Semantic Similarity Measures Research Articles

A Novel Measure for Semantic Similarity Computation of Gene Ontology Terms Using Weighted Aggregation of Information Contents

Background: Gene ontology (GO) is a well-structured knowledge of biological terms that describes roles of genes and their products in a standardized and organized controlled vocabulary format. Over the last decade, many measures are developed to exploit GO advantages to determine semantic similarities between biological entities. Using GO ontologies, there are some constraints that existing GO-based semantic similarity measures try to address them. For instance, (1) edges in a GO graph, do not indicate uniform distances and also have different densities, and (2) ignoring term levels in an ontology makes “shallow annotation” drawback, i.e., two terms with a certain distance near the root of GO graph have equal semantic similarity with two terms with the same distance but far from the root. Methods: Here, we present wAIC, a two-stage hybrid semantic similarity measure using weighted aggregation of information contents. In wAIC, the impact of each common ancestor on semantic similarity value is determined according to the location of the ancestor in the ontology graph. wAIC, also, filters (from annotating term set) terms that are in upper levels of the graph ontology to reduce shallow annotation constraints. Results: Experimental results confirm that the proposed measure is more consistent with major related constraints, such that, wAIC semantic similarity values have more correlation with both sequence similarity values and gene expression based similarity values than state-of-the-art semantic similarity measures. Conclusions: WAIC show using a weighted aggregation of common ancestors is completely consistent with the human perception and can improve accuracy of gene similarity measurement.

Detecting clinically relevant new information in clinical notes across specialties and settings

BackgroundAutomated methods for identifying clinically relevant new versus redundant information in electronic health record (EHR) clinical notes is useful for clinicians and researchers involved in patient care and clinical research, respectively. We evaluated methods to automatically identify clinically relevant new information in clinical notes, and compared the quantity of redundant information across specialties and clinical settings.MethodsStatistical language models augmented with semantic similarity measures were evaluated as a means to detect and quantify clinically relevant new and redundant information over longitudinal clinical notes for a given patient. A corpus of 591 progress notes over 40 inpatient admissions was annotated for new information longitudinally by physicians to generate a reference standard. Note redundancy between various specialties was evaluated on 71,021 outpatient notes and 64,695 inpatient notes from 500 solid organ transplant patients (April 2015 through August 2015).ResultsOur best method achieved at best performance of 0.87 recall, 0.62 precision, and 0.72 F-measure. Addition of semantic similarity metrics compared to baseline improved recall but otherwise resulted in similar performance. While outpatient and inpatient notes had relatively similar levels of high redundancy (61% and 68%, respectively), redundancy differed by author specialty with mean redundancy of 75%, 66%, 57%, and 55% observed in pediatric, internal medicine, psychiatry and surgical notes, respectively.ConclusionsAutomated techniques with statistical language models for detecting redundant versus clinically relevant new information in clinical notes do not improve with the addition of semantic similarity measures. While levels of redundancy seem relatively similar in the inpatient and ambulatory settings in the Fairview Health Services, clinical note redundancy appears to vary significantly with different medical specialties.

Developing a similarity searching module for patient safety event reporting system using semantic similarity measures

BackgroundThe most important knowledge in the field of patient safety is regarding the prevention and reduction of patient safety events (PSE) during treatment and care. The similarities and patterns among the events may otherwise go unnoticed if they are not properly reported and analyzed. There is an urgent need for developing a PSE reporting system that can dynamically measure the similarities of the events and thus promote event analysis and learning effect.MethodsIn this study, three prevailing algorithms of semantic similarity were implemented to measure the similarities of the 366 PSE annotated by the taxonomy of The Agency for Healthcare Research and Quality (AHRQ). The performance of each algorithm was then evaluated by a group of domain experts based on a 4-point Likert scale. The consistency between the scales of the algorithms and experts was measured and compared with the scales randomly assigned. The similarity algorithms and scores, as a self-learning and self-updating module, were then integrated into the system.ResultsThe result shows that the similarity scores reflect a high consistency with the experts’ review than those randomly assigned. Moreover, incorporating the algorithms into our reporting system enables a mechanism to learn and update based upon PSE similarity.ConclusionIn conclusion, integrating semantic similarity algorithms into a PSE reporting system can help us learn from previous events and provide timely knowledge support to the reporters. With the knowledge base in the PSE domain, the new generation reporting system holds promise in educating healthcare providers and preventing the recurrence and serious consequences of PSE.

MathIRs: Retrieval System for Scientific Documents

Effective retrieval of mathematical contents from vast corpus of scientific documents demands enhancement in the conventional indexing and searching mechanisms. Indexing mechanism and the choice of semantic similarity measures guide the results of Math Information Retrieval system (MathIRs) to perfection. Tokenization and formula unification are among the distinguishing i features of indexing mechanism, used in MathIRs, which facilitate sub-formula and similarity search. Besides, the scientific documents and the user queries in MathIRs will contain math as well as text contents and to match these contents we require three important modules: Text-Text Similarity (TS), Math-Math Similarity (MS) and Text-Math Similarity (TMS). In this paper we have proposed MathIRs comprising these important modules and a substitution tree based mechanism for indexing mathematical expressions. We have also presented experimental results for similarity search and argued that proposal of MathIRs will ease the task of scientific document retrieval.

An efficient path computing model for measuring semantic similarity using edge and density

The shortest path between two concepts in a taxonomic ontology is commonly used to represent the semantic distance between concepts in edge-based semantic similarity measures. In the past, edge counting, which is simple and intuitive and has low computational complexity, was considered the default method for path computation. However, a large lexical taxonomy, such as WordNet, has irregular link densities between concepts due to its broad domain, but edge counting-based path computation is powerless for this non-uniformity problem. In this paper, we advocate that the path computation can be separated from edge-based similarity measures and can form various general computing models. Therefore, to solve the problem of the non-uniformity of concept density in a large taxonomic ontology, we propose a new path computing model based on the compensation of local area density of concepts, which is equal to the number of direct hyponyms of the subsumers for concepts in the shortest path. This path model considers the local area density of concepts as an extension of the edge counting-based path according to the information theory. This model is a general path computing model and can be applied in various edge-based similarity approaches. The experimental results show that the proposed path model improves the average optimal correlation between edge-based measures and human judgments on the Miller and Charles benchmark for WordNet from less than 0.79 to more than 0.86, on the Pedersenet al. benchmark (average of both Physician and Coder) for SNOMED-CT from less than 0.75 to more than 0.82, and it has a large advantage in efficiency compared with information content computation in a dynamic ontology, thereby successfully improving the edge-based similarity measure as an excellent method with high performance and high efficiency.

Investigating Correlation between Protein Sequence Similarity and Semantic Similarity Using Gene Ontology Annotations.

Sequence similarity is a commonly used measure to compare proteins. With the increasing use of ontologies, semantic (function) similarity is getting importance. The correlation between these measures has been applied in the evaluation of new semantic similarity methods, and in protein function prediction. In this research, we investigate the relationship between the two similarity methods. The results suggest absence of a strong correlation between sequence and semantic similarities. There is a large number of proteins with low sequence similarity and high semantic similarity. We observe that Pearson's correlation coefficient is not sufficient to explain the nature of this relationship. Interestingly, the term semantic similarity values above 0 and below 1 do not seem to play a role in improving the correlation. That is, the correlation coefficient depends only on the number of common GO terms in proteins under comparison, and the semantic similarity measurement method does not influence it. Semantic similarity and sequence similarity have a distinct behavior. These findings are of significant effect for future works on protein comparison, and will help understand the semantic similarity between proteins in a better way.

Novel approach to Case Based Reasoning System by aggregating Semantic Similarity Measures using Fuzzy Aggregation for Case Retrieval

Novel approach to Case Based Reasoning System by aggregating Semantic Similarity Measures using Fuzzy Aggregation for Case Retrieval

Assessment of Semantic Similarity between Proteins Using Information Content and Topological Properties of the Gene Ontology Graph.

The semantic similarity between two interacting proteins can be estimated by combining the similarity scores of the GO terms associated with the proteins. Greater number of similar GO annotations between two proteins indicates greater interaction affinity. Existing semantic similarity measures make use of the GO graph structure, the information content of GO terms, or a combination of both. In this paper, we present a hybrid approach which utilizes both the topological features of the GO graph and information contents of the GO terms. More specifically, we 1) consider a fuzzy clustering of the GO graph based on the level of association of the GO terms, 2) estimate the GO term memberships to each cluster center based on the respective shortest path lengths, and 3) assign weightage to GO term pairs on the basis of their dissimilarity with respect to the cluster centers. We test the performance of our semantic similarity measure against seven other previously published similarity measures using benchmark protein-protein interaction datasets of Homo sapiens and Saccharomyces cerevisiae based on sequence similarity, Pfam similarity, area under ROC curve, and measure.

Exploring Approaches for Detecting Protein Functional Similarity within an Orthology-based Framework

Protein functional similarity based on gene ontology (GO) annotations serves as a powerful tool when comparing proteins on a functional level in applications such as protein-protein interaction prediction, gene prioritization, and disease gene discovery. Functional similarity (FS) is usually quantified by combining the GO hierarchy with an annotation corpus that links genes and gene products to GO terms. One large group of algorithms involves calculation of GO term semantic similarity (SS) between all the terms annotating the two proteins, followed by a second step, described as “mixing strategy”, which involves combining the SS values to yield the final FS value. Due to the variability of protein annotation caused e.g. by annotation bias, this value cannot be reliably compared on an absolute scale. We therefore introduce a similarity z-score that takes into account the FS background distribution of each protein. For a selection of popular SS measures and mixing strategies we demonstrate moderate accuracy improvement when using z-scores in a benchmark that aims to separate orthologous cases from random gene pairs and discuss in this context the impact of annotation corpus choice. The approach has been implemented in Frela, a fast high-throughput public web server for protein FS calculation and interpretation.

MultiWiki

In this article, we address the problem of text passage alignment across interlingual article pairs in Wikipedia. We develop methods that enable the identification and interlinking of text passages written in different languages and containing overlapping information. Interlingual text passage alignment can enable Wikipedia editors and readers to better understand language-specific context of entities, provide valuable insights in cultural differences, and build a basis for qualitative analysis of the articles. An important challenge in this context is the tradeoff between the granularity of the extracted text passages and the precision of the alignment. Whereas short text passages can result in more precise alignment, longer text passages can facilitate a better overview of the differences in an article pair. To better understand these aspects from the user perspective, we conduct a user study at the example of the German, Russian, and English Wikipedia and collect a user-annotated benchmark. Then we propose MultiWiki, a method that adopts an integrated approach to the text passage alignment using semantic similarity measures and greedy algorithms and achieves precise results with respect to the user-defined alignment. The MultiWiki demonstration is publicly available and currently supports four language pairs.

HESML: A scalable ontology-based semantic similarity measures library with a set of reproducible experiments and a replication dataset

This work is a detailed companion reproducibility paper of the methods and experiments proposed by Lastra-Díaz and García-Serrano in (2015, 2016) [56–58], which introduces the following contributions: (1) a new and efficient representation model for taxonomies, called PosetHERep, which is an adaptation of the half-edge data structure commonly used to represent discrete manifolds and planar graphs; (2) a new Java software library called the Half-Edge Semantic Measures Library (HESML) based on PosetHERep, which implements most ontology-based semantic similarity measures and Information Content (IC) models reported in the literature; (3) a set of reproducible experiments on word similarity based on HESMLand ReproZip with the aim of exactly reproducing the experimental surveys in the three aforementioned works; (4) a replication framework and dataset, called WNSimRep v1, whose aim is to assist the exact replication of most methods reported in the literature; and finally, (5) a set of scalability and performance benchmarks for semantic measures libraries. PosetHERep and HESML are motivated by several drawbacks in the current semantic measures libraries, especially the performance and scalability, as well as the evaluation of new methods and the replication of most previous methods. The reproducible experiments introduced herein are encouraged by the lack of a set of large, self-contained and easily reproducible experiments with the aim of replicating and confirming previously reported results. Likewise, the WNSimRep v1 dataset is motivated by the discovery of several contradictory results and difficulties in reproducing previously reported methods and experiments. PosetHERepproposes a memory-efficient representation for taxonomies which linearly scales with the size of the taxonomy and provides an efficient implementation of most taxonomy-based algorithms used by the semantic measures and IC models, whilst HESML provides an open framework to aid research into the area by providing a simpler and more efficient software architecture than the current software libraries. Finally, we prove the outperformance of HESML on the state-of-the-art libraries, as well as the possibility of significantly improving their performance and scalability without caching using PosetHERep.

Evaluating the effect of annotation size on measures of semantic similarity

BackgroundOntologies are widely used as metadata in biological and biomedical datasets. Measures of semantic similarity utilize ontologies to determine how similar two entities annotated with classes from ontologies are, and semantic similarity is increasingly applied in applications ranging from diagnosis of disease to investigation in gene networks and functions of gene products.ResultsHere, we analyze a large number of semantic similarity measures and the sensitivity of similarity values to the number of annotations of entities, difference in annotation size and to the depth or specificity of annotation classes. We find that most similarity measures are sensitive to the number of annotations of entities, difference in annotation size as well as to the depth of annotation classes; well-studied and richly annotated entities will usually show higher similarity than entities with only few annotations even in the absence of any biological relation.ConclusionsOur findings may have significant impact on the interpretation of results that rely on measures of semantic similarity, and we demonstrate how the sensitivity to annotation size can lead to a bias when using semantic similarity to predict protein-protein interactions.

Meaning preservation in Example-based Machine Translation with structural semantics

The main tasks in Example-based Machine Translation (EBMT) comprise of source text decomposition, following with translation examples matching and selection, and finally adaptation and recombination of the target translation. As the natural language is ambiguous in nature, the preservation of source text’s meaning throughout these processes is complex and challenging. A structural semantics is introduced, as an attempt towards meaning-based approach to improve the EBMT system. The structural semantics is used to support deeper semantic similarity measurement and impose structural constraints in translation examples selection. A semantic compositional structure is derived from the structural semantics of the selected translation examples. This semantic compositional structure serves as a representation structure to preserve the consistency and integrity of the input sentence’s meaning structure throughout the recombination process. In this paper, an English to Malay EBMT system is presented to demonstrate the practical application of this structural semantics. Evaluation of the translation test results shows that the new translation framework based on the structural semantics has outperformed the previous EBMT framework.

Featured correspondence topic model for semantic search on social image collections

Nowadays, due to the rapid growth of digital technologies, huge volumes of image data are created and shared on social media sites. User-provided tags attached to each social image are widely recognized as a bridge to fill the semantic gap between low-level image features and high-level concepts. Hence, a combination of images along with their corresponding tags is useful for intelligent retrieval systems, those are designed to gain high-level understanding from images and facilitate semantic search. However, user-provided tags in practice are usually incomplete and noisy, which may degrade the retrieval performance. To tackle this problem, we present a novel retrieval framework that automatically associates the visual content with textual tags and enables effective image search. To this end, we first propose a probabilistic topic model learned on social images to discover latent topics from the co-occurrence of tags and image features. Moreover, our topic model is built by exploiting the expert knowledge about the correlation between tags with visual contents and the relationship among image features that is formulated in terms of spatial location and color distribution. The discovered topics then help to predict missing tags of an unseen image as well as the ones partially labeled in the database. These predicted tags can greatly facilitate the reliable measure of semantic similarity between the query and database images. Therefore, we further present a scoring scheme to estimate the similarity by fusing textual tags and visual representation. Extensive experiments conducted on three benchmark datasets show that our topic model provides the accurate annotation against the noise and incompleteness of tags. Using our generalized scoring scheme, which is particularly advantageous to many types of queries, the proposed approach also outperforms state-of-the-art approaches in terms of retrieval accuracy.

What's the story? A computational analysis of narrative competence in autism.

Individuals with autism spectrum disorder demonstrate narrative (i.e. storytelling) difficulties which can significantly impact their ability to form and maintain social relationships. However, existing research has not comprehensively documented these impairments in more open-ended, emotionally evocative situations common to daily interactions. Computational linguistic measures offer a promising complement to traditional hand-coding methods of narrative analysis and in this study were applied together with hand coding of narratives elicited with emotionally salient scenes from the Thematic Apperception Test. In total, 19 individuals with autism spectrum disorder and 14 typically developing controls were asked to tell stories about six images from the Thematic Apperception Test. Both structural and qualitative aspects of narrative were assessed using a hand-coding system and Latent Semantic Analysis, an automated computational measure of semantic similarity. Individuals with autism spectrum disorder demonstrated significant difficulties with the use of complex syntax to integrate their narratives and problems explaining characters' intentions. These and other key narrative skills were strongly related to narrative competence scores derived from Latent Semantic Analysis, which also distinguished the autism spectrum disorder group from controls. Together, results underscore key narrative impairments in autism spectrum disorder and support the promise of Latent Semantic Analysis as a valuable tool for the quantitative assessment of complex language abilities.

MeSH-based disambiguation method using an intrinsic information content measure of semantic similarity

Word Sense Disambiguation represents a crucial task in many natural language processing applications such as information extraction, information retrieval and text summarization. It intends to identify the correct sense of an ambiguous term (target word) according to its context. In this paper, we present a biomedical knowledge-based disambiguation method for determining the adequate domain or sub-domain (sense) of an ambiguous biomedical term. This method uses the MeSH thesaurus as a knowledge resource and an intrinsic information content-based semantic similarity to measure the likeness between their different senses of the ambiguous term MeSH and its context. We define this latter as the set of unambiguous MeSH terms characterizing a document. We aim on one hand to minimize the time and computational complexity and on the other hand to increase the accuracy of the disambiguation by using such context. To evaluate this method, we involved it into a document indexing and retrieval system. The results of experiments, carried out on a sub-set of the OHSUMED collection, in terms of Mean Average Precision show that our method performs well.

Model and technology of lingware support in tasks of the nuclear knowledge management system

Model and technology of lingware application in the tasks of the knowledge management are presented. The used model of thesauri paradigmatic relationships allows measuring semantic similarity of documents. An algorithm for compiling a semantic intersection of conceptual patterns based on the coinciding maximum principle is described. The semantic neighborhood construction in accordance with the value of semantic similarity measure and descriptors ranking allows using in the expansion of the search query the descriptors, whose semantics are closer due to the links, which reflect the common semantics of genus-species and associative relationships. An algorithm for thesauri used in automatic classification tasks is considered. The technology of conceptual hierarchical dynamic formation of structures is proposed. The proposed method of constructing a terminological network structure with hierarchical relationships for the collection of texts of a particular subject area is based on the sequential formation of lexicographical neighborhoods of “nuclear” terms, which are used as high-frequency nouns. To build up a conceptual structure a documental databases are used.

A Hill-Climbing based Information Retrieval Mechanism for Formal Concept Analysis in Folksonomies

Folksonomies are a means by which communities of people annotate resources with tags. The information contributed by this action can be used to effectively search for an object. A method is required to represent and categorize objects based on their tags and search them effectively by providing words that are relevant to any object. Few existing systems use lattice structure that conceptually groups tagged objects into nodes. The lattice structure facilitates fast search and creates relations between groups of objects. In order to search objects existing statistical and mathematical measures solely take into account a few factors such as the co-occurrence of two tags rather than their meanings to determine their similarity. Hence, when searching for an object based on a user query in a lattice structure there is a need for semantic similarity measure for comparing the meaning of each query with the tags at each lattice node. Also, existing search algorithms like breadth first search process more number of concepts than necessary and so there is a need for a mechanism that performs a narrower search through the lattice. The proposed system represents and categorizes objects with a lattice structure and specifies an optimized search technique based on semantic similarity. Semantic similarity is found by extracting synonyms of the user's query and mapping them to the tags of the object in the lattice structure. In order to achieve an optimized top-down search a novel hill-climbing approach is performed over the lattice that processes minimal number of lattice nodes. Experimental results show the efficiency of the search algorithm along with the similarity measures with near accurate results to the user's query.

Comparison of Ontology-Based Semantic-Similarity Measures in the Biomedical Text

In recent years, there are many types of semantic similarity measures, which are used to measure the similarity between two concepts. It is necessary to define the differences between the measures, performance, and evaluations. The major contribution of this paper is to choose the best measure among different similarity measures that give us good result with less error rate. The experiment was done on a taxonomy built to measure the semantic distance between two concepts in the health domain, which are represented as nodes in the taxonomy. Similarity measures methods were evaluated relative to human experts’ ratings. Our experiment was applied on the ICD10 taxonomy to determine the similarity value between two concepts. The similarity between 30 pairs of the health domains has been evaluated using different types of semantic similarity measures equations. The experimental results discussed in this paper have shown that the Hoa A. Nguyen and Hisham Al-Mubaid measure has achieved high matching score by the expert’s judgment.

Object Identification for Visually Impaired

Objectives: This paper is based on a device that helps blind people to identify a set of objects used in everyday routine thereby enabling them to work independently. Method/Analysis: A new portable camera based method to recognize indoor objects for helping blind people is introduced. A more improvised algorithm called as the Coarse Description Technique has been employed. The coarse description technique can be implemented in two ways: Euclidian distance measurement method and another method that relies on semantic similarity measure modeled by means of Gaussian process estimation. Findings: The techniques used currently perform the recognition task by limiting it to a single predefined class of objects. The proposed concept in this paper utilizes a completely different alternative scheme termed as coarse description. Its main objective is to expand the recognition task to multiple objects and keep the processing period under control at the same time. While the Euclidian distance measurement method evaluates every image based on a matrix created using details regarding the pixels of each image, the Gaussian process estimation method compares images using a number of image attributes. Novelty/Improvement: PIC controller used in the existing system is replaced by a Raspberry Pi board which provides a computation speed twice as fast as the former.