The process of financial portfolio optimization involves choosing the most suitable mix of assets to meet a particular investment goal. Conventional portfolio optimization primarily focuses on maximizing returns and minimizing risks while overlooking the importance of social responsibility or sustainability in financial investments. In this paper, we present a Python-based multi-objective portfolio optimization library for sustainable investments (MOPO-LSI). MOPO-LSI is able to take Environmental, Social and Governance (ESG) factors into consideration in financial portfolio, where investors’ assets can be well allocated to mutual funds towards the ESG optimization along with their financial goals in the investment. MOPO-LSI is easy to be configured and used, and it is capable of production solutions in two scenarios – when client preferences are known or unknown. The developers can also easily customize the library to adapt it to their own financial objectives.

The FinRec workshop series offers a central forum for the study and discussion of the domain-specific aspects, challenges, and opportunities of RecSys and other related technologies in the financial services domain. Six years after the second edition of the workshop, the recent advances in the area of personalization and recommendation in financial services fostered the need for a new workshop aiming at bringing together researchers and practitioners working in financial services-related areas. Accordingly, the third edition of the event aims to: (1) understand and discuss open research challenges, (2) provide an overview of existing technologies using recommender systems in the financial services domain, and (3) provide an interactive platform for information exchange between industry and academia.

The development of recommender systems usually deal with single-objective optimizations, such as minimizing prediction errors or maximizing the ranking quality. There is an emerging demand in multi-objective recommendations in which the recommendation list can be generated by optimizing multiple objectives. For example, researchers may balance different evaluation metrics (e.g., accuracy, novelty, diversity) in their models, or consider different objectives in a multi-task recommender. This tutorial provides an overview of the multi-objective optimization and its applications in the area of recommender systems. More specifically, we summarize the multi-objective optimization methods, identify the circumstances in which a multi-objective recommender system could be useful, and point out the challenges in multi-objective recommendations.

Recommender systems have been applied in the area of education to suggest learning materials, assist informal learning, recommend projects or advisors, and so forth. In the process of teaching or learning, students may work together in a group. In this case, the group recommendations may come into play. In this paper, we propose to build a novel group recommendation model for education by introducing a virtual leader to each group. More specifically, we represent these virtual leaders by the personality traits, and learn these representations towards the optimization of the recommendations. Our experimental results demonstrate the effectiveness of the proposed models in comparison with other group recommendation models.

Recommender systems are able to produce a list of recommended items tailored to user preferences, while the end user is the only stakeholder in these traditional system. However, there could be multiple stakeholders in several applications domains (e.g., e-commerce, movies, music). Recommendations are necessary to be produced by balancing the needs of different stakeholders. First session of this tutorial introduces multi-stakeholder recommender systems (MSRS) with several case studies, and discusses the corresponding methods and challenges in MSRS. Reviews in an e-commerce platform may be mined to address cold-start problem and to generate explanations. Our earlier tutorial covered aspect-based sentiment analysis of products and topic models/distributed representations that bridge vocabulary gap between user reviews and product descriptions. Focus in the second session of this tutorial instead is on recent neural methods for review text mining - covering hands-on code for its use to enhance product recommendation. Each section will introduce topics from various mechanism (e.g., attention) and task (e.g., review ranking) perspectives, present cutting-edge research and a walk-through of programs executed on Jupyter notebook using real-world data sets.

Recommender systems (RS) have served as an effective technology-enhanced learning technique in the learning area. Traditional RS only consider the preferences of the end user who is the receiver the recommendations. Multi-stakeholder recommender systems (MSRS) are recently proposed to balance the needs of multiple stakeholders in the recommender systems. For example, the utility of the learning materials from the perspective of parents, instructors and even publishers may be also important in addition to the students' preferences in the area of educational learning. In this paper, we propose and exploit utility-based MSRS for personalized learning. Particularly, we develop our methods for preference corrections, in order to address the issue that instructors and students may have different perceptions on multiple aspects of the course projects. Our experimental results based on an educational data demonstrate the effectiveness of our proposed solutions.

Recommender systems have been applied as one of the technology-enhanced learning techniques for educations. Recently, multi-stakeholder recommender systems were proposed to balance the needs among different stakeholders, especially when there are conflicts of interests. In this paper, we create student communities by using the clustering technique, and seek the impact of these communities in the educational recommendations. Our experimental results identify the best community which can improve the multi-stakeholder recommendations.

Recommender systems are able to produce a list of recommended items tailored to user preferences, while the end user is the only stakeholder in the system. However, there could be multiple stakeholders in several applications or domains, e.g., e-commerce, advertising, educations, dating, job seeking, and so forth. Recommendations are necessary to be produced by balancing the needs of different stakeholders. This tutorial covers the introductions to multi-stakeholder recommender systems (MSRS), introduces multiple case studies, discusses the corresponding methods and challenges to develop MSRS. Particularly, a demo based on the MOEA framework will be given in the talk by using a speed-dating dataset.

Having data from multiple sources, cross-domain and context-aware recommender systems, with the help of transfer learning approaches, aim to integrate such data to improve recommendation quality and alleviate issues such as cold-start problem. With the advantages of these techniques, we host the second international workshop on intelligent recommender systems by knowledge transfer and learning (RecSysKTL) to provide such a forum for both academia and industry researchers as well as application developers from around the world to present their work and discuss exciting research ideas or outcomes. The workshop is held in conjunction with the ACM Conference on Recommender Systems 2018 on October 6th in Vancouver, Canada.

Recommender systems have been applied to several domains (e.g., online streaming, e-commerce, etc.) to assist decision making. Temporal information has been recognized and demonstrated as useful factor in improving the quality of recommendations. However, it is under investigated in the area of music recommendations. In this paper, we propose to integrate time dynamics and implicit feedback in the music recommender systems. More specifically, we develop a time-aware recommendation approach in which we produce simulated ratings by aggregating implicit feedback with time dynamics. The experimental results based on the last.fm music data demonstrate the effectiveness of our proposed approach.

In the recommender systems, the receiver of the recommendations may not be the only stakeholder in the system, while others may come into play. For example, job positions cannot be simply recommended to a user according to his or her tastes only without considering the expectations of the recruiters. In this paper, we propose a utility-based recommendation model which produces recommendations by optimizing the utilities of multiple stakeholders. Particularly, we take advantage of the multi-criteria ratings that are associated with user expectations and evaluations. And we propose to learn the user expectations by the learning-to-rank approaches if they are unknown in the data. We also propose to seek the optimal solutions by using the multi-objective optimization techniques. Our experiments based on a speed-dating data set demonstrate the effectiveness of the proposed methods in which we are able to keep the balance between multiple utilities and the recommendation performance by adopting the multi-objective optimization.

Recommender systems (RSs) have been successfully applied to alleviate the problem of information overload and assist users' decision makings. Multi-criteria recommender systems is one of the RSs which utilizes users' multiple ratings on different aspects of the items (i.e., multi-criteria ratings) to predict user preferences. Traditional approaches usually predict ratings on each criterion individually and aggregate them together to estimate the user preferences. In this paper, we propose an approach named as "Criteria Chains", where each combination of the criteria can be utilized in a way of contextual situations in order to better predict the multi-criteria ratings. Our experimental results based on the TripAdvisor and YahooMovies rating data sets demonstrate that our proposed approach is able to improve the performance of multi-criteria item recommendations.

Cross-domain recommender systems and transfer learning approaches are useful to help integrate knowledge from different places, so that we alleviate some existing problems (such as the cold-start problem), or improve the quality of recommender systems. With the advantages of these techniques, we host the first international workshop on intelligent recommender systems by knowledge transfer and learning (RecSysKTL) to provide such a forum for academia researchers and application developers from around the world to present their work and discuss exciting research ideas or outcomes. The workshop is held in conjunction with the ACM Conference on Recommender Systems 2017 on August 27th at Como, Italy.

Context suggestion refers to the task of recommending appropriate contexts to the users to improve the user experience. The suggested contexts could be time, location, companion, category, and so forth. In this paper, we particularly focus on the task of suggesting appropriate contexts to a user on a specific item. We evaluate the indirect context suggestion approaches over a movie data collected from user surveys, in comparison with direct context prediction approaches. Our experimental results reveal that indirect context suggestion is better and tensor factorization is generally the best way to suggest contexts to a user when given an item.

Recommender System has been successfully applied to assist user's decision making by providing a list of recommended items. Context-aware recommender system additionally incorporates contexts (such as time and location) into the system to improve the recommendation performance. The development of context-aware recommender systems brings a new opportunity - context suggestion which refers to the task of recommending appropriate contexts to the users to improve user experience. In this paper, we explore the question whether user's contextual ratings can be reused to produce context suggestions. We propose two evaluation mechanisms for context suggestion, and empirically compare direct context predictions and indirect context suggestions based on a movie data that was collected from user studies. The experimental results reveal that indirect context suggestion works better than the direct context prediction, and tensor factorization is the best approach to produce context suggestions in our movie data.

Recommender systems have been successfully applied to alleviate the information overload and assist user's decision makings. Emotional states have been demonstrated as effective factors in recommender systems. However, how to collect or predict a user's emotional state becomes one of the challenges to build affective recommender systems. In this paper, we explore and compare different solutions to predict emotions to be applied in the recommendation process. More specifically, we propose an approach named as collaborative chains. It predicts emotional states in a collaborative way and additionally takes correlations among emotions into consideration. Our experimental results based on a movie rating data demonstrate the effectiveness of affective prediction by collaborative chains in movie recommendations.

Recommender systems have been used in many domains to assist users' decision making by providing item recommendations and thereby reducing information overload. Context-aware recommender systems go further, incorporating the variability of users' preferences across contexts, and suggesting items that are appropriate in different contexts. In this paper, we present a novel recommendation task, "Context Suggestion", whereby the system recommends contexts in which items may be selected. We introduce the motivations behind the notion of context suggestion and discuss several potential solutions. In particular, we focus specifically on user-oriented context suggestion which involves recommending appropriate contexts based on a user's profile. We propose extensions of well-known context-aware recommendation algorithms such as tensor factorization and deviation-based contextual modeling and adapt them as methods to recommend contexts instead of items. In our empirical evaluation, we compare the proposed solutions to several baseline algorithms using four real-world data sets.

Context-aware recommender systems (CARS) help improve the effectiveness of recommendations by adapting to users' preferences in different contextual situations. One approach to CARS that has been shown to be particularly effective is Context-Aware Matrix Factorization (CAMF). CAMF incorporates contextual dependencies into the standard matrix factorization (MF) process, where users and items are represented as collections of weights over various latent factors. In this paper, we introduce another CARS approach based on an extension of matrix factorization, namely, the Sparse Linear Method (SLIM). We develop a family of deviation-based contextual SLIM (CSLIM) recommendation algorithms by learning rating deviations in different contextual conditions. Our CSLIM approach is better at explaining the underlying reasons behind contextual recommendations, and our experimental evaluations over five context-aware data sets demonstrate that these CSLIM algorithms outperform the state-of-the-art CARS algorithms in the top-N recommendation task. We also discuss the criteria for selecting the appropriate CSLIM algorithm in advance based on the underlying characteristics of the data.

Context-aware recommender systems (CARS) have been demonstrated to be able to enhance recommendations by adapting users' preferences to different contextual situations. In recent years, several CARS algorithms have been developed to incorporated into the recommender systems. For example, differential context modeling (DCM) was modified based on traditional neighborhood collaborative filtering (NBCF), context-aware matrix factorization (CAMF) coupled contextual dependency with the matrix factorization technique (MF), and tensor factorization directly models contexts as additional dimensions in the multi-dimensional space, etc. CAMF works well but it is difficult to interpret the latent features in the algorithm. DCM is good for explanation but it may only work well on data sets with dense contextual ratings. Recently, we successfully incorporate contexts into Sparse LInear Method (SLIM) and develop contextual SLIM (CSLIM) recommendation algorithms which take advantages of both NBCF and MF. CSLIM are demonstrated as more effective and promising context-aware recommenders. In this work, we provide the introduction on the framework of the CSLIM algorithms, present the current state of the research, and discuss our ongoing future work to develop and improve our CSLIM models for context-aware recommendations.

Context-aware recommender systems (CARS) take contextual conditions into account when providing item recommendations. In recent years, context-aware matrix factorization (CAMF) has emerged as an extension of the matrix factorization technique that also incorporates contextual conditions. In this paper, we introduce another matrix factorization approach for contextual recommendations, the contextual SLIM (CSLIM) recommendation approach. It is derived from the sparse linear method (SLIM) which was designed for Top-N recommendations in traditional recommender systems. Based on the experimental evaluations over several context-aware data sets, we demonstrate that CLSIM can be an effective approach for context-aware recommendations, in many cases outperforming state-of-the-art CARS algorithms in the Top-N recommendation task.