Wang-Chiew Tan

Wang-Chiew Tan, Yuliang Li, Pedro Rodriguez et al. · meta-ai

We present a reality check on large language models and inspect the promise of retrieval augmented language models in comparison. Such language models are semi-parametric, where models integrate model parameters and knowledge from external data sources to make their predictions, as opposed to the parametric nature of vanilla large language models. We give initial experimental findings that semi-parametric architectures can be enhanced with views, a query analyzer/planner, and provenance to make a significantly more powerful system for question answering in terms of accuracy and efficiency, and potentially for other NLP tasks

Wang-Chiew Tan, Jane Dwivedi-Yu, Yuliang Li et al.

Lifelogs are descriptions of experiences that a person had during their life. Lifelogs are created by fusing data from the multitude of digital services, such as online photos, maps, shopping and content streaming services. Question answering over lifelogs can offer personal assistants a critical resource when they try to provide advice in context. However, obtaining answers to questions over lifelogs is beyond the current state of the art of question answering techniques for a variety of reasons, the most pronounced of which is that lifelogs combine free text with some degree of structure such as temporal and geographical information. We create and publicly release TimelineQA1, a benchmark for accelerating progress on querying lifelogs. TimelineQA generates lifelogs of imaginary people. The episodes in the lifelog range from major life episodes such as high school graduation to those that occur on a daily basis such as going for a run. We describe a set of experiments on TimelineQA with several state-of-the-art QA models. Our experiments reveal that for atomic queries, an extractive QA system significantly out-performs a state-of-the-art retrieval-augmented QA system. For multi-hop queries involving aggregates, we show that the best result is obtained with a state-of-the-art table QA technique, assuming the ground truth set of episodes for deriving the answer is available.

Wang-Chiew Tan

This paper presents an opinion on the potential of using large language models to query on both unstructured and structured data. It also outlines some research challenges related to the topic of building question-answering systems for both types of data.

Yoshihiko Suhara, Jinfeng Li, Yuliang Li et al.

Inferring meta information about tables, such as column headers or relationships between columns, is an active research topic in data management as we find many tables are missing some of this information. In this paper, we study the problem of annotating table columns (i.e., predicting column types and the relationships between columns) using only information from the table itself. We develop a multi-task learning framework (called Doduo) based on pre-trained language models, which takes the entire table as input and predicts column types/relations using a single model. Experimental results show that Doduo establishes new state-of-the-art performance on two benchmarks for the column type prediction and column relation prediction tasks with up to 4.0% and 11.9% improvements, respectively. We report that Doduo can already outperform the previous state-of-the-art performance with a minimal number of tokens, only 8 tokens per column. We release a toolbox (https://github.com/megagonlabs/doduo) and confirm the effectiveness of Doduo on a real-world data science problem through a case study.

Hayate Iso, Xiaolan Wang, Yoshihiko Suhara et al.

Recent advances in text autoencoders have significantly improved the quality of the latent space, which enables models to generate grammatical and consistent text from aggregated latent vectors. As a successful application of this property, unsupervised opinion summarization models generate a summary by decoding the aggregated latent vectors of inputs. More specifically, they perform the aggregation via simple average. However, little is known about how the vector aggregation step affects the generation quality. In this study, we revisit the commonly used simple average approach by examining the latent space and generated summaries. We found that text autoencoders tend to generate overly generic summaries from simply averaged latent vectors due to an unexpected $L_2$-norm shrinkage in the aggregated latent vectors, which we refer to as summary vector degeneration. To overcome this issue, we develop a framework Coop, which searches input combinations for the latent vector aggregation using input-output word overlap. Experimental results show that Coop successfully alleviates the summary vector degeneration issue and establishes new state-of-the-art performance on two opinion summarization benchmarks. Code is available at \url{https://github.com/megagonlabs/coop}.

Jinfeng Li, Yuliang Li, Xiaolan Wang et al.

Semantic tagging, which has extensive applications in text mining, predicts whether a given piece of text conveys the meaning of a given semantic tag. The problem of semantic tagging is largely solved with supervised learning and today, deep learning models are widely perceived to be better for semantic tagging. However, there is no comprehensive study supporting the popular belief. Practitioners often have to train different types of models for each semantic tagging task to identify the best model. This process is both expensive and inefficient. We embark on a systematic study to investigate the following question: Are deep models the best performing model for all semantic tagging tasks? To answer this question, we compare deep models against "simple models" over datasets with varying characteristics. Specifically, we select three prevalent deep models (i.e. CNN, LSTM, and BERT) and two simple models (i.e. LR and SVM), and compare their performance on the semantic tagging task over 21 datasets. Results show that the size, the label ratio, and the label cleanliness of a dataset significantly impact the quality of semantic tagging. Simple models achieve similar tagging quality to deep models on large datasets, but the runtime of simple models is much shorter. Moreover, simple models can achieve better tagging quality than deep models when targeting datasets show worse label cleanliness and/or more severe imbalance. Based on these findings, our study can systematically guide practitioners in selecting the right learning model for their semantic tagging task.

Nofar Carmeli, Xiaolan Wang, Yoshihiko Suhara et al.

The Web is a major resource of both factual and subjective information. While there are significant efforts to organize factual information into knowledge bases, there is much less work on organizing opinions, which are abundant in subjective data, into a structured format. We present ExplainIt, a system that extracts and organizes opinions into an opinion graph, which are useful for downstream applications such as generating explainable review summaries and facilitating search over opinion phrases. In such graphs, a node represents a set of semantically similar opinions extracted from reviews and an edge between two nodes signifies that one node explains the other. ExplainIt mines explanations in a supervised method and groups similar opinions together in a weakly supervised way before combining the clusters of opinions together with their explanation relationships into an opinion graph. We experimentally demonstrate that the explanation relationships generated in the opinion graph are of good quality and our labeled datasets for explanation mining and grouping opinions are publicly available.

Sainyam Galhotra, Behzad Golshan, Wang-Chiew Tan

Creating and collecting labeled data is one of the major bottlenecks in machine learning pipelines and the emergence of automated feature generation techniques such as deep learning, which typically requires a lot of training data, has further exacerbated the problem. While weak-supervision techniques have circumvented this bottleneck, existing frameworks either require users to write a set of diverse, high-quality rules to label data (e.g., Snorkel), or require a labeled subset of the data to automatically mine rules (e.g., Snuba). The process of manually writing rules can be tedious and time consuming. At the same time, creating a labeled subset of the data can be costly and even infeasible in imbalanced settings. This is due to the fact that a random sample in imbalanced settings often contains only a few positive instances. To address these shortcomings, we present Darwin, an interactive system designed to alleviate the task of writing rules for labeling text data in weakly-supervised settings. Given an initial labeling rule, Darwin automatically generates a set of candidate rules for the labeling task at hand, and utilizes the annotator's feedback to adapt the candidate rules. We describe how Darwin is scalable and versatile. It can operate over large text corpora (i.e., more than 1 million sentences) and supports a wide range of labeling functions (i.e., any function that can be specified using a context free grammar). Finally, we demonstrate with a suite of experiments over five real-world datasets that Darwin enables annotators to generate weakly-supervised labels efficiently and with a small cost. In fact, our experiments show that rules discovered by Darwin on average identify 40% more positive instances compared to Snuba even when it is provided with 1000 labeled instances.

Yoshihiko Suhara, Xiaolan Wang, Stefanos Angelidis et al.

We present OpinionDigest, an abstractive opinion summarization framework, which does not rely on gold-standard summaries for training. The framework uses an Aspect-based Sentiment Analysis model to extract opinion phrases from reviews, and trains a Transformer model to reconstruct the original reviews from these extractions. At summarization time, we merge extractions from multiple reviews and select the most popular ones. The selected opinions are used as input to the trained Transformer model, which verbalizes them into an opinion summary. OpinionDigest can also generate customized summaries, tailored to specific user needs, by filtering the selected opinions according to their aspect and/or sentiment. Automatic evaluation on Yelp data shows that our framework outperforms competitive baselines. Human studies on two corpora verify that OpinionDigest produces informative summaries and shows promising customization capabilities.

Johannes Bjerva, Nikita Bhutani, Behzad Golshan et al.

Subjectivity is the expression of internal opinions or beliefs which cannot be objectively observed or verified, and has been shown to be important for sentiment analysis and word-sense disambiguation. Furthermore, subjectivity is an important aspect of user-generated data. In spite of this, subjectivity has not been investigated in contexts where such data is widespread, such as in question answering (QA). We therefore investigate the relationship between subjectivity and QA, while developing a new dataset. We compare and contrast with analyses from previous work, and verify that findings regarding subjectivity still hold when using recently developed NLP architectures. We find that subjectivity is also an important feature in the case of QA, albeit with more intricate interactions between subjectivity and QA performance. For instance, a subjective question may or may not be associated with a subjective answer. We release an English QA dataset (SubjQA) based on customer reviews, containing subjectivity annotations for questions and answer spans across 6 distinct domains.

Aaron Traylor, Chen Chen, Behzad Golshan et al.

Review comprehension has played an increasingly important role in improving the quality of online services and products and commonsense knowledge can further enhance review comprehension. However, existing general-purpose commonsense knowledge bases lack sufficient coverage and precision to meaningfully improve the comprehension of domain-specific reviews. In this paper, we introduce xSense, an effective system for review comprehension using domain-specific commonsense knowledge bases (xSense KBs). We show that xSense KBs can be constructed inexpensively and present a knowledge distillation method that enables us to use xSense KBs along with BERT to boost the performance of various review comprehension tasks. We evaluate xSense over three review comprehension tasks: aspect extraction, aspect sentiment classification, and question answering. We find that xSense outperforms the state-of-the-art models for the first two tasks and improves the baseline BERT QA model significantly, demonstrating the usefulness of incorporating commonsense into review comprehension pipelines. To facilitate future research and applications, we publicly release three domain-specific knowledge bases and a domain-specific question answering benchmark along with this paper.

Yuliang Li, Jinfeng Li, Yoshihiko Suhara et al.

We present Ditto, a novel entity matching system based on pre-trained Transformer-based language models. We fine-tune and cast EM as a sequence-pair classification problem to leverage such models with a simple architecture. Our experiments show that a straightforward application of language models such as BERT, DistilBERT, or RoBERTa pre-trained on large text corpora already significantly improves the matching quality and outperforms previous state-of-the-art (SOTA), by up to 29% of F1 score on benchmark datasets. We also developed three optimization techniques to further improve Ditto's matching capability. Ditto allows domain knowledge to be injected by highlighting important pieces of input information that may be of interest when making matching decisions. Ditto also summarizes strings that are too long so that only the essential information is retained and used for EM. Finally, Ditto adapts a SOTA technique on data augmentation for text to EM to augment the training data with (difficult) examples. This way, Ditto is forced to learn "harder" to improve the model's matching capability. The optimizations we developed further boost the performance of Ditto by up to 9.8%. Perhaps more surprisingly, we establish that Ditto can achieve the previous SOTA results with at most half the number of labeled data. Finally, we demonstrate Ditto's effectiveness on a real-world large-scale EM task. On matching two company datasets consisting of 789K and 412K records, Ditto achieves a high F1 score of 96.5%.

Aaron Feng, Shuwei Chen, Yuliang Li et al.

Existing e-commerce search engines typically support search only over objective attributes, such as price and locations, leaving the more desirable subjective attributes, such as romantic vibe and worklife balance unsearchable. We found that this is also the case for Recruit Group, which operates a wide range of online booking and search services, including jobs, travel, housing, bridal, dining, beauty, and where each service is among the biggest in Japan, if not internationally. We present our progress towards productionizing a recent subjective search prototype (OpineDB) developed by Megagon Labs for Recruit Group. Several components within OpineDB are enhanced to satisfy production demands, including adding a BERT language model pre-trained on massive hospitality domain review corpora. We also found that the challenges of productionizing the system are beyond enhancing the components. In particular, an important requirement in production-quality systems is to instrument a proper way of measuring the search quality, which is extremely tricky when the search results are subjective. This led to the creation of a high-quality benchmark dataset from scratch, involving over 600 queries by user interviews and a collection of more than 120,000 query-entity relevancy labels. Also, we found that the existing search algorithms do not meet the search quality standard required by production systems. Consequently, we enhanced the ranking model by fine-tuning several search algorithms and combining them under a learning-to-rank framework. The model achieves 5%-10% overall precision improvement and 90+% precision on more than half of the benchmark testing queries making these queries ready for AB-testing. While some enhancements can be immediately applied to other verticals, our experience reveals that benchmarking and fine-tuning ranking algorithms are specific to each domain and cannot be avoided.

Zhengjie Miao, Yuliang Li, Xiaolan Wang et al.

Online services are interested in solutions to opinion mining, which is the problem of extracting aspects, opinions, and sentiments from text. One method to mine opinions is to leverage the recent success of pre-trained language models which can be fine-tuned to obtain high-quality extractions from reviews. However, fine-tuning language models still requires a non-trivial amount of training data. In this paper, we study the problem of how to significantly reduce the amount of labeled training data required in fine-tuning language models for opinion mining. We describe Snippext, an opinion mining system developed over a language model that is fine-tuned through semi-supervised learning with augmented data. A novelty of Snippext is its clever use of a two-prong approach to achieve state-of-the-art (SOTA) performance with little labeled training data through: (1) data augmentation to automatically generate more labeled training data from existing ones, and (2) a semi-supervised learning technique to leverage the massive amount of unlabeled data in addition to the (limited amount of) labeled data. We show with extensive experiments that Snippext performs comparably and can even exceed previous SOTA results on several opinion mining tasks with only half the training data required. Furthermore, it achieves new SOTA results when all training data are leveraged. By comparison to a baseline pipeline, we found that Snippext extracts significantly more fine-grained opinions which enable new opportunities of downstream applications.

Xiong Zhang, Jonathan Engel, Sara Evensen et al.

Reviews are integral to e-commerce services and products. They contain a wealth of information about the opinions and experiences of users, which can help better understand consumer decisions and improve user experience with products and services. Today, data scientists analyze reviews by developing rules and models to extract, aggregate, and understand information embedded in the review text. However, working with thousands of reviews, which are typically noisy incomplete text, can be daunting without proper tools. Here we first contribute results from an interview study that we conducted with fifteen data scientists who work with review text, providing insights into their practices and challenges. Results suggest data scientists need interactive systems for many review analysis tasks. In response we introduce Teddy, an interactive system that enables data scientists to quickly obtain insights from reviews and improve their extraction and modeling pipelines.

Dan Zhang, Yoshihiko Suhara, Jinfeng Li et al.

Detecting the semantic types of data columns in relational tables is important for various data preparation and information retrieval tasks such as data cleaning, schema matching, data discovery, and semantic search. However, existing detection approaches either perform poorly with dirty data, support only a limited number of semantic types, fail to incorporate the table context of columns or rely on large sample sizes for training data. We introduce Sato, a hybrid machine learning model to automatically detect the semantic types of columns in tables, exploiting the signals from the context as well as the column values. Sato combines a deep learning model trained on a large-scale table corpus with topic modeling and structured prediction to achieve support-weighted and macro average F1 scores of 0.925 and 0.735, respectively, exceeding the state-of-the-art performance by a significant margin. We extensively analyze the overall and per-type performance of Sato, discussing how individual modeling components, as well as feature categories, contribute to its performance.

Danni Ma, Chen Chen, Behzad Golshan et al.

Paraphrases are important linguistic resources for a wide variety of NLP applications. Many techniques for automatic paraphrase mining from general corpora have been proposed. While these techniques are successful at discovering generic paraphrases, they often fail to identify domain-specific paraphrases (e.g., {staff, concierge} in the hospitality domain). This is because current techniques are often based on statistical methods, while domain-specific corpora are too small to fit statistical methods. In this paper, we present an unsupervised graph-based technique to mine paraphrases from a small set of sentences that roughly share the same topic or intent. Our system, Essentia, relies on word-alignment techniques to create a word-alignment graph that merges and organizes tokens from input sentences. The resulting graph is then used to generate candidate paraphrases. We demonstrate that our system obtains high-quality paraphrases, as evaluated by crowd workers. We further show that the majority of the identified paraphrases are domain-specific and thus complement existing paraphrase databases.

Wataru Hirota, Yoshihiko Suhara, Behzad Golshan et al.

We present Emu, a system that semantically enhances multilingual sentence embeddings. Our framework fine-tunes pre-trained multilingual sentence embeddings using two main components: a semantic classifier and a language discriminator. The semantic classifier improves the semantic similarity of related sentences, whereas the language discriminator enhances the multilinguality of the embeddings via multilingual adversarial training. Our experimental results based on several language pairs show that our specialized embeddings outperform the state-of-the-art multilingual sentence embedding model on the task of cross-lingual intent classification using only monolingual labeled data.

Sara Evensen, Yoshihiko Suhara, Alon Halevy et al.

Understanding what makes people happy is a central topic in psychology. Prior work has mostly focused on developing self-reporting assessment tools for individuals and relies on experts to analyze the periodic reported assessments. One of the goals of the analysis is to understand what actions are necessary to encourage modifications in the behaviors of the individuals to improve their overall well-being. In this paper, we outline a complementary approach; on the assumption that the user journals her happy moments as short texts, a system can analyze these texts and propose sustainable suggestions for the user that may lead to an overall improvement in her well-being. We prototype one necessary component of such a system, the Happiness Entailment Recognition (HER) module, which takes as input a short text describing an event, a candidate suggestion, and outputs a determination about whether the suggestion is more likely to be good for this user based on the event described. This component is implemented as a neural network model with two encoders, one for the user input and one for the candidate actionable suggestion, with additional layers to capture psychologically significant features in the happy moment and suggestion.

Vivian Li, Alon Halevy, Adi Zief-Balteriski Ph. D et al.

We introduce Jo, a mobile application that attempts to improve user's well-being. Jo is a journaling application--users log their important moments via short texts and optionally an attached photo. Unlike a static journal, Jo analyzes these moments and helps users take action towards increased well-being. For example, Jo annotates each moment with a set of values (e.g., family, socialization, mindfulness), thereby giving the user insights about the balance in their lives. In addition, Jo helps the user create reminders that enable them to create additional happy moments. We describe the results of fielding Jo in a study of 39 participants. The results illustrate the promise of a journaling application that provides personalized feedback, and points at further research.

Sara Evensen, Aaron Feng, Alon Halevy et al.

We describe Voyageur, which is an application of experiential search to the domain of travel. Unlike traditional search engines for online services, experiential search focuses on the experiential aspects of the service under consideration. In particular, Voyageur needs to handle queries for subjective aspects of the service (e.g., quiet hotel, friendly staff) and combine these with objective attributes, such as price and location. Voyageur also highlights interesting facts and tips about the services the user is considering to provide them with further insights into their choices.

Nikita Bhutani, Yoshihiko Suhara, Wang-Chiew Tan et al.

Open Information Extraction (OpenIE) extracts meaningful structured tuples from free-form text. Most previous work on OpenIE considers extracting data from one sentence at a time. We describe NeurON, a system for extracting tuples from question-answer pairs. Since real questions and answers often contain precisely the information that users care about, such information is particularly desirable to extend a knowledge base with. NeurON addresses several challenges. First, an answer text is often hard to understand without knowing the question, and second, relevant information can span multiple sentences. To address these, NeurON formulates extraction as a multi-source sequence-to-sequence learning task, wherein it combines distributed representations of a question and an answer to generate knowledge facts. We describe experiments on two real-world datasets that demonstrate that NeurON can find a significant number of new and interesting facts to extend a knowledge base compared to state-of-the-art OpenIE methods.

Xiaolan Wang, Aaron Feng, Behzad Golshan et al.

We present the KOKO system that takes declarative information extraction to a new level by incorporating advances in natural language processing techniques in its extraction language. KOKO is novel in that its extraction language simultaneously supports conditions on the surface of the text and on the structure of the dependency parse tree of sentences, thereby allowing for more refined extractions. KOKO also supports conditions that are forgiving to linguistic variation of expressing concepts and allows to aggregate evidence from the entire document in order to filter extractions. To scale up, KOKO exploits a multi-indexing scheme and heuristics for efficient extractions. We extensively evaluate KOKO over publicly available text corpora. We show that KOKO indices take up the smallest amount of space, are notably faster and more effective than a number of prior indexing schemes. Finally, we demonstrate KOKO's scale up on a corpus of 5 million Wikipedia articles.

Akari Asai, Sara Evensen, Behzad Golshan et al.

The science of happiness is an area of positive psychology concerned with understanding what behaviors make people happy in a sustainable fashion. Recently, there has been interest in developing technologies that help incorporate the findings of the science of happiness into users' daily lives by steering them towards behaviors that increase happiness. With the goal of building technology that can understand how people express their happy moments in text, we crowd-sourced HappyDB, a corpus of 100,000 happy moments that we make publicly available. This paper describes HappyDB and its properties, and outlines several important NLP problems that can be studied with the help of the corpus. We also apply several state-of-the-art analysis techniques to analyze HappyDB. Our results demonstrate the need for deeper NLP techniques to be developed which makes HappyDB an exciting resource for follow-on research.