interesting to suggest them[4]. Natural imbalances in the data (e.g., in the amount of observations for a subset of popular items) might be embedded in the patterns. As a consequence, the produced recommendations exacerbate these imbalances, thus strengthening inequalities and generating biases [2]. When these imbalances are associated to sensitive attributes of the users (e.g., gender or race), this might have negative societal consequences, such as unfairness [3]. Unfairness might affect all the stakeholders involved in a recommender system, such as the users (when the minority receives systematically worse recommendations), or the content providers (when the items offered by a group of providers are exposed less than those of their counterpart) [1].

The goal of this project is to analyze and mitigate bias in recommender systems. To do it, we will consider state-of-the-art recommender systems, covering both model- and memory -based approaches and point-wise and pair-wise algorithms. 

[1] Himan Abdollahpouri, Gediminas Adomavicius, Robin Burke, Ido Guy, Dietmar Jannach, Toshihiro Kamishima, Jan Krasnodebski, and Luiz AugustoPizzato. 2020. Multistakeholder recommendation: Survey and research directions.User Model. User-Adapt. Interact.30, 1 (2020), 127¿158. https://doi.org/10.1007/s11257-019-09256-1

[2] Ludovico Boratto, Gianni Fenu, and Mirko Marras. 2019. The Effect of Algorithmic Bias on Recommender Systems for Massive Open Online Courses.InAdvances in Information Retrieval - 41st European Conference on IR Research, ECIR 2019, Cologne, Germany, April 14-18, 2019, Proceedings, Part I(Lecture Notes in Computer Science, Vol. 11437), Leif Azzopardi, Benno Stein, Norbert Fuhr, Philipp Mayr, Claudia Hauff, and Djoerd Hiemstra (Eds.).Springer, 457¿472. https://doi.org/10.1007/978-3-030-15712-8_30

[3] Sara Hajian, Francesco Bonchi, and Carlos Castillo. 2016. Algorithmic Bias: From Discrimination Discovery to Fairness-aware Data Mining.InProceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, August 13-17, 2016, Balaji Krishnapuram, Mohak Shah, Alexander J. Smola, Charu C. Aggarwal, Dou Shen, and Rajeev Rastogi (Eds.). ACM, 2125¿2126.https://doi.org/10.1145/2939672.2945386

[4] Francesco Ricci, Lior Rokach, and Bracha Shapira. 2015. Recommender Systems: Introduction and Challenges. InRecommender Systems Handbook,Francesco Ricci, Lior Rokach, and Bracha Shapira (Eds.). Springer, 1¿34. https://doi.org/10.1007/978-1-4899-7637-6_

Note: Students interested in the proposal contact with the supervisor to an interview 

 The goal of this thesis is to enable the application of Deep Reinforcement Learning techniques for optimization problems in very large and complex scenarios, in this case Data Center Networks. Inspired by the last related work from OpenAI (https://openai.com/blog/evolution-strategies/), we will explore time-efficient alternatives to train a Graph Neural Network based DRL agent on large optimization scenarios in a reasonable amount of time.

The goal of this thesis is to enable the application of Deep Reinforcement Learning techniques for optimization problems in very large and complex scenarios, in this case LEO satellite communication networks. Inspired by the last related work from OpenAI (https://openai.com/blog/evolution-strategies/), we will explore time-efficient alternatives to train a Graph Neural Network based DRL agent on large optimization scenarios in a reasonable amount of time.