Search

The Web and social media have become the main source of information for citizens, with the risk that users rely on shallow information in sources prioritizing commercial or political incentives rather than the correctness and informational value. Non-experts tend to avoid scientific literature due to its complex language or their lack of prior background knowledge. Text simplification promises to remove some of these barriers. The CLEF 2022 SimpleText track addresses the challenges of text simplification approaches in the context of promoting scientific information access, by providing appropriate data and benchmarks, and creating a community of NLP and IR researchers working together to resolve one of the greatest challenges of today. The track will use a corpus of scientific literature abstracts and popular science requests. It features three tasks. First, content selection (what is in, or out?) challenges systems to select passages to include in a simplified summary in response to a query. Second, complexity spotting (what is unclear?) given a passage and a query, aims to rank terms/concepts that are required to be explained for understanding this passage (definitions, context, applications). Third, text simplification (rewrite this!) given a query, asks to simplify passages from scientific abstracts while preserving the main content.

In recent years, large pre-trained transformers have led to substantial gains in performance over traditional retrieval models and feedback approaches. However, these results are primarily based on the MS Marco/TREC Deep Learning Track setup, with its very particular setup, and our understanding of why and how these models work better is fragmented at best. We analyze effective BERT-based cross-encoders versus traditional BM25 ranking for the passage retrieval task where the largest gains have been observed, and investigate two main questions. On the one hand, what is similar? To what extent does the neural ranker already encompass the capacity of traditional rankers? Is the gain in performance due to a better ranking of the same documents (prioritizing precision)? On the other hand, what is different? Can it retrieve effectively documents missed by traditional systems (prioritizing recall)? We discover substantial differences in the notion of relevance identifying strengths and weaknesses of BERT that may inspire research for future improvement. Our results contribute to our understanding of (black-box) neural rankers relative to (well-understood) traditional rankers, help understand the particular experimental setting of MS-Marco-based test collections.

Various conceptual and descriptive models of conversational search have been proposed in the literature – while useful, they do not provide insights into how interaction between the agent and user would change in response to the costs and benefits of the different interactions. In this paper, we develop two economic models of conversational search based on patterns previously observed during conversational search sessions, which we refer to as: Feedback First where the agent asks clarifying questions then presents results, and Feedback After where the agent presents results, and then asks follow up questions. Our models show that the amount of feedback given/requested depends on its efficiency at improving the initial or subsequent query and the relative cost of providing said feedback. This theoretical framework for conversational search provides a number of insights that can be used to guide and inform the development of conversational search agents. However, empirical work is needed to estimate the parameters in order to make predictions specific to a given conversational search setting.

While much recent work has demonstrated that hard negative mining can be used to train better bi-encoder models, few have considered it in the context of cross-encoders, which are key ingredients in modern retrieval pipelines due to their high effectiveness. One noteworthy exception comes from Gao et al. [13], who propose to train cross-encoders by adapting the well-known NCE loss and augmenting it with a “localized” selection of hard negative examples from the first-stage retriever, which they call the Localized Contrastive Estimation (LCE) loss. In this work, we present a replication study of LCE on a different task and combine it with several other “tricks” (e.g., replacing BERTBase with ELECTRABase and replacing BM25 with TCT-ColBERTv2) to substantially improve ranking effectiveness. We attempt to more systematically explore certain parts of the hyperparameter space, including the choice of losses and the group size in the LCE loss. While our findings, for the most part, align with those from the original paper, we observe that for MS MARCO passage, orienting the retriever used for hard negative mining with the first-stage retriever used for inference is not as critical for improving effectiveness across all settings. Our code and documentation can be found in: https://github.com/castorini/replicate-lce.

Unstructured information in electronic health records provide an invaluable resource for medical research. To protect the confidentiality of patients and to conform to privacy regulations, de-identification methods automatically remove personally identifying information from these medical records. However, due to the unavailability of labeled data, most existing research is constrained to English medical text and little is known about the generalizability of de-identification methods across languages and domains. In this study, we construct a varied dataset consisting of the medical records of 1260 patients by sampling data from 9 institutes and three domains of Dutch healthcare. We test the generalizability of three de-identification methods across languages and domains. Our experiments show that an existing rule-based method specifically developed for the Dutch language fails to generalize to this new data. Furthermore, a state-of-the-art neural architecture performs strongly across languages and domains, even with limited training data. Compared to feature-based and rule-based methods the neural method requires significantly less configuration effort and domain-knowledge. We make all code and pre-trained de-identification models available to the research community, allowing practitioners to apply them to their datasets and to enable future benchmarks., Comment: Proceedings of the 1st ACM WSDM Health Search and Data Mining Workshop (HSDM2020), 2020

BACKGROUND: Among patients with acute ischemic stroke due to occlusions in the proximal anterior intracranial circulation, less than 40% regain functional independence when treated with intravenous tissue plasminogen activator (t-PA) alone. Thrombectomy with the use of a stent retriever, in addition to intravenous t-PA, increases reperfusion rates and may improve long-term functional outcome.METHODS: We randomly assigned eligible patients with stroke who were receiving or had received intravenous t-PA to continue with t-PA alone (control group) or to undergo endovascular thrombectomy with the use of a stent retriever within 6 hours after symptom onset (intervention group). Patients had confirmed occlusions in the proximal anterior intracranial circulation and an absence of large ischemic-core lesions. The primary outcome was the severity of global disability at 90 days, as assessed by means of the modified Rankin scale (with scores ranging from 0 [no symptoms] to 6 [death]).RESULTS: The study was stopped early because of efficacy. At 39 centers, 196 patients underwent randomization (98 patients in each group). In the intervention group, the median time from qualifying imaging to groin puncture was 57 minutes, and the rate of substantial reperfusion at the end of the procedure was 88%. Thrombectomy with the stent retriever plus intravenous t-PA reduced disability at 90 days over the entire range of scores on the modified Rankin scale (PCONCLUSIONS: In patients receiving intravenous t-PA for acute ischemic stroke due to occlusions in the proximal anterior intracranial circulation, thrombectomy with a stent retriever within 6 hours after onset improved functional outcomes at 90 days. (Funded by Covidien; SWIFT PRIME ClinicalTrials.gov number, NCT01657461.).

In text classification the amount and quality of training data is crucial for the performance of the classifier. The generation of training data is done by human labellers - a tedious and time-consuming work. To reduce the labelling time for single documents we propose to use condensed representations of text documents instead of the full-text document. These condensed representations are key sentences and key phrases and can be generated in a fully unsupervised way. We extended and evaluated the TextRank algorithm to automatically extract key sentences and key phrases. For representing key phrases we propose a layout similar to a tag cloud. In a user study with 37 participants we evaluated whether document labelling with these condensed representations can be done faster and equally accurate by the human labellers. Our evaluation shows that the users labelled tag clouds twice as fast and as accurately as full-text documents. While further investigations for different classification tasks are necessary, this insight could potentially reduce costs for the labelling process of text documents.

Hintergrund: Mit Dronedaron steht seit 2010 ein neues orales Antiarrhythmikum zur Behandlung von nicht-permanentem Vorhofflimmern (VHF) zur Verfügung. So zeigte die ATHENA-Studie, dass Dronedaron zu einer signifikanten Reduktion von kardiovaskulär-bedingter Hospitalisierung und Tod jeglich[for full text, please go to the a.m. URL], 10. Deutscher Kongress für Versorgungsforschung; 18. GAA-Jahrestagung

We describe a system with a multi-sensor object awareness and positioning solution for augmented tourist information systems in urban areas. The system offers technology of outdoors vision based object recognition that will extend state-of-the-art location and context aware services towards object based awareness in urban environments. In the proposed application scenario, tourist pedestrians are equipped with a GPRS or UMTS capable camera-phone. They are interested whether their field of view contains tourist sights that would point to more detailed information. Multimedia type data about related history might be explored by a mobile user who is intending to learn within the urban environment. Ambient learning is in this way achieved by pointing the device towards the urban sight, capturing an image, and consequently getting information about the object on site and within the focus of attention, i.e., the user's current field of view. The described mobile system offers multiple opportunities for application in both mobile business and commerce, and is currently developed towards an industrial prototype.

Au induced faceting of vicinal Si(1 1 1) has been studied during adsorption at elevated temperature by spot profile analyzing of low energy electron diffraction and after quenching to room temperature by scanning tunneling microscopy. On the surfaces inclined towards View the MathML source five different types of facets form with increasing Au coverage at adsorption temperatures Tads below 800 °C. They are (4 4 3), (7 7 5), (5 5 3), a stepped (2 2 1), and the (3 3 1) facets. Atomic models for the (5 5 3) and (7 7 5) facet planes are proposed on the basis of high resolution STM images. At View the MathML sourceC we found the formation of an ordered step train which covers the entire surface. With further increasing Au coverage the stepped surface decomposes again into (1 1 1) terraces and step bunches. Driving force is the formation of the Si(1 1 1)-(5×2)–Au reconstruction.

In an interconnected world, Linked Data is more important than ever before. However, it is still quite difficult to access this new wealth of semantic data directly without having in-depth knowledge about SPARQL and related semantic technologies. Also, most people are currently used to con- suming data as 2-dimensional tables. Linked Data is by defi- nition always a graph, and not that many people are used to handle data in graph structures. Therefore we present the Linked Data Query Wizard, a web-based tool for displaying, accessing, filtering, exploring, and navigating Linked Data stored in SPARQL endpoints. The main innovation of the interface is that it turns the graph structure of Linked Data into a tabular interface and provides easy-to-use interaction possibilities by using metaphors and techniques from current search engines and spreadsheet applications that regular web users are already familiar with.

In this work we present APA Labs, a generic framework for visualizing the news article domain. APA Labs is a web-based platform enabling retrieval and analysis of news repositories provided by the Austrian Press Agency. APA Labs is designed as a rich internet application combined with a modular system of interactive visualizations. News articles are analyzed using domain specific named entity extraction methods combined with language specific heuristics. The proposed methods were subject to an evaluation procedure outlined in this contribution. This article illustrates the domain, the underlying concepts and implementation details. Several visualization modules are presented and an outlook on planned modules is given. Being online for around six months the community feedback as well as the easy integration of new modules shows the success of the underlying concept and the platform itself.

Poster-Abstract Web tables comprise a rich source of factual information. However, without semantic annotation of the tables’ content the information is not usable for automatic integration and search. We propose a methodology to annotate table headers with semantic type information based on the content of column’s cells. In our experiments on 50 tables we achieved an F1 value of 0.55, where the accuracy greatly varies depending on the used ontology. Moreover, we found that for 94% of maximal F1 score only 20 cells (37%) need to be considered on average. Results suggest that for table disambiguation the choice of the ontology needs to be considered and the data input size can be reduced.

For the PAN 2011 authorship identification challenge we have developed a system based on a meta-classifier which selectively uses the results of multiple base classifiers. In addition we also performed feature engineering based on the given domain of e-mails. We present our system as well as results on the evaluation dataset. Our system performed second and third best in the authorship attribution task on the large data sets, and ranked middle for the small data set in the attribution task and in the verification task.