Sarvnaz Karimi

Bio: Sarvnaz Karimi is an academic researcher from Commonwealth Scientific and Industrial Research Organisation. The author has contributed to research in topic(s): Transliteration & Topic model. The author has an hindex of 22, co-authored 94 publication(s) receiving 1842 citation(s). Previous affiliations of Sarvnaz Karimi include University of Melbourne & RMIT University.

Evaluating topic models for digital libraries

Abstract: Topic models could have a huge impact on improving the ways users find and discover content in digital libraries and search interfaces through their ability to automatically learn and apply subject tags to each and every item in a collection, and their ability to dynamically create virtual collections on the fly. However, much remains to be done to tap this potential, and empirically evaluate the true value of a given topic model to humans. In this work, we sketch out some sub-tasks that we suggest pave the way towards this goal, and present methods for assessing the coherence and interpretability of topics learned by topic models. Our large-scale user study includes over 70 human subjects evaluating and scoring almost 500 topics learned from collections from a wide range of genres and domains. We show how scoring model -- based on pointwise mutual information of word-pair using Wikipedia, Google and MEDLINE as external data sources - performs well at predicting human scores. This automated scoring of topics is an important first step to integrating topic modeling into digital libraries

Cadec: A corpus of adverse drug event annotations

Abstract: CSIRO Adverse Drug Event Corpus (Cadec) is a new rich annotated corpus of medical forum posts on patient-reported Adverse Drug Events (ADEs). The corpus is sourced from posts on social media, and contains text that is largely written in colloquial language and often deviates from formal English grammar and punctuation rules. Annotations contain mentions of concepts such as drugs, adverse effects, symptoms, and diseases linked to their corresponding concepts in controlled vocabularies, i.e., SNOMED Clinical Terms and MedDRA. The quality of the annotations is ensured by annotation guidelines, multi-stage annotations, measuring inter-annotator agreement, and final review of the annotations by a clinical terminologist. This corpus is useful for studies in the area of information extraction, or more generally text mining, from social media to detect possible adverse drug reactions from direct patient reports. The corpus is publicly available at https://data.csiro.au.(1).

Location extraction from disaster-related microblogs

Abstract: Location information is critical to understanding the impact of a disaster, including where the damage is, where people need assistance and where help is available. We investigate the feasibility of applying Named Entity Recognizers to extract locations from microblogs, at the level of both geo-location and point-of-interest. Our experimental results show that such tools once retrained on microblog data have great potential to detect the where information, even at the granularity of point-of-interest.

Text and Data Mining Techniques in Adverse Drug Reaction Detection

Abstract: We review data mining and related computer science techniques that have been studied in the area of drug safety to identify signals of adverse drug reactions from different data sources, such as spontaneous reporting databases, electronic health records, and medical literature. Development of such techniques has become more crucial for public heath, especially with the growth of data repositories that include either reports of adverse drug reactions, which require fast processing for discovering signals of adverse reactions, or data sources that may contain such signals but require data or text mining techniques to discover them. In order to highlight the importance of contributions made by computer scientists in this area so far, we categorize and review the existing approaches, and most importantly, we identify areas where more research should be undertaken.

Best Topic Word Selection for Topic Labelling

Abstract: This paper presents the novel task of best topic word selection, that is the selection of the topic word that is the best label for a given topic, as a means of enhancing the interpretation and visualisation of topic models We propose a number of features intended to capture the best topic word, and show that, in combination as inputs to a reranking model, we are able to consistently achieve results above the baseline of simply selecting the highest-ranked topic word This is the case both when training in-domain over other labelled topics for that topic model, and cross-domain, using only labellings from independent topic models learned over document collections from different domains and genres

Machine learning

Abstract: Machine Learning is the study of methods for programming computers to learn. Computers are applied to a wide range of tasks, and for most of these it is relatively easy for programmers to design and implement the necessary software. However, there are many tasks for which this is difficult or impossible. These can be divided into four general categories. First, there are problems for which there exist no human experts. For example, in modern automated manufacturing facilities, there is a need to predict machine failures before they occur by analyzing sensor readings. Because the machines are new, there are no human experts who can be interviewed by a programmer to provide the knowledge necessary to build a computer system. A machine learning system can study recorded data and subsequent machine failures and learn prediction rules. Second, there are problems where human experts exist, but where they are unable to explain their expertise. This is the case in many perceptual tasks, such as speech recognition, hand-writing recognition, and natural language understanding. Virtually all humans exhibit expert-level abilities on these tasks, but none of them can describe the detailed steps that they follow as they perform them. Fortunately, humans can provide machines with examples of the inputs and correct outputs for these tasks, so machine learning algorithms can learn to map the inputs to the outputs. Third, there are problems where phenomena are changing rapidly. In finance, for example, people would like to predict the future behavior of the stock market, of consumer purchases, or of exchange rates. These behaviors change frequently, so that even if a programmer could construct a good predictive computer program, it would need to be rewritten frequently. A learning program can relieve the programmer of this burden by constantly modifying and tuning a set of learned prediction rules. Fourth, there are applications that need to be customized for each computer user separately. Consider, for example, a program to filter unwanted electronic mail messages. Different users will need different filters. It is unreasonable to expect each user to program his or her own rules, and it is infeasible to provide every user with a software engineer to keep the rules up-to-date. A machine learning system can learn which mail messages the user rejects and maintain the filtering rules automatically. Machine learning addresses many of the same research questions as the fields of statistics, data mining, and psychology, but with differences of emphasis. Statistics focuses on understanding the phenomena that have generated the data, often with the goal of testing different hypotheses about those phenomena. Data mining seeks to find patterns in the data that are understandable by people. Psychological studies of human learning aspire to understand the mechanisms underlying the various learning behaviors exhibited by people (concept learning, skill acquisition, strategy change, etc.).

Opportunities and obstacles for deep learning in biology and medicine.

Abstract: Deep learning describes a class of machine learning algorithms that are capable of combining raw inputs into layers of intermediate features. These algorithms have recently shown impressive results across a variety of domains. Biology and medicine are data-rich disciplines, but the data are complex and often ill-understood. Hence, deep learning techniques may be particularly well suited to solve problems of these fields. We examine applications of deep learning to a variety of biomedical problems-patient classification, fundamental biological processes and treatment of patients-and discuss whether deep learning will be able to transform these tasks or if the biomedical sphere poses unique challenges. Following from an extensive literature review, we find that deep learning has yet to revolutionize biomedicine or definitively resolve any of the most pressing challenges in the field, but promising advances have been made on the prior state of the art. Even though improvements over previous baselines have been modest in general, the recent progress indicates that deep learning methods will provide valuable means for speeding up or aiding human investigation. Though progress has been made linking a specific neural network's prediction to input features, understanding how users should interpret these models to make testable hypotheses about the system under study remains an open challenge. Furthermore, the limited amount of labelled data for training presents problems in some domains, as do legal and privacy constraints on work with sensitive health records. Nonetheless, we foresee deep learning enabling changes at both bench and bedside with the potential to transform several areas of biology and medicine.

Automatic Evaluation of Topic Coherence

Abstract: This paper introduces the novel task of topic coherence evaluation, whereby a set of words, as generated by a topic model, is rated for coherence or interpretability. We apply a range of topic scoring models to the evaluation task, drawing on WordNet, Wikipedia and the Google search engine, and existing research on lexical similarity/relatedness. In comparison with human scores for a set of learned topics over two distinct datasets, we show a simple co-occurrence measure based on pointwise mutual information over Wikipedia data is able to achieve results for the task at or nearing the level of inter-annotator correlation, and that other Wikipedia-based lexical relatedness methods also achieve strong results. Google produces strong, if less consistent, results, while our results over WordNet are patchy at best.