In order to address current water challenges, scientific research on water-related issues is crucial. However, traditional techniques for selecting research topics, such as literature reviews and expert opinions, can be time-consuming and may not provide a comprehensive overview of available information. We propose using Natural Language Processing (NLP) techniques in this study to extract, align, and compare water research topics from different corpora. We apply these techniques to the research paper abstracts from the New Mexico Water Resources Research Institute (NMWRRI) and the U.S. Geological Survey (USGS) to assess these institutions’ current research interests and identify potential new research directions. We use a Latent Dirichlet Allocation (LDA) model for topic extraction and a Word2Vec model for topic alignment. This study highlights the benefits of using NLP techniques to analyze trends and identify novel research directions in water studies.

Local Memory Store (LMStore) is a novel scratchpad memory (SPM) design, with recent research evaluation showing its capability for improving program performance. However, the performance of LMStore depends on its memory layout decided by its allocation scheme. In this paper, we evaluate the impact of SPM allocation on LMStore performance. Our experimental results, using benchmarks from the Malardalen WCET benchmark suite executing on LMStore architecture modeled in the PyCacheSim simulator, demonstrate that LMStore with a stack distance-based SPM allocation scheme significantly improves data movement by an average of 44.46% compared to a Cache-only architecture, and by an average of 23.89% compared to LMStore with a frequency-based SPM allocation scheme.

Recommender Systems ( RecSys ) provide suggestions in many decision-making processes. Given that groups of people can perform many real-world activities (e.g., a group of people attending a conference looking for a place to dine), the need for recommendations for groups has increased. A wide range of Group Recommender Systems ( GRecSys ) has been developed to aggregate individual preferences to group preferences. We analyze 175 studies related to GRecSys . Previous works evaluate their systems using different types of groups (sizes and cohesiveness), and most of such works focus on testing their systems using only one type of item, called Experience Goods (EG). As a consequence, it is hard to get consistent conclusions about the performance of GRecSys . We present the aggregation strategies and aggregation functions that GRecSys commonly use to aggregate group members’ preferences. This study experimentally compares the performance (i.e., accuracy, ranking quality, and usefulness) using four metrics (Hit Ratio, Normalize Discounted Cumulative Gain, Diversity, and Coverage) of eight representative RecSys for group recommendations on ephemeral groups. Moreover, we use two different aggregation strategies, 10 different aggregation functions, and two different types of items on two types of datasets (EG and Search Goods (SG)) containing real-life datasets. The results show that the evaluation of GRecSys needs to use both EG and SG types of data, because the different characteristics of datasets lead to different performance. GRecSys using Singular Value Decomposition or Neural Collaborative Filtering methods work better than others. It is observed that the Average aggregation function is the one that produces better results.

Most recommender services help individual users by recommending items within a single category based on the items' overall ratings. However, this may not be sufficient for group activities. For example, a group of friends using an online travel website look for a weekend getaway package (with hotel and restaurant), where the group members have different preferences over the characteristics (e.g., price, service, ambient) of these items. We call items with multiple characteristics as multi-criteria items. The items may come from different categories (e.g., hotel, restaurant). This paper proposes a novel problem of recommending packages of multi-criteria items to a group of users by leveraging users' preferences over categories. As far as we know, our work is the first paper studying this problem. We propose two models to measure the preference of a group to a package. The first model utilizes users' preferences for all the items and all categories, while the second model further leverages the influence of different group members to user preferences. We further introduce a new metric, to measure the fairness of the recommendations to different group members. We present an approach that utilizes co-clustering to incorporate items' characteristics in the calculation of user preferences and creating recommendations. Finally, we conduct extensive experiments with three real datasets. The experiments show that the second model can find packages that balance better the preferences of all the group members.