The modern-day research community has an embarrassment of riches regarding pre-trained AI models. Even for a simple task such as lexicon set expansion, where an AI model suggests new entities to add to a predefined seed set of entities, thousands of models are available. However, deciding which model to use for a given set expansion task is non-trivial. In hindsight, some models can be 'off topic' for specific set expansion tasks, while others might work well initially but quickly exhaust what they have to offer. Additionally, certain models may require more careful priming in the form of samples or feedback before being finetuned to the task at hand. In this work, we frame this model selection as a sequential non-stationary problem, where there exist a large number of diverse pre-trained models that may or may not fit a task at hand, and an expert is shown one suggestion at a time to include in the set or not, i.e., accept or reject the suggestion. The goal is to expand the list with the most entities as quickly as possible. We introduce MArBLE, a hierarchical multi-armed bandit method for this task, and two strategies designed to address cold-start problems. Experimental results on three set expansion tasks demonstrate MArBLE's effectiveness compared to baselines.

Recent advances in text representation have shown that training on large amounts of text is crucial for natural language understanding. However, models trained without predefined notions of topical interest typically require careful fine-tuning when transferred to specialized domains. When a sufficient amount of within-domain text may not be available, expanding a seed corpus of relevant documents from large-scale web data poses several challenges. First, corpus expansion requires scoring and ranking each document in the collection, an operation that can quickly become computationally expensive as the web corpora size grows. Relying on dense vector spaces and pairwise similarity adds to the computational expense. Secondly, as the domain concept becomes more nuanced, capturing the long tail of domain-specific rare terms becomes non-trivial, especially under limited seed corpora scenarios.

The use of AI in knowledge dense domains, e.g., chemistry, medicine, biology, etc. - is extremely promising, but often suffers from slow deployment and adaptation to different tasks. We propose a methodology to quickly capture the intent and expertise of a domain expert in order to train personalized AI models for specific tasks. Specifically we focus on the domain of polymer materials design and discovery: it often takes 10 years or more to design, synthesize, test, and introduce a new polymer material into the market. One way to accelerate up the design of polymer materials is through the use of computational methods to design the material, such as combinatorial screening, generative models, inverse design, etc. The drawback of these methods is that they generate a large number of candidates for new molecules, which then need to be manually reviewed by subject matter experts who select only a dozen for further investigation. Our solution is a human-in-the-loop methodology where we rank the candidates according to a utility function that is learned via the continued interaction with the subject matter experts, but which is also constrained by specific chemical knowledge. We prove the viability of our proposed methodology in a polymer production lab and we (i) evaluate against datasets of polymers previously produced in the lab as well as (ii) producing several novel materials that are undergoing experimental development, and (iii) quantitatively show that standard synthetic accessibility scores do not inform about patterns of SME decisions.

Many Knowledge Extraction systems rely on semantic resources - dictionaries, ontologies, lexical resources - to extract information from unstructured text. A key for successful information extraction is to consider such resources as evolving artifacts and keep them up-to-date. In this paper, we tackle the problem of dictionary expansion and we propose a human-in-the-loop approach: we couple neural language models with tight human supervision to assist the user in building and maintaining domain-specific dictionaries. The approach works on any given input text corpus and is based on the explore and exploit paradigm: starting from a few seeds (or an existing dictionary) it effectively discovers new instances (explore) from the text corpus as well as predicts new potential instances which are not in the corpus, i.e. “unseen” , using the current dictionary entries (exploit). We evaluate our approach on five real-world dictionaries, achieving high accuracy with a rapid expansion rate.

In many Information Extraction tasks, dictionaries and lexica are powerful building blocks for sophisticated extractions. The success of the Semantic Web in the last 10 years has produced an unprecedented quantity of available structured data that can be leveraged to produce dictionaries on countless concepts in many domains. While being an invaluable resource, these automatically built dictionaries may contain "problematic" items, such as spurious words, which have been included by mistake, or ambiguous words, which appear with multiple different meanings in the target corpus and therefore necessitating an expensive disambiguation task. In this paper, we propose a simple and effective method to identify problematic terms in a given dictionary, which are ambiguous or spurious with respect to a given corpus, with the aim to facilitate subsequent Information Extraction tasks. We prove the effectiveness of the method with a systematic experiment on publicly available concept dictionaries, using a very large Web corpus as target, with an average precision in identifying a problem term above 85%.

Recognizing mentions of Adverse Drug Reactions (ADR) in social media is challenging: ADR mentions are contextdependent and include long, varied and unconventional descriptions as compared to more formal medical symptom terminology.We use the CADEC corpus to train a recurrent neural network (RNN) transducer, integrated with knowledge graph embeddings of DBpedia, and show the resulting model to be highly accurate (93.4 F1).Furthermore, even when lacking high quality expert annotations, we show that by employing an active learning technique and using purpose built annotation tools, we can train the RNN to perform well (83.9F1).

Ontologies are dynamic artifacts that evolve both in structure and content. Keeping them up-to-date is a very expensive and critical operation for any application relying on semantic Web technologies. In this paper we focus on evolving the content of an ontology by extracting relevant instances of ontological concepts from text. We propose a novel technique which is (i) completely language independent, (ii) combines statistical methods with human-in-the-loop and (iii) exploits Linked Data as bootstrapping source. Our experiments on a publicly available medical corpus and on a Twitter dataset show that the proposed solution achieves comparable performances regardless of language, domain and style of text. Given that the method relies on a human-in-the-loop, our results can be safely fed directly back into Linked Data resources.

Data analysis tasks often require grouping of information to identify trends and associations. However, as the number of elements rises to the hundreds and thousands the cost of having a person perform the groupings unassisted quickly becomes prohibitive. Previous approaches have combined traditional clustering techniques with manual interaction steps, yielding human-in-the-loop clustering algorithms that incorporate user feedback by reweighting features or adjusting a similarity function. But in the real world, many grouping tasks lack both a feature set and a well-defined (dis)similarity metric, having only a subject matter expert with an implicit understanding of the correct relationships between elements based on the domain and the task at hand. We present a refine-and-lock clustering interaction model and demonstrate its effectiveness for cognitive-assisted human clustering over other interaction models such as split/merge and must-link/can't-link. Our approach offers effective automatic clustering assistance even in the absence of clear features or a definitive similarity metric; ensures that every cluster has final user approval; and exhibits at least a 3.94× improvement over other interactive clustering approaches in time to completion.

With the rise of social media, learning from informal text has become increasingly important. We present a novel semantic lexicon induction approach that is able to learn new vocabulary from social media. Our method is robust to the idiosyncrasies of informal and open-domain text corpora. Unlike previous work, it does not impose restrictions on the lexical features of candidate terms — e.g. by restricting entries to nouns or noun phrases —while still being able to accurately learn multiword phrases of variable length. Starting with a few seed terms for a semantic category, our method first explores the context around seed terms in a corpus, and identifies context patterns that are relevant to the category. These patterns are used to extract candidate terms — i.e. multiword segments that are further analyzed to ensure meaningful term boundary segmentation. We show that our approach is able to learn high quality semantic lexicons from informally written social media text of Twitter, and can achieve accuracy as high as 92% in the top 100 learned category members.

There is still significant investment in legacy healthcare information technology (HIT). Current systems are a diverse mix of technologies, standards, platforms and versions. Many of which were never intended to be used together to achieve a common goal. In order to deliver care effectively and effi

The ever increasing amount of content on the Internet has fostered many efforts seeking to leverage this potentially yottascale information source. Service systems using advanced data and text analytics techniques have been developed to perform knowledge gathering and information discovery over Web data. Information gathered from free and public sources on the Web is frequently integrated with enterprise and proprietary data to create sophisticated service systems able to provide insight in an increasing number of business critical areas. Unfortunately, for fixed and or limited resource projects, consistent and reliable ingestion and integration of content often dominates the effort, reducing the time available for developing core analytics and presentations that differentiate and define an information service. If this initial data extraction, translation and loading of information (known as ETL in the database world) can be abstracted for these web sources, it would provide an important core technology on which Web-based information services could be more rapidly and inexpensively developed and deployed. This paper presents such a system - MONGOOSE - an approach that seeks to reduce the time spent creating a reliable data ingest and integration system and thus reducing the time-to-impact of advanced analytics service solutions.

There are a large number of Web sites serving valuable content that can be used by higher-level applications, Web services, Mashups, etc. Yet, due to various reasons (lack of computing resources, financial constraints etc.) they are unable to provide Web service APIs to access their data. In their desire to incorporate the latest and greatest technologies, as well as to adapt layouts that are more preferred by users, Web sites undergo changes over time. These changes can range from minor, e.g. function name changes, to major, e.g., shifting the Web platform to AJAX technologies. This paper addresses the problem of detecting layout changes for Web sites which are unable to provide any Web service to access their content, yet do not mind others harvesting said content.

Popularity based search engines have served to stagnate information retrieval from the web. Developed to deal with the very real problem of degrading quality within keyword based search they have had the unintended side effect of creating "icebergs" around topics, where only a small minority of the information is above the popularity water-line. This problem is especially pronounced with emerging information--new sites are often hidden until they become popular enough to be considered above the water-line. In domains new to a user this is often helpful--they can focus on popular sites first. Unfortunately it is not the best tool for a professional seeking to keep up-to-date with a topic as it emerges and evolves.We present a tool focused on this audience--a system that addresses the very large scale information gathering, filtering and routing, and presentation problems associated with creating a useful incremental stream of information from the web as a whole. Utilizing the WebFountain platform as the primary data engine and Really Simple Syndication (RSS) as the delivery mechanism, our "Daily Deltas" (Delta) application is able to provide an informative feed of relevant content directly to a user. Individuals receive a personalized, incremental feed of pages related to their topic allowing them to track their interests independent of the overall popularity of the topic.

XML provides a universal and portable format for document and data exchange. While the syntax and specification of XML makes documents both human readable and machine parsable, it is often at the expense of efficiency when representing simple data structures.We investigate the ``costs'' associated with XML serialization from several resource perspectives: storage, transport, processing and human readability. These experiments are done within the context of a large text-centric service oriented architecture -- IBM's WebFountain project.We find that for several applications, human readable formats outperform binary equivalents, especially in the area of data size, and that the costs of processing encoded binary data often exceeds that of processing terse human readable formats.

An increasing fraction of the global discourse is migrating online in the form of blogs, bulletin boards, web pages, wikis, editorials, and a dizzying array of new collaborative technologies. The migration has now proceeded to the point that topics reflecting certain individual products are sufficiently popular to allow targeted online tracking of the ebb and flow of chatter around these topics. Based on an analysis of around half a million sales rank values for 2,340 books over a period of four months, and correlating postings in blogs, media, and web pages, we are able to draw several interesting conclusions.First, carefully hand-crafted queries produce matching postings whose volume predicts sales ranks. Second, these queries can be automatically generated in many cases. And third, even though sales rank motion might be difficult to predict in general, algorithmic predictors can use online postings to successfully predict spikes in sales rank.

This paper provides an objective evaluation of the performance impacts of binary XML encodings, using a fast stream-based XQuery processor as our representative application. Instead of proposing one binary format and comparing it against standard XML parsers, we investigate the individual effects of several binary encoding techniques that are shared by many proposals. Our goal is to provide a deeper understanding of the performance impacts of binary XML encodings in order to clarify the ongoing and often contentious debate over their merits, particularly in the domain of high performance XML stream processing.

We study the dynamics of information propagation in environments of low-overhead personal publishing, using a large collection of weblogs over time as our example domain. We characterize and model this collection at two levels. First, we present a macroscopic characterization of topic propagation through our corpus, formalizing the notion of long-running "chatter" topics consisting recursively of "spike" topics generated by outside world events, or more rarely, by resonances within the community. Second, we present a microscopic characterization of propagation from individual to individual, drawing on the theory of infectious diseases to model the flow. We propose, validate, and employ an algorithm to induce the underlying propagation network from a sequence of posts, and report on the results.

This paper describes Seeker, a platform for large-scale text analytics, and SemTag, an application written on the platform to perform automated semantic tagging of large corpora. We apply SemTag to a collection of approximately 264 million web pages, and generate approximately 434 million automatically disambiguated semantic tags, published to the web as a label bureau providing metadata regarding the 434 million annotations. To our knowledge, this is the largest scale semantic tagging effort to date.We describe the Seeker platform, discuss the architecture of the SemTag application, describe a new disambiguation algorithm specialized to support ontological disambiguation of large-scale data, evaluate the algorithm, and present our final results with information about acquiring and making use of the semantic tags. We argue that automated large scale semantic tagging of ambiguous content can bootstrap and accelerate the creation of the semantic web.

YouServ is a system that allows its users to pool existing desktop computing resources for high availability web hosting and file sharing. By exploiting standard web and internet protocols (e.g. HTTP and DNS), YouServ does not require those who access YouServ-published content to install special purpose software. Because it requires minimal server-side resources and administration, YouServ can be provided at a very low cost. We describe the design, implementation, and a successful intranet deployment of the YouServ system, and compare it with several alternatives.

Article Share on Vinci: a service-oriented architecture for rapid development of web applications Authors: Rakesh Agrawal IBM Almaden Research Center, 650 Harry Road, San Jose, CA IBM Almaden Research Center, 650 Harry Road, San Jose, CAView Profile , Roberto J. Bayardo IBM Almaden Research Center, 650 Harry Road, San Jose, CA IBM Almaden Research Center, 650 Harry Road, San Jose, CAView Profile , Daniel Gruhl IBM Almaden Research Center, 650 Harry Road, San Jose, CA IBM Almaden Research Center, 650 Harry Road, San Jose, CAView Profile , Spiros Papadimitriou Carnegie Mellon University, Dept. of Computer Science and IBM Almaden Research Center, 650 Harry Road, San Jose, CA Carnegie Mellon University, Dept. of Computer Science and IBM Almaden Research Center, 650 Harry Road, San Jose, CAView Profile Authors Info & Claims WWW '01: Proceedings of the 10th international conference on World Wide WebMay 2001Pages 355–365https://doi.org/10.1145/371920.372088Published:01 April 2001Publication History 24citation2,443DownloadsMetricsTotal Citations24Total Downloads2,443Last 12 Months6Last 6 weeks3 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access