Showing 1–10 of 10 results for author: Mather, B

DAHRS: Divergence-Aware Hallucination-Remediated SRL Projection

Authors: Sangpil Youm, Brodie Mather, Chathuri Jayaweera, Juliana Prada, Bonnie Dorr

Abstract: Semantic role labeling (SRL) enriches many downstream applications, e.g., machine translation, question answering, summarization, and stance/belief detection. However, building multilingual SRL models is challenging due to the scarcity of semantically annotated corpora for multiple languages. Moreover, state-of-the-art SRL projection (XSRL) based on large language models (LLMs) yields output that… ▽ More Semantic role labeling (SRL) enriches many downstream applications, e.g., machine translation, question answering, summarization, and stance/belief detection. However, building multilingual SRL models is challenging due to the scarcity of semantically annotated corpora for multiple languages. Moreover, state-of-the-art SRL projection (XSRL) based on large language models (LLMs) yields output that is riddled with spurious role labels. Remediation of such hallucinations is not straightforward due to the lack of explainability of LLMs. We show that hallucinated role labels are related to naturally occurring divergence types that interfere with initial alignments. We implement Divergence-Aware Hallucination-Remediated SRL projection (DAHRS), leveraging linguistically-informed alignment remediation followed by greedy First-Come First-Assign (FCFA) SRL projection. DAHRS improves the accuracy of SRL projection without additional transformer-based machinery, beating XSRL in both human and automatic comparisons, and advancing beyond headwords to accommodate phrase-level SRL projection (e.g., EN-FR, EN-ES). Using CoNLL-2009 as our ground truth, we achieve a higher word-level F1 over XSRL: 87.6% vs. 77.3% (EN-FR) and 89.0% vs. 82.7% (EN-ES). Human phrase-level assessments yield 89.1% (EN-FR) and 91.0% (EN-ES). We also define a divergence metric to adapt our approach to other language pairs (e.g., English-Tagalog). △ Less

Submitted 12 July, 2024; originally announced July 2024.

Comments: 15 pages, 6 figures

Sizing and Location Selection of Medium-Voltage Back-to-Back Converter for DER-Dominated Distribution Systems

Authors: Xiangqi Zhu, Akanksha Singh, Barry Mather

Abstract: Medium-voltage back-to-back (MVB2B) converters can connect two distribution systems and quantifiably transfer power between them. This function can enable the MVB2B converter to exchange distributed energy resource (DER)-generated power between two systems and bring significant value to enhancing distribution system DER adoption. Our previous work analyzed and demonstrated the value the MVB2B conv… ▽ More Medium-voltage back-to-back (MVB2B) converters can connect two distribution systems and quantifiably transfer power between them. This function can enable the MVB2B converter to exchange distributed energy resource (DER)-generated power between two systems and bring significant value to enhancing distribution system DER adoption. Our previous work analyzed and demonstrated the value the MVB2B converter can bring to DER integration. As continuous work, this paper presents a methodology that helps address the MVB2B converter sizing and location selection problem in distribution systems with high DER penetrations. The proposed methodology aims to address three critical problems for MVB2B converter implementation in the real world: 1) which distribution systems are better to be connected, 2) what converter size is appropriate for connecting the distribution systems, and 3) where the optimal connection points are in the systems for connecting the MVB2B converter. The proposed methodology has been demonstrated by case studies that include various scenarios involving distribution systems with different dominated load types and high photovoltaic penetrations. △ Less

Submitted 12 August, 2022; originally announced August 2022.

Behavioral and Population Data Driven Distribution System Load Modeling

Authors: Isaac Bromley-Dulfano, Xiangqi Zhu, Barry Mather

Abstract: Distribution system residential load modeling and analysis for different geographic areas within a utility or an independent system operator territory are critical for enabling small-scale, aggregated distributed energy resources to participate in grid services under Federal Energy Regulatory Commission Order No. 2222 [1]. In this study, we develop a methodology of modeling residential load profil… ▽ More Distribution system residential load modeling and analysis for different geographic areas within a utility or an independent system operator territory are critical for enabling small-scale, aggregated distributed energy resources to participate in grid services under Federal Energy Regulatory Commission Order No. 2222 [1]. In this study, we develop a methodology of modeling residential load profiles in different geographic areas with a focus on human behavior impact. First, we construct a behavior-based load profile model leveraging state-of-the-art appliance models. We simulate human activity and occupancy using Markov chain Monte Carlo methods calibrated with the American Time Use Survey data set. Second, we link our model with cleaned Current Population Survey data from the U.S. Census Bureau. Finally, we populate two sets of 500 households using California and Texas census data, respectively, to perform an initial analysis of the load in different geographic areas with various group features (e.g., different income levels). To distinguish the effect of population behavior differences on aggregated load, we simulate load profiles for both sets assuming fixed physical household parameters and weather data. Analysis shows that average daily load profiles vary significantly by income and income dependency varies by locality. △ Less

Submitted 12 August, 2022; originally announced August 2022.

Grid Impact Analysis and Mitigation of En-Route Charging Stations for Heavy-Duty Electric Vehicles

Authors: Xiangqi Zhu, Partha Mishra, Barry Mather, Mingzhi Zhang, Andrew Meintz

Abstract: This paper presents a consolidated grid impact analysis design and corresponding mitigation strategies for heavy-duty electric vehicle (EV) charging stations. The charging load of heavy-duty charging station can reach several megawatts, which could induce adverse impacts on the distribution grid if not effectively mitigated. To analyze the impacts and provide corresponding solutions, we select fou… ▽ More This paper presents a consolidated grid impact analysis design and corresponding mitigation strategies for heavy-duty electric vehicle (EV) charging stations. The charging load of heavy-duty charging station can reach several megawatts, which could induce adverse impacts on the distribution grid if not effectively mitigated. To analyze the impacts and provide corresponding solutions, we select four representative distribution systems - including both single-feeder cases and a multi-feeder case - and design thorough test metrics for the impact analysis. The charging load profiles used in the analysis are derived from realistic conventional heavy-duty vehicle travel data. Based on the analysis results, charging stations are placed at three different representative locations in each distribution system: best, good, and worst locations. Mitigation strategies using a combination of smart charger functionality, on-site photovoltaic (PV) generation, and on-site energy storage (ES) are proposed and tested. A sizing method is also proposed to find the optimal PV-ES-charger capacity that minimizes the capital cost. △ Less

Submitted 12 August, 2022; originally announced August 2022.

From Stance to Concern: Adaptation of Propositional Analysis to New Tasks and Domains

Authors: Brodie Mather, Bonnie J Dorr, Adam Dalton, William de Beaumont, Owen Rambow, Sonja M. Schmer-Galunder

Abstract: We present a generalized paradigm for adaptation of propositional analysis (predicate-argument pairs) to new tasks and domains. We leverage an analogy between stances (belief-driven sentiment) and concerns (topical issues with moral dimensions/endorsements) to produce an explanatory representation. A key contribution is the combination of semi-automatic resource building for extraction of domain-d… ▽ More We present a generalized paradigm for adaptation of propositional analysis (predicate-argument pairs) to new tasks and domains. We leverage an analogy between stances (belief-driven sentiment) and concerns (topical issues with moral dimensions/endorsements) to produce an explanatory representation. A key contribution is the combination of semi-automatic resource building for extraction of domain-dependent concern types (with 2-4 hours of human labor per domain) and an entirely automatic procedure for extraction of domain-independent moral dimensions and endorsement values. Prudent (automatic) selection of terms from propositional structures for lexical expansion (via semantic similarity) produces new moral dimension lexicons at three levels of granularity beyond a strong baseline lexicon. We develop a ground truth (GT) based on expert annotators and compare our concern detection output to GT, to yield 231% improvement in recall over baseline, with only a 10% loss in precision. F1 yields 66% improvement over baseline and 97.8% of human performance. Our lexically based approach yields large savings over approaches that employ costly human labor and model building. We provide to the community a newly expanded moral dimension/value lexicon, annotation guidelines, and GT. △ Less

Submitted 20 March, 2022; originally announced March 2022.

Comments: Accepted to Findings of the Association for Computational Linguistics, 2022

MSC Class: 68T50 ACM Class: I.2.7

Grid Value Analysis of Medium Voltage Back-to-Back Converter on DER Hosting Enhancement

Authors: Xiangqi Zhu, Akanksha Singh, Barry Mather

Abstract: This paper presents an analysis of the value that can be realized by medium-voltage back-to-back (MVB2B) converters in terms of increased utilization rate of distributed energy resource (DER) and the improvement in operational conditions. A systematic, transferrable, and scalable methodology has been designed to analyze and quantify the increased DER value from three perspectives: 1) curtailment r… ▽ More This paper presents an analysis of the value that can be realized by medium-voltage back-to-back (MVB2B) converters in terms of increased utilization rate of distributed energy resource (DER) and the improvement in operational conditions. A systematic, transferrable, and scalable methodology has been designed to analyze and quantify the increased DER value from three perspectives: 1) curtailment reduction of the DER generation, 2) size reduction of the energy storage needed to otherwise realize DER hosting levels, and 3) hosting capacity improvement of DER compared to base distribution circuit capability. In the case study, the proposed methodology is applied to two utility distribution systems for analysis and quantification of the grid value of the MVB2B converter, installed in the distribution circuit, and provided to the solar photovoltaic (PV) DERs. The analysis results demonstrate that the MVB2B converter can deliver significant value to PV hosting enhancement of two adjacent distribution systems when they are connected by the MVB2B converter. Based on this case study, this paper analyzes and summarizes the approximate realized grid value of the MVB2B converter for distribution systems dominated by different shares of customer classes. △ Less

Submitted 8 March, 2022; originally announced March 2022.

Learning to Plan and Realize Separately for Open-Ended Dialogue Systems

Authors: Sashank Santhanam, Zhuo Cheng, Brodie Mather, Bonnie Dorr, Archna Bhatia, Bryanna Hebenstreit, Alan Zemel, Adam Dalton, Tomek Strzalkowski, Samira Shaikh

Abstract: Achieving true human-like ability to conduct a conversation remains an elusive goal for open-ended dialogue systems. We posit this is because extant approaches towards natural language generation (NLG) are typically construed as end-to-end architectures that do not adequately model human generation processes. To investigate, we decouple generation into two separate phases: planning and realization… ▽ More Achieving true human-like ability to conduct a conversation remains an elusive goal for open-ended dialogue systems. We posit this is because extant approaches towards natural language generation (NLG) are typically construed as end-to-end architectures that do not adequately model human generation processes. To investigate, we decouple generation into two separate phases: planning and realization. In the planning phase, we train two planners to generate plans for response utterances. The realization phase uses response plans to produce an appropriate response. Through rigorous evaluations, both automated and human, we demonstrate that decoupling the process into planning and realization performs better than an end-to-end approach. △ Less

Submitted 4 October, 2020; v1 submitted 25 September, 2020; originally announced September 2020.

Comments: Accepted at EMNLP 2020 (Findings)

The Panacea Threat Intelligence and Active Defense Platform

Authors: Adam Dalton, Ehsan Aghaei, Ehab Al-Shaer, Archna Bhatia, Esteban Castillo, Zhuo Cheng, Sreekar Dhaduvai, Qi Duan, Md Mazharul Islam, Younes Karimi, Amir Masoumzadeh, Brodie Mather, Sashank Santhanam, Samira Shaikh, Tomek Strzalkowski, Bonnie J. Dorr

Abstract: We describe Panacea, a system that supports natural language processing (NLP) components for active defenses against social engineering attacks. We deploy a pipeline of human language technology, including Ask and Framing Detection, Named Entity Recognition, Dialogue Engineering, and Stylometry. Panacea processes modern message formats through a plug-in architecture to accommodate innovative appro… ▽ More We describe Panacea, a system that supports natural language processing (NLP) components for active defenses against social engineering attacks. We deploy a pipeline of human language technology, including Ask and Framing Detection, Named Entity Recognition, Dialogue Engineering, and Stylometry. Panacea processes modern message formats through a plug-in architecture to accommodate innovative approaches for message analysis, knowledge representation and dialogue generation. The novelty of the Panacea system is that uses NLP for cyber defense and engages the attacker using bots to elicit evidence to attribute to the attacker and to waste the attacker's time and resources. △ Less

Submitted 20 April, 2020; originally announced April 2020.

Comments: Accepted at STOC

Adaptation of a Lexical Organization for Social Engineering Detection and Response Generation

Authors: Archna Bhatia, Adam Dalton, Brodie Mather, Sashank Santhanam, Samira Shaikh, Alan Zemel, Tomek Strzalkowski, Bonnie J. Dorr

Abstract: We present a paradigm for extensible lexicon development based on Lexical Conceptual Structure to support social engineering detection and response generation. We leverage the central notions of ask (elicitation of behaviors such as providing access to money) and framing (risk/reward implied by the ask). We demonstrate improvements in ask/framing detection through refinements to our lexical organi… ▽ More We present a paradigm for extensible lexicon development based on Lexical Conceptual Structure to support social engineering detection and response generation. We leverage the central notions of ask (elicitation of behaviors such as providing access to money) and framing (risk/reward implied by the ask). We demonstrate improvements in ask/framing detection through refinements to our lexical organization and show that response generation qualitatively improves as ask/framing detection performance improves. The paradigm presents a systematic and efficient approach to resource adaptation for improved task-specific performance. △ Less

Submitted 20 April, 2020; originally announced April 2020.

Comments: Accepted at STOC

Detecting Asks in SE attacks: Impact of Linguistic and Structural Knowledge

Authors: Bonnie J. Dorr, Archna Bhatia, Adam Dalton, Brodie Mather, Bryanna Hebenstreit, Sashank Santhanam, Zhuo Cheng, Samira Shaikh, Alan Zemel, Tomek Strzalkowski

Abstract: Social engineers attempt to manipulate users into undertaking actions such as downloading malware by clicking links or providing access to money or sensitive information. Natural language processing, computational sociolinguistics, and media-specific structural clues provide a means for detecting both the ask (e.g., buy gift card) and the risk/reward implied by the ask, which we call framing (e.g.… ▽ More Social engineers attempt to manipulate users into undertaking actions such as downloading malware by clicking links or providing access to money or sensitive information. Natural language processing, computational sociolinguistics, and media-specific structural clues provide a means for detecting both the ask (e.g., buy gift card) and the risk/reward implied by the ask, which we call framing (e.g., lose your job, get a raise). We apply linguistic resources such as Lexical Conceptual Structure to tackle ask detection and also leverage structural clues such as links and their proximity to identified asks to improve confidence in our results. Our experiments indicate that the performance of ask detection, framing detection, and identification of the top ask is improved by linguistically motivated classes coupled with structural clues such as links. Our approach is implemented in a system that informs users about social engineering risk situations. △ Less

Submitted 25 February, 2020; originally announced February 2020.

Comments: Accepted at AAAI 2020