Medical semantic similarity with a neural language model

Advances in neural network language models have demonstrated that these models can effectively learn representations of words meaning. In this paper, we explore a variation of neural language models that can learn on concepts taken from structured ontologies and extracted from free-text, rather than directly from terms in free-text. This model is employed for the task of measuring semantic similarity between medical concepts, a task that is central to a number of techniques in medical informatics and information retrieval. The model is built with two medical corpora (journal abstracts and patient records) and empirically validated on two ground-truth datasets of human-judged concept pairs assessed by medical professionals. Empirically, our approach correlates closely with expert human assessors ($\approx$ 0.9) and outperforms a number of state-of-the-art benchmarks for medical semantic similarity. The demonstrated superiority of this model for providing an effective semantic similarity measure is promising in that this may translate into effectiveness gains for techniques in medical information retrieval and medical informatics (e.g., query expansion and literature-based discovery).

Explorations in the distributional and semantic similarity of words

A straightforward computation of the list of the words (the `tail words' of the list) that are distributionally most similar to a given word (the `head word' of the list) leads to the question: How semantically similar to the head word are the tail words; that is: how similar are their meanings to its meaning? And can we do better? The experiment was done on nearly 18,000 most frequent nouns in a Finnish newsgroup corpus. These nouns are considered to be distributionally similar to the extent that they occur in the same direct dependency relations with the same nouns, adjectives and verbs. The extent of the similarity of their computational representations is quantified with the information radius. The semantic classification of head-tail pairs is intuitive; some tail words seem to be semantically similar to the head word, some do not. Each such pair is also associated with a number of further distributional variables. Individually, their overlap for the semantic classes is large, but the trained classification-tree models have some success in using combinations to predict the semantic class. The training data consists of a random sample of 400 head-tail pairs with the tail word ranked among the 20 distributionally most similar to the head word, excluding names. The models are then tested on a random sample of another 100 such pairs. The best success rates range from 70% to 92% of the test pairs, where a success means that the model predicted my intuitive semantic class of the pair. This seems somewhat promising when distributional similarity is used to capture semantically similar words. This analysis also includes a general discussion of several different similarity formulas, arranged in three groups: those that apply to sets with graded membership, those that apply to the members of a vector space, and those that apply to probability mass functions.

Extracting Query Interfaces Based on Form Structures and Semantic Similarity

Web databases are now pervasive. Such a database can be accessed via its query interface (usually HTML query form) only. Extracting Web query interfaces is a critical step in data integration across multiple Web databases, which creates a formal representation of a query form by extracting a set of query conditions in it. This paper presents a novel approach to extracting Web query interfaces. In this approach, a generic set of query condition rules are created to define query conditions that are semantically equivalent to SQL search conditions. Query condition rules represent the semantic roles that labels and form elements play in query conditions, and how they are hierarchically grouped into constructs of query conditions. To group labels and form elements in a query form, we explore both their structural proximity in the hierarchy of structures in the query form, which is captured by a tree of nested tags in the HTML codes of the form, and their semantic similarity, which is captured by various short texts used in labels, form elements and their properties. We have implemented the proposed approach and our experimental results show that the approach is highly effective.

Comparison of Statistical and Semantic Similarity Techniques for Paraphrase Identification

This paper compares statistical technique of paraphrase identification to semantic technique of paraphrase identification. The statistical techniques used for comparison are word set and word-order based methods where as the semantic technique used is the WordNet similarity matrix method described by Stevenson and Fernando in [3].

SSONDE: Semantic Similarity On liNked Data Entities

Linked data offers a promising setting to encode, publish and share metadata of resources. As the matter of fact, it is already adopted by data producers such as European Environment Agency, US and some EU Governs, whose first ambition is to share (meta)data making their processes more effective and transparent. Such as an increasing interest and involvement of data providers surely represents a genuine witness of the web of data success, but in a longer perspective, frameworks supporting linked data consumers in their decision making processes will be a compelling need. In this respect, the talk is introducing SSONDE, a framework enabling in detailed comparison, ranking and selection of linked data resources through the analysis of their RDF ontology driven metadata. SSONDE implements an instance similarity especially designed to support in resource selection, namely the process stakeholders engage to choose a set of resources suitable for a given analysis purpose: (i) it deploys an asymmetric similarity assessment to emphasize information about gains and losses the stakeholders get adopting a resource in place of another; (ii) it relies on an explicit formalization of contexts to tailor the similarity assessment with respect to specific user-defined selection goals. The talk aims at providing an insight on SSONDE instance similarity and it will briefly describe some examples of SSONDE deployment in the context of linked data consumption.

Exploring Semantic Textual Similarity

[EN]Measuring semantic similarity and relatedness between textual items (words, sentences, paragraphs or even documents) is a very important research area in Natural Language Processing (NLP). In fact, it has many practical applications in other NLP tasks. For instance, Word Sense Disambiguation, Textual Entailment, Paraphrase detection, Machine Translation, Summarization and other related tasks such as Information Retrieval or Question Answering. In this masther thesis we study di erent approaches to compute the semantic similarity between textual items. In the framework of the european PATHS project1, we also evaluate a knowledge-base method on a dataset of cultural item descriptions. Additionaly, we describe the work carried out for the Semantic Textual Similarity (STS) shared task of SemEval-2012. This work has involved supporting the creation of datasets for similarity tasks, as well as the organization of the task itself.

Similarity metrics within a point of view

In vector space based approaches to natural language processing, similarity is commonly measured by taking the angle between two vectors representing words or documents in a semantic space. This is natural from a mathematical point of view, as the angle between unit vectors is, up to constant scaling, the only unitarily invariant metric on the unit sphere. However, similarity judgement tasks reveal that human subjects fail to produce data which satisfies the symmetry and triangle inequality requirements for a metric space. A possible conclusion, reached in particular by Tversky et al., is that some of the most basic assumptions of geometric models are unwarranted in the case of psychological similarity, a result which would impose strong limits on the validity and applicability vector space based (and hence also quantum inspired) approaches to the modelling of cognitive processes. This paper proposes a resolution to this fundamental criticism of of the applicability of vector space models of cognition. We argue that pairs of words imply a context which in turn induces a point of view, allowing a subject to estimate semantic similarity. Context is here introduced as a point of view vector (POVV) and the expected similarity is derived as a measure over the POVV's. Different pairs of words will invoke different contexts and different POVV's. Hence the triangle inequality ceases to be a valid constraint on the angles. We test the proposal on a few triples of words and outline further research.

Semantic judgement of medical concepts : combining syntagmatic and paradigmatic information with the tensor encoding model

This paper outlines a novel approach for modelling semantic relationships within medical documents. Medical terminologies contain a rich source of semantic information critical to a number of techniques in medical informatics, including medical information retrieval. Recent research suggests that corpus-driven approaches are effective at automatically capturing semantic similarities between medical concepts, thus making them an attractive option for accessing semantic information. Most previous corpus-driven methods only considered syntagmatic associations. In this paper, we adapt a recent approach that explicitly models both syntagmatic and paradigmatic associations. We show that the implicit similarity between certain medical concepts can only be modelled using paradigmatic associations. In addition, the inclusion of both types of associations overcomes the sensitivity to the training corpus experienced by previous approaches, making our method both more effective and more robust. This finding may have implications for researchers in the area of medical information retrieval.

An evaluation of corpus-driven measures of medical concept similarity for information retrieval

Measures of semantic similarity between medical concepts are central to a number of techniques in medical informatics, including query expansion in medical information retrieval. Previous work has mainly considered thesaurus-based path measures of semantic similarity and has not compared different corpus-driven approaches in depth. We evaluate the effectiveness of eight common corpus-driven measures in capturing semantic relatedness and compare these against human judged concept pairs assessed by medical professionals. Our results show that certain corpus-driven measures correlate strongly (approx 0.8) with human judgements. An important finding is that performance was significantly affected by the choice of corpus used in priming the measure, i.e., used as evidence from which corpus-driven similarities are drawn. This paper provides guidelines for the implementation of semantic similarity measures for medical informatics and concludes with implications for medical information retrieval.

Erroneous and veridical recall are not two sides of the same coin: evidence from semantic distraction in free recall

Two experiments examined the extent to which erroneous recall blocks veridical recall using, as a vehicle for study, the disruptive impact of distractors that are semantically similar to a list of words presented for free recall. Instructing participants to avoid erroneous recall of to-be-ignored spoken distractors attenuated their recall but this did not influence the disruptive effect of those distractors on veridical recall (Experiment 1). Using an externalised output-editing procedure—whereby participants recalled all items that came to mind and identified those that were erroneous—the usual between-sequence semantic similarity effect on erroneous and veridical recall was replicated but the relationship between the rate of erroneous and veridical recall was weak (Experiment 2). The results suggest that forgetting is not due to veridical recall being blocked by similar events.

Spatially structured cognitve models of semantic information:the implications for computerised databases

Existing theories of semantic cognition propose models of cognitive processing occurring in a conceptual space, where ‘meaning’ is derived from the spatial relationships between concepts’ mapped locations within the space. Information visualisation is a growing area of research within the field of information retrieval, and methods for presenting database contents visually in the form of spatial data management systems (SDMSs) are being developed. This thesis combined these two areas of research to investigate the benefits associated with employing spatial-semantic mapping (documents represented as objects in two- and three-dimensional virtual environments are proximally mapped dependent on the semantic similarity of their content) as a tool for improving retrieval performance and navigational efficiency when browsing for information within such systems. Positive effects associated with the quality of document mapping were observed; improved retrieval performance and browsing behaviour were witnessed when mapping was optimal. It was also shown using a third dimension for virtual environment (VE) presentation provides sufficient additional information regarding the semantic structure of the environment that performance is increased in comparison to using two-dimensions for mapping. A model that describes the relationship between retrieval performance and browsing behaviour was proposed on the basis of findings. Individual differences were not found to have any observable influence on retrieval performance or browsing behaviour when mapping quality was good. The findings from this work have implications for both cognitive modelling of semantic information, and for designing and testing information visualisation systems. These implications are discussed in the conclusions of this work.

Automatic Identification of False Friends in Parallel Corpora: Statistical and Semantic Approach

False friends are pairs of words in two languages that are perceived as similar but have different meanings. We present an improved algorithm for acquiring false friends from sentence-level aligned parallel corpus based on statistical observations of words occurrences and co-occurrences in the parallel sentences. The results are compared with an entirely semantic measure for cross-lingual similarity between words based on using the Web as a corpus through analyzing the words’ local contexts extracted from the text snippets returned by searching in Google. The statistical and semantic measures are further combined into an improved algorithm for identification of false friends that achieves almost twice better results than previously known algorithms. The evaluation is performed for identifying cognates between Bulgarian and Russian but the proposed methods could be adopted for other language pairs for which parallel corpora and bilingual glossaries are available.

Automatic Discovery of Word Semantic Relations

In this paper, we propose an unsupervised methodology to automatically discover pairs of semantically related words by highlighting their local environment and evaluating their semantic similarity in local and global semantic spaces. This proposal di®ers from previous research as it tries to take the best of two different methodologies i.e. semantic space models and information extraction models. It can be applied to extract close semantic relations, it limits the search space and it is unsupervised.

Semantic features for context organization

In recent years the technological world has grown by incorporating billions of small sensing devices, collecting and sharing real-world information. As the number of such devices grows, it becomes increasingly difficult to manage all these new information sources. There is no uniform way to share, process and understand context information. In previous publications we discussed efficient ways to organize context information that is independent of structure and representation. However, our previous solution suffers from semantic sensitivity. In this paper we review semantic methods that can be used to minimize this issue, and propose an unsupervised semantic similarity solution that combines distributional profiles with public web services. Our solution was evaluated against Miller-Charles dataset, achieving a correlation of 0.6.