An effective approach to verbose queries using a limited dependencies language model

Intuitively, any ‘bag of words’ approach in IR should benefit from taking term dependencies into account. Unfortunately, for years the results of exploiting such dependencies have been mixed or inconclusive. To improve the situation, this paper shows how the natural language properties of the target documents can be used to transform and enrich the term dependencies to more useful statistics. This is done in three steps. The term co-occurrence statistics of queries and documents are each represented by a Markov chain. The paper proves that such a chain is ergodic, and therefore its asymptotic behavior is unique, stationary, and independent of the initial state. Next, the stationary distribution is taken to model queries and documents, rather than their initial distributions. Finally, ranking is achieved following the customary language modeling paradigm. The main contribution of this paper is to argue why the asymptotic behavior of the document model is a better representation then just the document’s initial distribution. A secondary contribution is to investigate the practical application of this representation in case the queries become increasingly verbose. In the experiments (based on Lemur’s search engine substrate) the default query model was replaced by the stable distribution of the query. Just modeling the query this way already resulted in significant improvements over a standard language model baseline. The results were on a par or better than more sophisticated algorithms that use fine-tuned parameters or extensive training. Moreover, the more verbose the query, the more effective the approach seems to become.

Classifying the user intent of web queries using k-means clustering

Purpose: Web search engines are frequently used by people to locate information on the Internet. However, not all queries have an informational goal. Instead of information, some people may be looking for specific web sites or may wish to conduct transactions with web services. This paper aims to focus on automatically classifying the different user intents behind web queries. Design/methodology/approach: For the research reported in this paper, 130,000 web search engine queries are categorized as informational, navigational, or transactional using a k-means clustering approach based on a variety of query traits. Findings: The research findings show that more than 75 percent of web queries (clustered into eight classifications) are informational in nature, with about 12 percent each for navigational and transactional. Results also show that web queries fall into eight clusters, six primarily informational, and one each of primarily transactional and navigational. Research limitations/implications: This study provides an important contribution to web search literature because it provides information about the goals of searchers and a method for automatically classifying the intents of the user queries. Automatic classification of user intent can lead to improved web search engines by tailoring results to specific user needs. Practical implications: The paper discusses how web search engines can use automatically classified user queries to provide more targeted and relevant results in web searching by implementing a real time classification method as presented in this research. Originality/value: This research investigates a new application of a method for automatically classifying the intent of user queries. There has been limited research to date on automatically classifying the user intent of web queries, even though the pay-off for web search engines can be quite beneficial. © Emerald Group Publishing Limited.

Pairwise similarity of TopSig document signatures

This paper analyses the pairwise distances of signatures produced by the TopSig retrieval model on two document collections. The distribution of the distances are compared to purely random signatures. It explains why TopSig is only competitive with state of the art retrieval models at early precision. Only the local neighbourhood of the signatures is interpretable. We suggest this is a common property of vector space models.

Reasoning about component similarity in building product models from the construction perspective

Identifying the design features that impact construction is essential to developing cost effective and constructible designs. The similarity of building components is a critical design feature that affects method selection, productivity, and ultimately construction cost and schedule performance. However, there is limited understanding of what constitutes similarity in the design of building components and limited computer-based support to identify this feature in a building product model. This paper contributes a feature-based framework for representing and reasoning about component similarity that builds on ontological modelling, model-based reasoning and cluster analysis techniques. It describes the ontology we developed to characterize component similarity in terms of the component attributes, the direction, and the degree of variation. It also describes the generic reasoning process we formalized to identify component similarity in a standard product model based on practitioners' varied preferences. The generic reasoning process evaluates the geometric, topological, and symbolic similarities between components, creates groupings of similar components, and quantifies the degree of similarity. We implemented this reasoning process in a prototype cost estimating application, which creates and maintains cost estimates based on a building product model. Validation studies of the prototype system provide evidence that the framework is general and enables a more accurate and efficient cost estimating process.

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.

Securities market regulators’ use of the ‘please explain’ queries sanction and its power of compliance

The aim of this study is to investigate the compliance impact of price queries issued by a securities market operator to its participating firms. Market operators in Australia and New Zealand, such as the Australian Securities Exchange and the New Zealand Securities Exchange, have the regulatory power in their rules to issue queries to its market participants to explain unusual fluctuations in trading price or volume in the market. The operator will issue a price query where it believes that the market has not been fully informed as to price relevant information. Responsive regulation has informed much of the regulatory debate in securities laws in our region. We posit that price queries are one strategy that a market operator can use in communicating its enforcement expectations to its stakeholder. However, whilst responsive regulation informs regulatory choices, an alternate view seeks to explain why participants respond to these regulatory strategies, and we use disclosure behaviour after price queries to test compliance behaviour

Optimization of an integrated model for automatic reduction and expansion of long queries

A long query provides more useful hints for searching relevant documents, but it is likely to introduce noise which affects retrieval performance. In order to smooth such adverse effect, it is important to reduce noisy terms, introduce and boost additional relevant terms. This paper presents a comprehensive framework, called Aspect Hidden Markov Model (AHMM), which integrates query reduction and expansion, for retrieval with long queries. It optimizes the probability distribution of query terms by utilizing intra-query term dependencies as well as the relationships between query terms and words observed in relevance feedback documents. Empirical evaluation on three large-scale TREC collections demonstrates that our approach, which is automatic, achieves salient improvements over various strong baselines, and also reaches a comparable performance to a state of the art method based on user’s interactive query term reduction and expansion.

Similarity solutions for flow and heat transfer of non-Newtonian fluid over a stretching surface

Similarity solutions are carried out for flow of power law non-Newtonian fluid film on unsteady stretching surface subjected to constant heat flux. Free convection heat transfer induces thermal boundary layer within a semi-infinite layer of Boussinesq fluid. The nonlinear coupled partial differential equations (PDE) governing the flow and the boundary conditions are converted to a system of ordinary differential equations (ODE) using two-parameter groups. This technique reduces the number of independent variables by two, and finally the obtained ordinary differential equations are solved numerically for the temperature and velocity using the shooting method. The thermal and velocity boundary layers are studied by the means of Prandtl number and non-Newtonian power index plotted in curves.

University of Glasgow at ImageCLEFPhoto 2009 : optimising similarity and diversity in image retrieval

In this paper we describe the approaches adopted to generate the runs submitted to ImageCLEFPhoto 2009 with an aim to promote document diversity in the rankings. Four of our runs are text based approaches that employ textual statistics extracted from the captions of images, i.e. MMR [1] as a state of the art method for result diversification, two approaches that combine relevance information and clustering techniques, and an instantiation of Quantum Probability Ranking Principle. The fifth run exploits visual features of the provided images to re-rank the initial results by means of Factor Analysis. The results reveal that our methods based on only text captions consistently improve the performance of the respective baselines, while the approach that combines visual features with textual statistics shows lower levels of improvements.

Analysing the similarity of proteins based on a new approach to empirical mode decomposition

The function of a protein can be partially determined by the information contained in its amino acid sequence. It can be assumed that proteins with similar amino acid sequences normally have closer functions. Hence analysing the similarity of proteins has become one of the most important areas of protein study. In this work, a layered comparison method is used to analyze the similarity of proteins. It is based on the empirical mode decomposition (EMD) method, and protein sequences are characterized by the intrinsic mode functions (IMFs). The similarity of proteins is studied with a new cross-correlation formula. It seems that the EMD method can be used to detect the functional relationship of two proteins. This kind of similarity method is a complement of traditional sequence similarity approaches which focus on the alignment of amino acids

Searching for semantic person queries using channel representations

It is not uncommon to hear a person of interest described by their height, build, and clothing (i.e. type and colour). These semantic descriptions are commonly used by people to describe others, as they are quick to relate and easy to understand. However such queries are not easily utilised within intelligent surveillance systems as they are difficult to transform into a representation that can be searched for automatically in large camera networks. In this paper we propose a novel approach that transforms such a semantic query into an avatar that is searchable within a video stream, and demonstrate state-of-the-art performance for locating a subject in video based on a description.

Using internet search queries for infectious disease surveillance: screening diseases for suitability

Background Internet-based surveillance systems provide a novel approach to monitoring infectious diseases. Surveillance systems built on internet data are economically, logistically and epidemiologically appealing and have shown significant promise. The potential for these systems has increased with increased internet availability and shifts in health-related information seeking behaviour. This approach to monitoring infectious diseases has, however, only been applied to single or small groups of select diseases. This study aims to systematically investigate the potential for developing surveillance and early warning systems using internet search data, for a wide range of infectious diseases. Methods Official notifications for 64 infectious diseases in Australia were downloaded and correlated with frequencies for 164 internet search terms for the period 2009–13 using Spearman’s rank correlations. Time series cross correlations were performed to assess the potential for search terms to be used in construction of early warning systems. Results Notifications for 17 infectious diseases (26.6%) were found to be significantly correlated with a selected search term. The use of internet metrics as a means of surveillance has not previously been described for 12 (70.6%) of these diseases. The majority of diseases identified were vaccine-preventable, vector-borne or sexually transmissible; cross correlations, however, indicated that vector-borne and vaccine preventable diseases are best suited for development of early warning systems. Conclusions The findings of this study suggest that internet-based surveillance systems have broader applicability to monitoring infectious diseases than has previously been recognised. Furthermore, internet-based surveillance systems have a potential role in forecasting emerging infectious disease events, especially for vaccine-preventable and vector-borne diseases

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).

Similarity-based birdcall retrieval from environmental audio

Automated digital recordings are useful for large-scale temporal and spatial environmental monitoring. An important research effort has been the automated classification of calling bird species. In this paper we examine a related task, retrieval of birdcalls from a database of audio recordings, similar to a user supplied query call. Such a retrieval task can sometimes be more useful than an automated classifier. We compare three approaches to similarity-based birdcall retrieval using spectral ridge features and two kinds of gradient features, structure tensor and the histogram of oriented gradients. The retrieval accuracy of our spectral ridge method is 94% compared to 82% for the structure tensor method and 90% for the histogram of gradients method. Additionally, this approach potentially offers a more compact representation and is more computationally efficient.