More Like this

1. Designing a System to Facilitate Intergenerational Story Sharing and Preservation for Older Adults

   https://doi.org/10.1145/3294109.3302957  

   Li, Cun (March 2019, Thirteenth International Conference on Tangible, Embedded, and Embodied Interaction (TEI ’19).)

   My PhD research focuses on intergenerational story sharing for older adults, which is conducted in a Research-through-Design manner. It includes five iterations: It started from an exploration prototype Interactive Gallery (1st iteration), and its findings helped to narrow down my research area and define my research question. To answer it, the 2nd iteration was continued, which was a co-design process of developing prototypes. 3rd and 4th iteration focused on older adults' life stories and memento stories respectively. While the 5th iteration is in the process, which aims to facilitate intergenerational story sharing and preservation in a sustainable manner. 

   more » « less
2. Towards the production of net-negative greenhouse gas emission bio-based plastics from 2nd and 3rd generation feedstocks

   https://doi.org/10.1016/j.jclepro.2024.141203  

   Van_Roijen, Elisabeth; Miller, Sabbie A (March 2024, Journal of Cleaner Production)

   Here, we show production pathways for greenhouse gas (GHG)-negative bio-based plastics from 2nd and 3rd generation feedstocks. We focus on bio-based plastics that are technically capable of replacing 80% of the global plastic market. By presenting life cycle inventories and discussing GHG-emissions hotspots, this work will inform stakeholders along the plastic supply chain of the necessary steps to achieving net-zero emissions by 2050, and potentially, how to drive net-uptake. This work is of critical importance given the overwhelming mass of plastic produced annually and the resulting CO2 emissions. To conduct this assessment, we derive life cycle inventories for nine different bio-based plastics and address the impact of methodological choices, such as allocation method, on the resulting 100a global warming potential (GWP). Our findings show that resources used and processing methods implemented have significant effects on the potential for us to derive carbon-negative plastics. Furthermore, we find that environmental impact quantification methods greatly influence the perceived GWP of such processes. For example, economic and mass allocation methods resulted in an apparent increase in GWP of up to 39% and 166%, respectively, compared to no allocation for bio-based plastics made from 2nd generation crops, whereas mass allocation resulted in the lowest GWP for bio-based plastics made from 1st generation crops. In considering environmental impact hotspots, our findings show that decarbonization of thermal energy and electricity, reduced use of ammonia-based fertilizer, renewable hydrogen production, use of bio-based alternatives for petrochemicals and plasticizers, enzyme production pathways from 2nd generation crops, and more efficient biomass conversion processes to reduce feedstock inputs may be critical steps in creating GHG negative bio-based plastics in the future. 

   more » « less
3. CAN NEURAL NETWORKS ACQUIRE A STRUCTURAL BIAS FROM RAW LINGUISTIC DATA?

   Warstadt, Alex; Bowman, Samuel R. (January 2020, Proceedings of the Annual Meeting of the Cognitive Science Society)

   null (Ed.)

   We evaluate whether BERT, a widely used neural network for sentence processing, acquires an inductive bias towards forming structural generalizations through pretraining on raw data. We conduct four experiments testing its preference for structural vs. linear generalizations in different structure-dependent phenomena. We find that BERT makes a structural generalization in 3 out of 4 empirical domains---subject-auxiliary inversion, reflexive binding, and verb tense detection in embedded clauses---but makes a linear generalization when tested on NPI licensing. We argue that these results are the strongest evidence so far from artificial learners supporting the proposition that a structural bias can be acquired from raw data. If this conclusion is correct, it is tentative evidence that some linguistic universals can be acquired by learners without innate biases. However, the precise implications for human language acquisition are unclear, as humans learn language from significantly less data than BERT. 

   more » « less
4. EmpatheticFIG at WASSA 2024 Empathy and Personality Shared Task: Predicting Empathy and Emotion in Conversations with Figurative Language

   Lee, Gyeongeun; Wang, Zhu; Ravi, Sathya N; Parde, Natalie (August 2024, Proceedings of the 14th Workshop on Computational Approaches to Subjectivity, Sentiment, & Social Media Analysis)

   Recent research highlights the importance of figurative language as a tool for amplifying emotional impact. In this paper, we dive deeper into this phenomenon and outline our methods for Track 1, Empathy Prediction in Conversations (CONV-dialog) and Track 2, Empathy and Emotion Prediction in Conversation Turns (CONV-turn) of the WASSA 2024 shared task. We leveraged transformer-based large language models augmented with figurative language prompts, specifically idioms, metaphors and hyperbole, that were selected and trained for each track to optimize system performance. For Track 1, we observed that a fine-tuned BERT with metaphor and hyperbole features outperformed other models on the development set. For Track 2, DeBERTa, with different combinations of figurative language prompts, performed well for different prediction tasks. Our method provides a novel framework for understanding how figurative language influences emotional perception in conversational contexts. Our system officially ranked 4th in the 1st track and 3rd in the 2nd track. 

   more » « less
5. Reanalyzing K’iche’ as a Tensed Language

   Maša Bešlin (October 2022, Proceedings of the 24th Workshop on American Indigenous Languages)

   Mayan languages have been claimed to lack the category Tense. Temporal interpretation is instead said to be guided by grammatical Aspect (see e.g. Larsen 1988 for K’iche’, Vázquez Álvarez 2002 for Chol, Bohnemeyer 2002 for Yucatec Maya, Coon 2016 for an overview, a.o.). In this paper, I examine the distribution and interpretation of the Tense/Aspect markers x- and k- in K’iche’, traditionally said to mark perfective (completive) and imperfective (incompletive) Aspect, respectively. I consider the co-occurrence possibilities of these markers with temporal adverbials (including temporal clauses), aspectual adverbials (‘in/for an hour’), the adverb na ‘still’, and individual level predicates. The evidence converges on the conclusion that the K’iche’ prefixes x- and k- mark (past and non-past) Tense rather than Aspect. The analysis is also shown to make the correct predictions for temporal matching in embedded clauses. Finally, I consider some uses of k- in past contexts and conclude that they are best seen as instances of the narrative present. 

   more » « less