In this talk, I survey a small, non-random sample of research projects in information access carried out as part of the Thomson Reuters family of companies over the course of a 10+-year period. I analyse into how these projects are similar and different when compared to academic research efforts and attempt a critical (and personal, so certainly subjective) assessment of what academia can do for industry, and what industry can do for research in terms of R&D efforts. I will conclude with some advice for academic-industry collaboration initiatives in several areas of vertical information services (legal, finance, pharma and regulatory/compliance) as well as news.

The automatic extraction of breaking news events from natural language text is a valuable capability for decision support systems. Traditional systems tend to focus on extracting events from a single media source and often ignore cross-media references. Here, we describe a large-scale automated system for extracting natural disasters and critical events from both newswire text and social media. We outline a comprehensive architecture that can identify, categorize and summarize seven different event types - namely floods, storms, fires, armed conflict, terrorism, infrastructure breakdown, and labour unavailability. The system comprises fourteen modules and is equipped with a novel coreference mechanism, capable of linking events extracted from the two complementary data sources. Additionally, the system is easily extensible to accommodate new event types. Our experimental evaluation demonstrates the effectiveness of the system.

We present a contrastive study of document-level event classification of a range of seven different event types, namely floods, storms, fires, armed conflict, terrorism, infrastructure breakdown and labour unavailability from English-language news. Our study compares different supervised classification approaches, namely Support Vector Machine (SVM), Random Forest (RF), Convolutional Neural Network (CNN) and Hierarchical Attention Network (HAN). While past systems for Topic Detection and Tracking (TDT) and event extraction have proposed different machine learning models, to date SVMs, RFs, CNNs and HANs have not been compared on this task. Our classifiers are also informed by word embeddings trained on large amounts of high-quality agency news, which leads to improvements compared to the use of pre-trained embedding vectors. We report a detailed quantitative error analysis.

Financial and economic data are typically available in the form of tables and comprise mostly of monetary amounts, numeric and other domain-specific fields. They can be very hard to search and they are often made available out of context, or in forms which cannot be integrated with systems where text is required, such as voice-enabled devices. This work presents a novel system that enables both experts in the finance domain and non-expert users to search financial data with both keyword and natural language queries. Our system answers the queries with an automatically generated textual description using Natural Language Generation (NLG). The answers are further enriched with derived information, not explicitly asked in the user query, to provide the context of the answer. The system is designed to be flexible in order to accommodate new use cases without significant development effort, thus allowing fast integration of new datasets.

Financial fines imposed by regulatory bodies to penalize illegal activities and violations against regulations (cases of non-compliance) have recently become more common, and the sizes of fines have increased. This development coincides with the ongoing increase of complexity of regulatory rules. Huge fines have been imposed on banks for financial fraud and regulations have been made more stringent after 9/11 to curb funding of terrorist groups. Market players would also like to have available a database of fine events for a range of applications, such as to benchmark their competitors performance, or to use it as an early warning system for detecting shifts in regulators' enforcement behavior. To this end, we introduce the task of extracting fines from regulatory enforcement actions and we present a method to extract such fine event instances from timeline-like descriptions of regulatory investigation activities authored by legal professionals for a commercial product. We evaluate how well a rule-based method can extract information about fine events and we compare its performance to a machine-learning baseline. To the best of our knowledge, this work is the first one addressing this task.

It is often argued that in information extraction (IE), certain machine learning (ML) approaches save development time over others, or that certain ML methods (e.g. Active Learning) require less training data than others, thus saving development cost. However, such development cost claims are not normally backed up by controlled studies which show that such development cost savings actually occur. This situation in Language Engineering is contrasted with Software Engineering in general, where a lot of studies investigating system development cost have been carried out. We argue for the need of controlled studies that measure actual system development time in language engineering. To this end, we carry out an experiment in resource monitoring for an IE task: three named entity taggers for the same "surprise" domain are developed in parallel, using competing methods. Their human development time is accounted forusing a logging facility.We report development cost results for parallel implementations of a named entity tagger and present a breakdown of the development time for the three alternative methods. We are not aware of detailed previous parallel studies that detail how system development time is spent when creating a named entity tagger.

In question answering, two main kinds of matching methods for finding answer sentences for a question are term-based approaches -- which are simple, efficient, effective, and yield high recall -- and event-based approaches that take syntactic and semantic information into account. The latter often sacrifice recall for increased precision, but actually capture the meaning of the events denoted by the textual units of a passage or sentence. We propose a robust, data-driven method that learns the mapping between questions and answers using logistic regression and show that combining term-based and event-based approaches significantly outperforms the individual methods.

Web search using stationary (desktop) computers has become a pervasive activity. The mobile user in need of information, however, faces several problems in his or her quest to satisfy an information need. Mobile devices have small displays, and mobile user interfaces are often less then usable, because they impose the desktop Web search paradigm on the mobile user. We present a wireless search engine based on natural language queries transmitted via popular Small Message Service (SMS) text messages. Besides traditional keyword based queries, the system can accept questions or phrases and returns responses that contain likely answers (Figure 1) instead of traditional lists of hyperlinks. The additional precision gained from performing a linguistic analysis of the query helps extracting answers from Web pages directly, which requires no navigation. The system is implemented using a NLIR system residing on a server, which can translate questions or phrases into search engine queries or queries to SOAP Web services, where a gateway mediates between the mobile network and the Internet (Figure 2). Whereas on the desktop keyboard-based search still prevails, we find that in a mobile context question answering techniques can help overcome the output constraints.

Named entity tagging comprises the sub-tasks of identifying a text span and classifying it, but this view ignores the relationship between the entities and the world. Spatial and temporal entities ground events in space-time, and this relationship is vital for applications such as question answering and event tracking. There is much recent work regarding the temporal dimension (Setzer and Gaizauskas 2002, Mani and Wilson 2000), but no detailed study of the spatial dimension.I propose to investigate how spatial named entities (which are often referentially ambiguous) can be automatically resolved with respect to an extensional coordinate model (toponym resolution). To this end, various information sources including linguistic cue patterns, co-occurrence information, discourse/positional information, world knowledge (such as size and population) as well as minimality heuristics (Leidner et al. 2003) will be combined in a supervised machine learning regime.The major contributions of this research project will be a corpus of text manually annotated for spatial named entities with their model correlates as a training and evaluation resource, a novel method to spatially ground toponyms in text and a component-based evaluation based on this new reference corpus.