Web Search Queries Research Articles

Web search query data and prescription volumes of antidepressants

IntroductionPersons using the Internet generate large amounts of health-related data, which are increasingly used in modern health sciences.Objectives/aimsWe analysed the relation between annual prescription volumes (APV) of several antidepressants with marketing approval in Germany and corresponding web search query data generated in Google to test, if web search query volume may be a proxy for medical prescription practice.MethodsWe obtained APVs of several antidepressants related to corresponding prescriptions at the expense of the statutory health insurance in Germany from 2004–2013. Web search query data generated in Germany and related to defined search-terms (active substance or brand name) were obtained with Google Trends. We calculated correlations (Pearson's r) between the APVs of each substance and the respective annual “search share” values; coefficients of determination (R2) were computed to determine the amount of variability shared by the two variables.ResultsSignificant and strong correlations between substance-specific APVs and corresponding annual query volume were found for each substance during the observational interval: agomelatine (r = 0.968; R2 = 0.932; P = 0.01), bupropion (r = 0.962; R2 = 0.925; P = 0.01), citalopram (r = 0.970; R2 = 0.941; P = 0.01), escitalopram (r = 0.824; R2 = 0.682; P = 0.01), fluoxetine (r = 0.885; R2 = 0.783; P = 0.01), paroxetine (r = 0.801; R2 = 0.641; P = 0.01), and sertraline (r = 0.880; R2 = 0.689; P = 0.01).ConclusionsAlthough the used data did not allow to perform an analysis with a higher temporal resolution our results suggest that web search query volume may be a proxy for corresponding prescription behaviour. However, further studies analysing other pharmacologic agents and prescription data that facilitates an increased temporal resolution are needed to confirm this hypothesis.Disclosure of interestThe authors have not supplied their declaration of competing interest.

Red Roses and Gift Chocolates Are Judged More Positively in the U.S. Near Valentine's Day: Evidence of Naturally Occurring Cultural Priming.

Attitudes are not static, but constructed at the moment of the evaluation, incorporating temporary contextual influences. How do meaningful events that naturally occur within a culture, such as a national holiday, shape evaluative judgments of objects related to the holiday? We focused on evaluations of red roses and gift chocolates, which are everyday objects, but also iconic of Valentine’s Day in the U.S. We reasoned that if cultural events shape evaluations, then roses and chocolates would be evaluated differently near Valentine’s Day. Using a large and diverse U.S. sample, we found that as Valentine’s Day neared, evaluations of roses and chocolates (but not a comparison object) were evaluated more positively. Increases in positivity of roses and chocolates covaried with their increased cultural relevance, as quantified by the volume of web search queries involving these terms. These findings provide a demonstration of naturally occurring cultural priming by which the salience of cultural events shape evaluations.

Now-Casting Building Permits with Google Trends

We propose a useful way to predict building permits in the US, exploiting rich real-time data from web search queries. The time series on building permits is usually considered as a leading indicator of economic activity in the construction sector. Nevertheless, new data on building permits are released with a lag close to two months. Therefore, an accurate now-cast of this leading indicator is desirable. We show that models including Google search queries now-cast and forecast better than our good, not naive, univariate benchmarks both in-sample and out-of-sample. We also show that our results are robust to different specifications, the use of rolling or recursive windows and, in some cases, to the forecasting horizon. Since Google queries information is free, our approach is a simple and inexpensive way to predict building permits in the United States.

Analysis of the prediction capability of web search data based on the HE-TDC method ‒ prediction of the volume of daily tourism visitors

Web search query data are obtained to reflect social spots and serve as novel economic indicators. When faced with high-dimensional query data, selecting keywords that have plausible predictive ability and can reduce dimensionality is critical. This paper presents a new integrative method that combines Hurst Exponent (HE) and Time Difference Correlation (TDC) analysis to select keywords with powerful predictive ability. The method is called the HE-TDC screening method and requires keywords with predictive ability to satisfy two characteristics, namely, high correlation and fluctuation memorability similar to the predicting target series. An empirical study is employed to predict the volume of tourism visitors in the Jiuzhai Valley scenic area. The study shows that keywords selected using HE-TDC method produce a model with better robustness and predictive ability.

Permutation of Web Search Query Types for User Intent Privacy

Privacy remains a major concern when using search engines to find for information on the web due to the fact that search engines own massive resources in preserving search logs of each user and organizations. However, many of the present query search privacy practices require the very same search engine and third party to collaborate, making privacy even more difficult. Therefore, as a contribution, we present a heuristic, permutation of web search query types, a non-cryptographic heuristic that works by formation of obfuscated search queries via permutation of query keyword categories. Preliminary results from this study show that web search query and specific user intent privacy might be achievable from the user side without involvement of the search engine or other third parties by the permutation of web search query types.

Web Search Query Privacy, an End-User Perspective

While search engines have become vital tools for searching information on the Internet, privacy issues remain a growing concern due to the technological abilities of search engines to retain user search logs. Although such capabilities might provide enhanced personalized search results, the confidentiality of user intent remains uncertain. Even with web search query obfuscation techniques, another challenge remains, namely, reusing the same obfuscation methods is problematic, given that search engines have enormous computation and storage resources for query disambiguation. A number of web search query privacy procedures involve the cooperation of the search engine, a non-trusted entity in such cases, making query obfuscation even more challenging. In this study, we provide a review on how search engines work in regards to web search queries and user intent. Secondly, this study reviews material in a manner accessible to those outside computer science with the intent to introduce knowledge of web search engines to enable non-computer scientists to approach web search query privacy innovatively. As a contribution, we identify and highlight areas open for further investigative and innovative research in regards to end-user personalized web search privacy—that is methods that can be executed on the user side without third party involvement such as, search engines. The goal is to motivate future web search obfuscation heuristics that give users control over their personal search privacy.

Syntactic complexity of Web search queries through the lenses of language models, networks and users

Across the world, millions of users interact with search engines every day to satisfy their information needs. As the Web grows bigger over time, such information needs, manifested through user search queries, also become more complex. However, there has been no systematic study that quantifies the structural complexity of Web search queries. In this research, we make an attempt towards understanding and characterizing the syntactic complexity of search queries using a multi-pronged approach. We use traditional statistical language modeling techniques to quantify and compare the perplexity of queries with natural language (NL). We then use complex network analysis for a comparative analysis of the topological properties of queries issued by real Web users and those generated by statistical models. Finally, we conduct experiments to study whether search engine users are able to identify real queries, when presented along with model-generated ones. The three complementary studies show that the syntactic structure of Web queries is more complex than what n-grams can capture, but simpler than NL. Queries, thus, seem to represent an intermediate stage between syntactic and non-syntactic communication.

A Joint Model for Entity Set Expansion and Attribute Extraction from Web Search Queries

Entity Set Expansion (ESE) and Attribute Extraction (AE) are usually treated as two separate tasks in Information Extraction (IE). However, the two tasks are tightly coupled, and each task can benefit significantly from the other by leveraging the inherent relationship between entities and attributes. That is, 1) an attribute is important if it is shared by many typical entities of a class; 2) an entity is typical if it owns many important attributes of a class. Based on this observation, we propose a joint model for ESE and AE, which models the inherent relationship between entities and attributes as a graph. Then a graph reinforcement algorithm is proposed to jointly mine entities and attributes of a specific class. Experimental results demonstrate the superiority of our method for discovering both new entities and new attributes.

Ensembling Classifiers for Detecting User Intentions behind Web Queries

Discovering user intentions behind Web search queries is key to improving user experience. Usually, this task is seen as a classification problem, in which a sample of annotated user query intentions are provided to a supervised machine learning algorithm or classifier that learns from these examples and then can classify unseen user queries. This article proposes a new approach based on an ensemble of classifiers. The method combines syntactic and semantic features so as to effectively detect user intentions. Different setting experiments show the promise of this linguistically motivated ensembling approach, by reducing the ranking variance of single classifiers across user intentions.

Why do people google movement disorders? An infodemiological study of information seeking behaviors.

Millions of people worldwide everyday search Google or Wikipedia to look for health-related information. Aim of this study was to evaluate and interpret web search queries for terms related to movement disorders (MD) in English-speaking countries and their changes over time. We analyzed information regarding the volume of online searches in Google and Wikipedia for the most common MD and their treatments. We determined the highest search volume peaks to identify possible relation with online news headlines. The volume of searches for some queries related to MD entered in Google enormously increased over time. Most queries were related to definition, subtypes, symptoms and treatment (mostly to adverse effects, or alternatively, to possible alternative treatments). The highest peaks of MD search queries were temporally related to news about celebrities suffering from MD, to specific mass-media events or to news concerning pharmaceutic companies or scientific discoveries on MD. An increasing number of people use Google and Wikipedia to look for terms related to MD to obtain information on definitions, causes and symptoms, possibly to aid initial self-diagnosis. MD information demand and the actual prevalence of different MDs do not travel together: web search volume may mirrors patients' fears and worries about some particular disorders perceived as more serious than others, or may be driven by release of news about celebrities suffering from MD, "breaking news" or specific mass-media events regarding MD.

Googling environmental issues

Purpose– Queries as a pioneering measure of public attention on various social issues have elicited considerable scholarly attention. The purpose of this paper is to address two fundamental questions, as follows: first, how do we identify niche queries that internet users search for on specific social issues?; and second, what are the measurement properties of queries data in gauging public attention on social issues?Design/methodology/approach– The present study uses public attention on environmental issues in the USA as the empirical setting of research. An iterative framework is developed to identify niche queries to measure public attention on environmental issues. The measurement properties of queries data are assessed by comparing the dynamics of public attention on environmental issues captured by queries data with that measured by the “most important problem” (MIP) question in Gallup opinion polls.Findings– A list of 39 niche queries that internet users search for on environmental issues is identified. The dynamics of public attention on environmental issues determined by the search trends of these 39 queries is found to positively correlate with that measured by Gallup MIP polls, whereas both dynamics can forecast each other well in a 12-month time frame.Originality/value– The findings of the study possess methodological and practical implications. The study shows that queries data are complementary to, rather than substitutes of, public opinion polls in measuring public attention on environmental issues. The iterative framework developed in the study can be applied in future studies to help researchers identify valid queries to measure public attention on other social issues, as it can minimize researchers’ subjective biases in selecting search queries. Policymakers and environmentalists can utilize our approach to monitor the status of public attention on environmental issues and implement campaigns to mobilize favorable public opinion when the decline of public attention is predicted by the trends of web search queries.

Nowcasting Mobile Games Ranking Using Web Search Query Data

In recent years, the Internet has become embedded into the purchasing decision of consumers. The purpose of this paper is to study whether the Internet behavior of users correlates with their actual behavior in computer games market. Rather than proposing the most accurate model for computer game sales, we aim to investigate to what extent web search query data can be exploited to nowcast (contraction of “now” and “forecasting” referring to techniques used to make short-term forecasts) (predict the present status of) the ranking of mobile games in the world. Google search query data is used for this purpose, since this data can provide a real-time view on the topics of interest. Various statistical techniques are used to show the effectiveness of using web search query data to nowcast mobile games ranking.

National economic image and soft economic power evaluation

The article provides a review of national economic image, brand, and soft economic power evaluation methods based on opinion polls or objective indicators. The author elaborates and tests a webometric method for evaluation of specific components of national economic image and soft power. National economic image is measured as the share of positive messages in total number of messages about economy of a country. Soft economic power is described as national economic image adjusted for absolute amount of positive economic information about a country. Various key words and expressions are used for web search query to analyze specific components of economic image (trade, inflation, finance, budget, quality, competitiveness, investment, business, corruption, corporate governance, wages, economic cooperation or conflicts, economic growth or crisis). The advantages and limitations of the suggested approach are discussed. The sample for testing the method includes Poland, Ukraine, Russia, Turkey, Romania, Germany, and Moldova. Germany and Poland have almost no major weaknesses, while Russia and Ukraine have almost no major strengths among economic image components. Germany is a soft power leader in the majority of areas, followed by Russia, Turkey and Poland. Ukraine, Romania, and Moldova can improve their image and soft power by better real economic performance and wider coverage of positive economic information about these countries.

Modified Privacy Protection in Personalized Web Search

This proposed work aims to show the importance of personalized web search with improved privacy protection. To protect user profile based on Personalized Web Search (PWS), we need to consider few major issues during web site search processing. It is possible to improve efficient search based method with the personalized usage method of the accessing new user profile and prevent the privacy contents, which exists in the user profile method to adopt the risk privacy under control. User privacy can be safe without compromising the web search if personalized search quality. User Profile Search Engine (UPSE) is typically offline with all queries received from a same user. This article provides an inexpensive method for the user clientto conform whether to personalize a web search query in UPSE. This can avoid the unnecessary experience of the profile. Expected results can be evident that UPSE can achieve better quality based search results. So this proposed method is more applicable for personalized web portal.

Linking Annual Prescription Volume of Antidepressants to Corresponding Web Search Query Data: A Possible Proxy for Medical Prescription Behavior?

Persons using the Internet to retrieve medical information generate large amounts of health-related data, which are increasingly used in modern health sciences. We analyzed the relation between annual prescription volumes (APVs) of several antidepressants with marketing approval in Germany and corresponding web search query data generated in Google to test whether web search query volume may be a proxy for medical prescription practice. We obtained APVs of several antidepressants related to corresponding prescriptions at the expense of the statutory health insurance in Germany from 2004 to 2013. Web search query data generated in Germany and related to defined search terms (active substance or brand name) were obtained with Google Trends. We calculated correlations (Person's r) between the APVs of each substance and the respective annual "search share" values; coefficients of determination (R) were computed to determine the amount of variability shared by the 2 variables. Significant and strong correlations between substance-specific APVs and corresponding annual query volumes were found for each substance during the observational interval: agomelatine (r = 0.968, R = 0.932, P = 0.01), bupropion (r = 0.962, R = 0.925, P = 0.01), citalopram (r = 0.970, R = 0.941, P = 0.01), escitalopram (r = 0.824, R = 0.682, P = 0.01), fluoxetine (r = 0.885, R = 0.783, P = 0.01), paroxetine (r = 0.801, R = 0.641, P = 0.01), and sertraline (r = 0.880, R = 0.689, P = 0.01). Although the used data did not allow to perform an analysis with a higher temporal resolution (quarters, months), our results suggest that web search query volume may be a proxy for corresponding prescription behavior. However, further studies analyzing other pharmacologic agents and prescription data that facilitate an increased temporal resolution are needed to confirm this hypothesis.

Native-2-native: automated cross-platform code synthesis from web-based programming resources

For maximal market penetration, popular mobile applications are typically supported on all major platforms, including Android and iOS. Despite the vast differences in the look-and-feel of major mobile platforms, applications running on these platforms in essence provide the same core functionality. As an application is maintained and evolved, the resulting changes must be replicated on all the supported platforms, a tedious and error-prone programming process. Existing automated source-to-source translation tools prove inadequate due to the structural and idiomatic differences in how functionalities are expressed across major platforms. In this paper, we present a new approach---Native-2-Native---that automatically synthesizes code for a mobile application to make use of native resources on one platform, based on the equivalent program transformations performed on another platform. First, the programmer modifies a mobile application's Android version to make use of some native resource, with a plugin capturing code changes. Based on the changes, the system then parameterizes a web search query over popular programming resources (e.g., Google Code, StackOverflow, etc.), to discover equivalent iOS code blocks with the closest similarity to the programmer-written Android code. The discovered iOS code block is then presented to the programmer as an automatically synthesized Swift source file to further fine-tune and subsequently integrate in the mobile application's iOS version. Our evaluation, enhancing mobile applications to make use of common native resources, shows that the presented approach can correctly synthesize more than 86% of Swift code for the subject applications' iOS versions.

Query suggestion with diversification and personalization

Web search query suggestion is an important functionality that facilitates information seeking of search engine users. In existing work, the concepts of diversification and personalization have been individually introduced to query suggestion systems. In this paper, we propose a new query suggestion paradigm, Query Suggestion With Diversification and Personalization (QS-DP) to effectively integrate diversification and personalization into one unified framework. In the QS-DP, the suggested queries are effectively diversified to cover different facets of the input query while the ranking of the suggested queries are personalized to ensure that the top ones are those that align with a user’s personal preferences. We evaluate QS-DP on a commercial search engine query log against several existing query suggestion methods. The experimental results verify our hypothesis that diversification and personalization can be effectively integrated and they are able to enhance each other within the QS-DP framework, which significantly outperforms several strong baselines with respect to a series of metrics.

Latent entity space: a novel retrieval approach for entity-bearing queries

Analysis on Web search query logs has revealed that there is a large portion of entity-bearing queries, reflecting the increasing demand of users on retrieving relevant information about entities such as persons, organizations, products, etc. In the meantime, significant progress has been made in Web-scale information extraction, which enables efficient entity extraction from free text. Since an entity is expected to capture the semantic content of documents and queries more accurately than a term, it would be interesting to study whether leveraging the information about entities can improve the retrieval accuracy for entity-bearing queries. In this paper, we propose a novel retrieval approach, i.e., latent entity space (LES), which models the relevance by leveraging entity profiles to represent semantic content of documents and queries. In the LES, each entity corresponds to one dimension, representing one semantic relevance aspect. We propose a formal probabilistic framework to model the relevance in the high-dimensional entity space. Experimental results over TREC collections show that the proposed LES approach is effective in capturing latent semantic content and can significantly improve the search accuracy of several state-of-the-art retrieval models for entity-bearing queries.

Automatic Taxonomy Construction from Keywords via Scalable Bayesian Rose Trees

In this paper, we study a challenging problem of deriving a taxonomy from a set of keyword phrases. A solution can benefit many real-world applications because i) keywords give users the flexibility and ease to characterize a specific domain; and ii) in many applications, such as online advertisements, the domain of interest is already represented by a set of keywords. However, it is impossible to create a taxonomy out of a keyword set itself. We argue that additional knowledge and context are needed. To this end, we first use a general-purpose knowledgebase and keyword search to supply the required knowledge and context. Then, we develop a Bayesian approach to build a hierarchical taxonomy for a given set of keywords. We reduce the complexity of previous hierarchical clustering approaches from $O(n^2\; \log\; n)$ to $O(n\; \log\; n)$ using a nearest-neighbor-based approximation, so that we can derive a domain-specific taxonomy from one million keyword phrases in less than an hour. Finally, we conduct comprehensive large scale experiments to show the effectiveness and efficiency of our approach. A real life example of building an insurance-related web search query taxonomy illustrates the usefulness of our approach for specific domains.

Discovering and Understanding Word Level User Intent in Web Search Queries

Identifying and interpreting user intent are fundamental to semantic search. In this paper, we investigate the association of intent with individual words of a search query. We propose that words in queries can be classified as either content or intent, where content words represent the central topic of the query, while users add intent words to make their requirements more explicit. We argue that intelligent processing of intent words can be vital to improving result quality, and in this work we focus on intent word discovery and understanding. Our approach towards intent word detection is motivated by the hypotheses that query intent words satisfy certain distributional properties in large query logs similar to function words in natural language corpora. Following this idea, we first prove the effectiveness of our corpus distributional features, namely, word co-occurrence counts and entropies, towards function word detection for five natural languages. Next, we show that reliable detection of intent words in queries is possible using these same features computed from query logs. To make the distinction between content and intent words more tangible, we additionally provide operational definitions of content and intent words as those words that should match, and those that need not match, respectively, in the text of relevant documents. In addition to a standard evaluation against human annotations, we also provide an alternative validation of our ideas using click through data. Concordance of the two orthogonal evaluation approaches provide further support to our original hypothesis of the existence of two distinct word classes in search queries. Finally, we provide a taxonomy of intent words derived through rigorous manual analysis of large query logs.