Unsupervised low-dimensional vector representations for words, phrases and text that are transparent, scalable, and produce similarity metrics that are not redundant with neural embeddings

Refining electronic medical records representation in manifold subspace

Background

Electronic medical records (EMR) contain detailed information about patient health. Developing an effective representation model is of great significance for the downstream applications of EMR. However, processing data directly is difficult because EMR data has such characteristics as incompleteness, unstructure and redundancy. Therefore, preprocess of the original data is the key step of EMR data mining. The classic distributed word representations ignore the geometric feature of the word vectors for the representation of EMR data, which often underestimate the similarities between similar words and overestimate the similarities between distant words. This results in word similarity obtained from embedding models being inconsistent with human judgment and much valuable medical information being lost.

Results

In this study, we propose a biomedical word embedding framework based on manifold subspace. Our proposed model first obtains the word vector representations of the EMR data, and then re-embeds the word vector in the manifold subspace. We develop an efficient optimization algorithm with neighborhood preserving embedding based on manifold optimization. To verify the algorithm presented in this study, we perform experiments on intrinsic evaluation and external classification tasks, and the experimental results demonstrate its advantages over other baseline methods.

Conclusions

Manifold learning subspace embedding can enhance the representation of distributed word representations in electronic medical record texts. Reduce the difficulty for researchers to process unstructured electronic medical record text data, which has certain biomedical research value.

A web-based tool for automatically linking clinical trials to their publications

OBJECTIVE

Evidence synthesis teams, physicians, policy makers, and patients and their families all have an interest in following the outcomes of clinical trials and would benefit from being able to evaluate both the results posted in trial registries and in the publications that arise from them. Manual searching for publications arising from a given trial is a laborious and uncertain process. We sought to create a statistical model to automatically identify PubMed articles likely to report clinical outcome results from each registered trial in ClinicalTrials.gov.

MATERIALS AND METHODS

A machine learning-based model was trained on pairs (publications known to be linked to specific registered trials). Multiple features were constructed based on the degree of matching between the PubMed article metadata and specific fields of the trial registry, as well as matching with the set of publications already known to be linked to that trial.

RESULTS

Evaluation of the model using known linked articles as gold standard showed that they tend to be top ranked (median best rank = 1.0), and 91% of them are ranked in the top 10.

DISCUSSION

Based on this model, we have created a free, public web-based tool that, given any registered trial in ClinicalTrials.gov, presents a ranked list of the PubMed articles in order of estimated probability that they report clinical outcome data from that trial. The tool should greatly facilitate studies of trial outcome results and their relation to the original trial designs.

Anne O'Tate: Value-added PubMed search engine for analysis and text mining

Over a decade ago, we introduced Anne O'Tate, a free, public web-based tool http://arrowsmith.psych.uic.edu/cgi-bin/arrowsmith_uic/AnneOTate.cgi to support user-driven summarization, drill-down and mining of search results from PubMed, the leading search engine for biomedical literature. A set of hotlinked buttons allows the user to sort and rank retrieved articles according to important words in titles and abstracts; topics; author names; affiliations; journal names; publication year; and clustered by topic. Any result can be further mined by choosing any other button, and small search results can be expanded to include related articles. It has been deployed continuously, serving a wide range of biomedical users and needs, and over time has also served as a platform to support the creation of new tools that address additional needs. Here we describe the current, greatly expanded implementation of Anne O'Tate, which has added additional buttons to provide new functionalities: We now allow users to sort and rank search results by important phrases contained in titles and abstracts; the number of authors listed on the article; and pairs of topics that co-occur significantly more than chance. We also display articles according to NLM-indexed publication types, as well as according to 50 different publication types and study designs as predicted by a novel machine learning-based model. Furthermore, users can import search results into two new tools: e) Mine the Gap!, which identifies pairs of topics that are under-represented within set of the search results, and f) Citation Cloud, which for any given article, allows users to visualize the set of articles that cite it; that are cited by it; that are co-cited with it; and that are bibliographically coupled to it. We invite the scientific community to explore how Anne O'Tate can assist in analyzing biomedical literature, in a variety of use cases.

A review of author name disambiguation techniques for the PubMed bibliographic database

Author names in bibliographic databases often suffer from ambiguity owing to the same author appearing under different names and multiple authors possessing similar names. It creates difficulty in associating a scholarly work with the person who wrote it, thereby introducing inaccuracy in credit attribution, bibliometric analysis, search-by-author in a digital library and expert discovery. A plethora of techniques for disambiguation of author names has been proposed in the literature. In this article, we focus on the research efforts targeted to disambiguate author names specifically in the PubMed bibliographic database. We believe this concentrated review will be useful to the research community because it discusses techniques applied to a very large real database that is actively used worldwide. We make a comprehensive survey of the existing author name disambiguation (AND) approaches that have been applied to the PubMed database: we organise the approaches into a taxonomy; describe the major characteristics of each approach including its performance, strengths, and limitations; and perform a comparative analysis of them. We also identify the datasets from PubMed that are publicly available for researchers to evaluate AND algorithms. Finally, we outline a few directions for future work.

BioWordVec, improving biomedical word embeddings with subword information and MeSH

Distributed word representations have become an essential foundation for biomedical natural language processing (BioNLP), text mining and information retrieval. Word embeddings are traditionally computed at the word level from a large corpus of unlabeled text, ignoring the information present in the internal structure of words or any information available in domain specific structured resources such as ontologies. However, such information holds potentials for greatly improving the quality of the word representation, as suggested in some recent studies in the general domain. Here we present BioWordVec: an open set of biomedical word vectors/embeddings that combines subword information from unlabeled biomedical text with a widely-used biomedical controlled vocabulary called Medical Subject Headings (MeSH). We assess both the validity and utility of our generated word embeddings over multiple NLP tasks in the biomedical domain. Our benchmarking results demonstrate that our word embeddings can result in significantly improved performance over the previous state of the art in those challenging tasks.

Unsupervised low-dimensional vector representations for words, phrases and text that are transparent, scalable, and produce similarity metrics that are not redundant with neural embeddings

Neural embeddings are a popular set of methods for representing words, phrases or text as a low dimensional vector (typically 50-500 dimensions). However, it is difficult to interpret these dimensions in a meaningful manner, and creating neural embeddings requires extensive training and tuning of multiple parameters and hyperparameters. We present here a simple unsupervised method for representing words, phrases or text as a low dimensional vector, in which the meaning and relative importance of dimensions is transparent to inspection. We have created a near-comprehensive vector representation of words, and selected bigrams, trigrams and abbreviations, using the set of titles and abstracts in PubMed as a corpus. This vector is used to create several novel implicit word-word and text-text similarity metrics. The implicit word-word similarity metrics correlate well with human judgement of word pair similarity and relatedness, and outperform or equal all other reported methods on a variety of biomedical benchmarks, including several implementations of neural embeddings trained on PubMed corpora. Our implicit word-word metrics capture different aspects of word-word relatedness than word2vec-based metrics and are only partially correlated (rho = 0.5-0.8 depending on task and corpus). The vector representations of words, bigrams, trigrams, abbreviations, and PubMed title + abstracts are all publicly available from http://arrowsmith.psych.uic.edu/arrowsmith_uic/word_similarity_metrics.html for release under CC-BY-NC license. Several public web query interfaces are also available at the same site, including one which allows the user to specify a given word and view its most closely related terms according to direct co-occurrence as well as different implicit similarity metrics.

Design of a generic, open platform for machine learning-assisted indexing and clustering of articles in PubMed, a biomedical bibliographic database

Many investigators have carried out text mining of the biomedical literature for a variety of purposes, ranging from the assignment of indexing terms to the disambiguation of author names. A common approach is to define positive and negative training examples, extract features from article metadata, and employ machine learning algorithms. At present, each research group tackles each problem from scratch, and in isolation of other projects, which causes redundancy and great waste of effort. Here, we propose and describe the design of a generic platform for biomedical text mining, which can serve as a shared resource for machine learning projects, and can serve as a public repository for their outputs. We will initially focus on a specific goal, namely, classifying articles according to Publication Type, and emphasize how feature sets can be made more powerful and robust through the use of multiple, heterogeneous similarity measures as input to machine learning models. We then discuss how the generic platform can be extended to include a wide variety of other machine learning based goals and projects, and can be used as a public platform for disseminating the results of NLP tools to end-users as well.