A5031651864- Topic Modeling
- Natural Language Processing Techniques
- Speech and dialogue systems
- Weed Control and Herbicide Applications
- Multimodal Machine Learning Applications
- Agronomic Practices and Intercropping Systems
- Mobile Crowdsensing and Crowdsourcing
- Plant Disease Resistance and Genetics
- Insect Pest Control Strategies
- Agricultural Science and Fertilization
- AI in Service Interactions
- Plant Disease Management Techniques
- Glaucoma and retinal disorders
- Insect behavior and control techniques
- Plant Growth Enhancement Techniques
- Biological Control of Invasive Species
- Expert finding and Q&A systems
- Retinal Diseases and Treatments
- Corneal surgery and disorders
- Plant Pathogens and Fungal Diseases
- Cryptography and Data Security
- Speech Recognition and Synthesis
- Sugarcane Cultivation and Processing
- Blockchain Technology Applications and Security
- Allelopathy and phytotoxic interactions
University of Chicago
2025
Baylor College of Medicine
2017-2024
University of California, Los Angeles
2024
Stanford University
2015-2023
University of California, Berkeley
2023
Laboratoire d'Informatique de Paris-Nord
2021
The University of Texas at Austin
2020
Baylor School
2018-2020
ORCID
2020
Catholic University of America
1999
Zero-shot transfer learning for multi-domain dialogue state tracking can allow us to handle new domains without incurring the high cost of data acquisition. This paper proposes zero-short technique where in-domain training are all synthesized from an abstract model and ontology domain. We show that augmentation through improve accuracy zero-shot both TRADE BERT-based SUMBT on MultiWOZ 2.1 dataset. with only reach about 2/3 obtained full art average across by 21%.
This paper presents the architecture of Almond, an open, crowdsourced, privacy-preserving and programmable virtual assistant for online services Internet Things (IoT). Included in Almond is Thingpedia, a crowdsourced public knowledge base natural language interfaces open APIs. Our proposal addresses four challenges technology: generality, interoperability, privacy, usability. Generality addressed by crowdsourcing while interoperability provided ThingTalk, high-level domain-specific that...
To understand diverse natural language commands, virtual assistants today are trained with numerous labor-intensive, manually annotated sentences. This paper presents a methodology and the Genie toolkit that can handle new compound commands significantly less manual effort. We advocate formalizing capability of Virtual Assistant Programming Language (VAPL) using neural semantic parser to translate into VAPL code. needs only small realistic set input sentences for validating model. Developers...
This article contains two cases with accompanying surgical videos, illustrating the use of a Tutoplast plug to create watertight seal in corneoscleral fistula formed after tube shunt removal. These demonstrate that can be cut into strategic geometric shape both at entry site and reinforce adjacent areas scleral thinning during removal or repositioning surgery. Specifically, piece dehydrated was shaped small rectangle an attached larger rectangle. The smaller "head" would fistula, while...
We propose AutoQA, a methodology and toolkit to generate semantic parsers that answer questions on databases, with no manual effort. Given database schema its data, AutoQA automatically generates large set of high-quality for training covers different operations. It uses automatic paraphrasing combined template-based parsing find alternative expressions an attribute in parts speech. also novel filtered auto-paraphraser correct paraphrases entire sentences. apply the Schema2QA dataset obtain...
We propose Semantic Parser Localizer (SPL), a toolkit that leverages Neural Machine Translation (NMT) systems to localize semantic parser for new language. Our methodology is (1) generate training data automatically in the target language by augmenting machine-translated datasets with local entities scraped from public websites, (2) add few-shot boost of human-translated sentences and train novel XLMR-LSTM parser, (3) test model on natural utterances curated using human translators. assess...
While Alexa can perform over 100,000 skills, its capability covers only a fraction of what is possible on the web. Individuals need and want to automate long tail web-based tasks which often involve visiting different websites require programming concepts such as function composition, conditional, iterative evaluation. This paper presents DIYA (Do-It-Yourself Assistant), new system that empowers users create personalized virtual assistant skills full generality composable control constructs,...
The molecular and phenotypic effects of a brown seaweed extract (BSE) were assessed in sugar beet (Beta vulgaris L.). Transcript levels BSE-treated untreated plants studied by RNA-seq validated quantitative real-time PCR analysis (RT-qPCR). Root morphology, yield, processing quality traits also analyzed to better elucidate the treatment effects. revealed 1019 differentially expressed genes (DEGs) between plants. An adjusted p-value < 0.1 an absolute value log2 (fold change) greater than one...
Purpose: To evaluate the impact of race and ethnicity, surgical technique, level surgeon training on recurrence rates after primary pterygium excision. Methods: A retrospective review patients who underwent excision at our academic institution was performed. The technique—conjunctival autografting (CAG) or amniotic membrane grafting (AMG)—was chosen attending surgeon's discretion, all surgeries were performed by surgeons in training. outcome measure recurrence, defined as regrowth...
Rhizoctonia solani, causing crown and root rot, is a major risk to sugar beet (Beta vulgaris L.) cultivation. The development of resistant varieties accelerated by marker-assisted selection priority breeding programs. We report the identification single-nucleotide polymorphism (SNP) marker linked resistance using restriction site-associated DNA (RAD) sequencing two geographically discrete sets plant materials with different degrees resistance/susceptibility enable wider superior genotypes....
The present study aimed to explore the effects of foliar application a leonardite-based product on sugar beet (Beta vulgaris L.) plants grown in field. approach concerned evaluation community compositional structure plant endophytic bacteria through metabarcoding approach, expression level gene panel related hormonal metabolism and signaling, main productivity traits. Results indicated that treated with leonardite (dosage 2,000 ml ha-1, dilution 1:125, 4 mg C l-1) compared untreated ones had...
The identification of efficient molecular markers related to low bolting tendency is a priority in sugar beet ( Beta vulgaris L.) breeding. This study aimed identify SNP associated with by establishing genome-wide association study. An elaborate 3-year field trial comprising 13 lines identified L14 as the one exhibiting lowest along an increased survival rate after autumnal sowing. For discovery following phenotyping, contrasting phenotypes 24 non-bolting and 15 plants line were sequenced...
Nancy Xu, Sam Masling, Michael Du, Giovanni Campagna, Larry Heck, James Landay, Monica Lam. Proceedings of the 2021 Conference North American Chapter Association for Computational Linguistics: Human Language Technologies. 2021.
This paper proposes a novel approach to let consumers share data from their existing web accounts and devices easily, securely, with fine granularity of control. Our proposal is have our personal virtual assistant be responsible for sharing digital assets. The owner can specify fine-grain access control in natural language; the executes requests on behalf requesters returns results, if conform owner's policies. Specifically, we allow any ThingTalk command--an event-driven task composed...
Building a question-answering agent currently requires large annotated datasets, which are prohibitively expensive. This paper proposes Schema2QA, an open-source toolkit that can generate Q&A system from database schema augmented with few annotations for each field. The key concept is to cover the space of possible compound queries on number in-domain questions synthesized help corpus generic query templates. data and small paraphrase set used train novel neural network based BERT...
Building a question-answering agent currently requires large annotated datasets, which are prohibitively expensive. This paper proposes Schema2QA, an open-source toolkit that can generate Q&A system from database schema augmented with few annotations for each field. The key concept is to cover the space of possible compound queries on number in-domain questions synthesized help corpus generic query templates. data and small paraphrase set used train novel neural network based BERT pretrained...
Background: In the absence of confirmatory biopsy, criteria for diagnosis neuro-ophthalmic sarcoidosis are not well established. Diagnostic both intraocular and neurosarcoidosis have been proposed, but remains challenging. It is our intention to augment what currently known about by providing a series biopsy-proven cases that contribute continued development diagnostic this enigmatic condition.Methods: Case four Caucasian women with sarcoidosis.Results: The first patient was initially...
To report a case of sequential open globe rupture in young patient with osteogenesis imperfecta type I following minor accidental blunt injury. This represented the patient's sole clinical manifestation connective tissue disease, leading to diagnosis at age 12 years old.A 12-year-old male presented right eye pain trauma school while wearing protective lenses. One year ago, he required surgical repair left during middle basketball game. His exam was significant for full-thickness corneal...
In distributed applications, the transmission of non-contiguous data structures is greatly slowed down by need to serialize them into a buffer before sending. We describe Compact Normal Forms, an API that allows programmers explicitly place immutable heap objects regions, which can both be accessed like ordinary as well efficiently transmitted over network. The process placing compact regions (essentially copy) faster than any serializer and amortized series functional updates structure in...
Leonardite-based biostimulants are a large class of compounds, including humic acid substances. Foliar application at field level improves plant growth, yield and quality through metabolic changes stimulation proton pumps. The present study aimed identifying optimum dosage BLACKJAK, acid-based substance, which is able to modify genes involved in sugar beet growth. Thirty-three belonging various biochemical pathway categories were tested leaves treated (Beta vulgaris L.) samples assess gene...