More Like this

1. An Exploratory Interface for Dataset Repositories Using Cell-Centric Indexing

   https://doi.org/10.1109/BigData50022.2020.9378057  

   Johnson, Drake ; Register, Keith ; Davison, Brian D. ; Heflin, Jeff ( December 2020 , IEEE International Conference on Big Data (IEEE BigData 2020))

   null (Ed.)

   Large collections of datasets are being published on the Web at an increasing rate. This poses a problem to researchers and data journalists who must sift through these large quantities of data to find datasets that meet their needs. Our solution to this problem is cell-centric indexing, a novel approach which considers the individual cell of a dataset to be the fundamental unit of search, indexing the corresponding metadata to each individual cell. This facilitates a new style of user interface that allows users to explore the collection via histograms that show the distributions of various terms organized by how they are used in the dataset. 

   more » « less
2. An Architecture for Cell-Centric Indexing of Datasets

   Qiu, Lixuan ; Jia, Haiyan ; Davison, Brian D. ; Heflin, Jeff ( December 2020 , CEUR workshop proceedings)

   null (Ed.)

   ncreasingly, large collections of datasets are made available to the public via the Web, ranging from government-curated datasets like those of data.gov to communally-sourced datasets such as Wikipedia tables. It has become clear that traditional search techniques are insufficient for such sources, especially when the user is unfamiliar with the terminology used by the creators of the relevant datasets. We propose to address this problem by elevating the datum to a first-class object that is indexed, thereby making it less dependent on how a dataset is structured. In a data table, a cell contains a value for a particular row as described by a particular column. In our cell-centric indexing approach, we index the metadata of each cell, so that information about its dataset and column simply become metadata rather than constraining concepts. In this paper we define cell-centric indexing and present a system architecture that supports its use in exploring datasets. We describe how cell-centric indexing can be implemented using traditional information retrieval technology and evaluate the scalability of the architecture. 

   more » « less
3. Asking Participant-specific Outcome Questions for a Deeper Understanding of Complex Events

   Vallurupalli, Sai ; Ghosh, Sayontan ; Erk, Katrin ; Balasubramian, Niranjan ; Ferraro, Francis ( December 2022 , Conference on Empirical Methods in Natural Language Processing (EMNLP))

   Knowledge about outcomes is critical for complex event understanding but is hard to acquire. We show that by pre-identifying a participant in a complex event, crowdworkers are able to (1) infer the collective impact of salient events that make up the situation, (2) annotate the volitional engagement of participants in causing the situation, and (3) ground the outcome of the situation in state changes of the participants. By creating a multi-step interface and a careful quality control strategy, we collect a high quality annotated dataset of 8K short newswire narratives and ROCStories with high inter-annotator agreement (0.74-0.96 weighted Fleiss Kappa). Our dataset, POQue (Participant Outcome Questions), enables the exploration and development of models that address multiple aspects of semantic understanding. Experimentally, we show that current language models lag behind human performance in subtle ways through our task formulations that target abstract and specific comprehension of a complex event, its outcome, and a participant’s influence over the event culmination. 

   more » « less
4. POQue: Asking Participant-specific Outcome Questions for a Deeper Understanding of Complex Events

   https://doi.org/10.18653/v1/2022.emnlp-main.594  

   Vallurupalli, Sai ; Ghosh, Sayontan ; Erk, Katrin ; Balasubramanian, Niranjan ; Ferraro, Francis ( December 2022 , Conference on Empirical Methods in Natural Language Processing (EMNLP))

   Knowledge about outcomes is critical for com- plex event understanding but is hard to acquire. We show that by pre-identifying a participant in a complex event, crowdworkers are able to (1) infer the collective impact of salient events that make up the situation, (2) annotate the volitional engagement of participants in causing the situation, and (3) ground the outcome of the situation in state changes of the participants. By creating a multi-step interface and a careful quality control strategy, we collect a high quality annotated dataset of 8K short newswire narratives and ROCStories with high inter-annotator agreement (0.74-0.96 weighted Fleiss Kappa). Our dataset, POQue (Participant Outcome Questions), enables the exploration and development of models that address multiple aspects of semantic understanding. Experimentally, we show that current language models lag behind human performance in subtle ways through our task formulations that target abstract and specific comprehension of a complex event, its outcome, and a participant’s influence over the event culmination. 

   more » « less
5. Learning to Hash Robustly, Guaranteed

   Andoni, Alexandr ; Beaglehole, Daniel ( July 2022 , International Conference on Machine Learning)

   The indexing algorithms for the high-dimensional nearest neighbor search (NNS) with the best worst-case guarantees are based on the randomized Locality Sensitive Hashing (LSH), and its derivatives. In practice, many heuristic approaches exist to "learn" the best indexing method in order to speed-up NNS, crucially adapting to the structure of the given dataset. Oftentimes, these heuristics outperform the LSH-based algorithms on real datasets, but, almost always, come at the cost of losing the guarantees of either correctness or robust performance on adversarial queries, or apply to datasets with an assumed extra structure/model. In this paper, we design an NNS algorithm for the Hamming space that has worst-case guarantees essentially matching that of theoretical algorithms, while optimizing the hashing to the structure of the dataset (think instance-optimal algorithms) for performance on the minimum-performing query. We evaluate the algorithm's ability to optimize for a given dataset both theoretically and practically. On the theoretical side, we exhibit a natural setting (dataset model) where our algorithm is much better than the standard theoretical one. On the practical side, we run experiments that show that our algorithm has a 1.8x and 2.1x better recall on the worst-performing queries to the MNIST and ImageNet datasets. https://arxiv.org/abs/2108.05433 

   more » « less