Neural methods for interacting with Named Entities and DatabasesThe challenging problem of using neural methods to extract information from databases is made even more difficult by the presence of many Named Entities (NE) in databases. Adding each NE to the vocabulary and learning individual representations for them is problematic because it causes an explosion in vocabulary size, and because in many cases the NEs occur only a small number of times. Additionally, it is common for new NEs to arise during test time, which leads to issues because they will be out of vocabulary words. We propose a general method for interacting with NEs, where each NE is encoded on the fly using a distributed representation and stored in a specialized NE table. This representation then acts as a key to retrieve the NE whenever required. We then propose a multiple-attention based method for retrieving information from large databases, which can use the above idea to handle the NEs present in them.
Translation of Complex Natural Language Queries into SQLIn natural language, the same idea can be expressed in many different ways: not only using different words but also using entirely different sentence structures. In databases too we can have radically different schema designs to represent the same information. Even against a specific schema, we can have multiple equivalent query expressions. This great variety makes it difficult to learn translations, and makes it unlikely that any rule-based translation that works for one natural language statement will work for another. To address these challenges, we propose to divide and conquer the problem. We use deep-learning to address variations in natural language vocabulary; we use past history to determine desired query structure; we use database statistics for entity disambiguation.
World Knowledge for Semantic Parsing with Abstract Meaning RepresentationIn our experiments, we introduce modifications to a parser for Abstract Meaning Representation that allow it to more accurately identify concepts and named entities in semantic sentence representations. We look at the types of errors that currently exist in a state of the art parser and explore the problem of how to integrate world knowledge to reduce these errors. An examination of the limitations of these types of features is included, which provides insight into the potential for world knowledge to benefit future work in AMR parsing.
Crowd Annotation of Expert Content
AI for disentanglement and conversation graph extractionDialogue often consists of multiple threads of conversation mixed together, with a complex structure in which an utterance could be responding to multiple previous utterances of which some might be far back in the conversation, and there could be multiple utterances that respond to a single one. One key step in learning to understand conversation is disentangling these threads. We intend to develop neural network based models that can extract the graph representing the conversation, along with a hand-labeled dataset for evaluation. We will focus on the Ubuntu corpus, but also consider smaller IRC datasets used in prior work.
H. Kaur, M. Gordon, Y. Yang, J.P. Bigham, J. Teevan, E. Kamar, W.S. Lasecki. CrowdMask: Using Crowds to Preserve Privacy in Crowd-Powered Systems via Progressive Filtering. In Proceedings of the AAAI Conference on Human Computation (HCOMP 2017). Quebec City, Canada.
Youxuan (Lucy) Jiang, Jonathan K. Kummerfeld, and Walter S. Lasecki. Understanding Task Design Trade-offs in Crowdsourced Paraphrase Collection. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (ACL 2017). Vancouver, Canada.
Xiaoxiao Guo, Tim Klinger, Clemens Rosenbaum, Joseph P. Bigus, Murray Campbell, Ban Kawas, Kartik Talamadupula, Gerald Tesauro, and Satinder Singh. Learning to Query, Reason, and Answer Questions On Ambiguous Texts. In Proceedings of the 6th International Conference on Learning Representations (ICLR 2017). Toulon, France.
Charles Welch and Rada Mihalcea. Targeted Sentiment to Understand Student Comments. In Proceedings of the 26th International Conference on Computational Linguistics (COLING 2016). Osaka, Japan.
Yelin Kim and Emily Mower Provost. ISLA: Temporal Segmentation and Labeling for Audio-Visual Emotion Recognition. IEEE Transactions on Affective Computing, vol: To appear, 2017.
Soheil Khorram, Zakariah Aldeneh, Dimitrios Dimitriadis, Melvin McInnis, and Emily Mower Provost. Capturing Long-term Temporal Dependencies with Convolutional Networks for Continuous Emotion. Interspeech. Stockholm, Sweden, August 2017.
John Gideon, Soheil Khorram, Zakariah Aldeneh, Dimitrios Dimitriadis, and Emily Mower Provost. Progressive Neural Networks for Transfer Learning in Emotion Recognition. Interspeech. Stockholm, Sweden, August 2017.
Duc Le, Zakariah Aldeneh, and Emily Mower Provost. Discretized Continuous Speech Emotion Recognition with Multi-Task Deep Recurrent Neural Network. Interspeech. Stockholm, Sweden, August 2017.
Quan Chen, Hailong Yang, Jason Mars, and Lingjia Tang. Baymax: Qos awareness and increased utilization for non-preemptive accelerators in warehouse scale computers. In ACM SIGPLAN Notices, vol. 51, no. 4, pp. 681-696. ACM, 2016.
Quan Chen, Hailong Yang, Minyi Guo, Ram Srivatsa Kannan, Jason Mars, and Lingjia Tang. Prophet: Precise QoS Prediction on Non-Preemptive Accelerators to Improve Utilization in Warehouse-Scale Computers. In Proceedings of the Twenty-Second International Conference on Architectural Support for Programming Languages and Operating Systems, pp. 17-32. ACM, 2017.
Hailong Yang, Quan Chen, Moeiz Riaz, Zhongzhi Luan, Lingjia Tang, and Jason Mars. PowerChief: Intelligent Power Allocation for Multi-Stage Applications to Improve Responsiveness on Power Constrained CMP. In Proceedings of the 44th Annual International Symposium on Computer Architecture, pp. 133-146. ACM, 2017.