Including Search Engines Research Articles

Market Definition and Market Power in Data: The Case of Online Platforms

With the emergence of data as an asset for market players operating in the digital economy, questions have risen about the relevance of data for competition enforcement. This article focuses on the role of data in the competitive process between online platforms including search engines, social networks and e-commerce platforms. After arguing that situations can be identified in which access to data is a competitive advantage for incumbents and an entry barrier for new entrants, it is analysed how a relevant market for data can be defined and how market power can be established in such a market. Most providers of online platforms do not trade data as a stand-alone product as a result of which no supply and demand exists and no relevant product market for data can be defined under current competition law standards. However, it may still be appropriate for competition authorities in this situation to assess proposed concentrations and alleged abusive behaviour of dominant firms beyond the relevant markets for the services provided to users and advertisers. By defining a wider market for data, a form of potential competition can be taken into account whereby market players also compete for the asset that is used as an input to develop or improve services offered on online platforms.

Market Definition and Market Power in Data: The Case of Online Platforms

With the emergence of data as an asset for market players operating in the digital economy, questions have risen about the relevance of data for competition enforcement. This article focuses on the role of data in the competitive process between online platforms including search engines, social networks and e-commerce platforms. After arguing that situations can be identified in which access to data is a competitive advantage for incumbents and an entry barrier for new entrants, it is analyzed how a relevant market for data can be defined and how market power can be established in such a market. Most providers of online platforms do not trade data as a stand-alone product as a result of which no supply and demand exists and no relevant product market for data can be defined under current competition law standards. However, it may still be appropriate for competition authorities in this situation to assess proposed concentrations and alleged abusive behaviour of dominant firms beyond the relevant markets for the services provided to users and advertisers. By defining a wider market for data, a form of potential competition can be taken into account whereby market players also compete for the asset that is used as an input to develop or improve services offered on online platforms.

Search Personalization Using Machine Learning

Query-based search is commonly used by many businesses to help consumers find information/products on their websites. Examples include search engines (Google, Bing), online retailers (Amazon, Macy's), and entertainment sites (Hulu, YouTube). Nevertheless, a significant portion of search sessions are unsuccessful, i.e., do not provide information that the user was looking for. We present a machine learning framework that improves the quality of search results through automated personalization based on a user's search history. Our framework consists of three modules -- (a) Feature generation, (b) NDCG-based LambdaMART algorithm, and (c) Feature selection wrapper. We estimate our framework on large-scale data from a leading search engine using Amazon EC2 servers. We show that our framework offers a significant improvement in search quality compared to non-personalized results. We also show that the returns to personalization are monotonically, but concavely increasing with the length of user history. Next, we find that personalization based on short-term history or within-session behavior is less valuable than long-term or across-session personalization. We also derive the value of different feature sets -- user-specific features contribute over 50% of the improvement and click-specific over 28%. Finally, we demonstrate scalability to big data and derive the set of optimal features that maximize accuracy while minimizing computing speed.

Project Information Literacy's Research Summary: Lifelong Learning Study, Phase Two and the Online Survey

A trends report with preliminary findings from the lifelong learning survey Project Information Literacy administered to recent college graduates from 10 US colleges and universities during October 2014 through December 2014. Survey data was collected from college graduates (n=1,651) about their lifelong learning needs and information practices for their personal development, engagement in local communities, and for staying employable and competitive in the workplace. A majority of grads had a diffuse set of complex learning needs, which often overlapped. Grads relied on a mixed variety of information sources in their lives for staying informed, including search engines, books, blogs, social networks, co-workers, and friends. About half of the respondents credited college with helping them develop their competencies for extracting the information needed and evaluating the credibility of content. The survey results are revealing about the information competencies a majority of students may develop and master during their college years. Still, lifelong learning was not necessarily easier for grads: Many "strongly agreed" with the statement that finding time as well as locating affordable sources for learning after college was challenging.

Governance of Online Intermediaries: Observations from a Series of National Case Studies

Online intermediaries in various forms – including search engines, social media, or app platforms – play a constitutive role in today’s digital environment. They have become a new type of powerful institution in the 21st century that shape the public networked sphere, and are subject to intense and often controversial policy debates. This paper focuses on one particular force shaping the emergence and future evolution of online intermediaries: the rapidly changing landscape of intermediary governance at the intersection of law, technology, norms, and markets. Building upon eight in-depth case studies and use cases, respectively, this paper seeks to distill key observations and provide a high-level analysis of some of the structural elements that characterize varying governance regimes, with a focus on intermediary liability regimes and their evolution.Analyzing online intermediary governance issues from multiple perspectives, and in the context of different cultures and regulatory frameworks, immediately creates basic problems of semantic interoperability. Lacking a universally agreed-upon definition, the synthesis paper and its’ underlying case studies are based on a broad and phenomenon-oriented notion of online intermediaries, as further described below. In methodological terms, the observations shared in the synthesis paper offer a selective reading and interpretation by the authors of the broader take-ways of a diverse set of case studies examining online intermediary governance frameworks and issues in Brazil, the European Union, India, South Korea, the United States, Thailand, Turkey, and Vietnam. These case studies, in turn, have emerged in the context of an international research pilot by the Global Network of Internet & Society Research Centers (NoC), through a process of in-person consultations and remote collaborations among the researchers, and are based on a set of broader questions regarding the role of online intermediaries in the digital age.

Survey of Machine Learning Techniques in Textual Document Classification

Classification of Document points towards associating one or more predefined categories based on the likelihood expressed by the training set of labeled documents. Many machine learning algorithms plays an important role in training the system with predefined categories. The importance of Machine learning approach has felt because of which the study has been taken up for text classification based on the statistical event models available. The aim of this paper is to present the important techniques and methodologies that are employed for text documents classification, at the same time making awareness of some of the interesting challenges that remain to be solved, focused mainly on text representation and machine learning techniques. Keywords: mining, Web mining, Documents classification, Information retrieval, Event models. I. Introduction With the rapid growth of the World Wide Web and increasing availability of electronic documents, the task of automatic categorization of documents became important for organizing the information and knowledge discovery. Proper categorization of electronic documents, online news, blogs, e-mails and digital libraries requires text mining, machine learning and natural language processing techniques to extract required knowledge information. The term Text document refers to written, printed, or online that presents or communicates narrative or tabulated data in the form of an article, letter, memorandum, report, etc. The expresses a vast range of information, but encodes the information in the form that is difficult to decipher automatically. In the existing online word huge amount of textual information is available in textual form in databases and various sources. The information may be available in structured and unstructured form. Unstructured means data that does not reside in fixed locations. The term generally refers to free-form text, which is present everywhere. Data that resides in fixed fields within a record or file that data is termed as a structured data. Relational databases and spreadsheets are examples of structured data. In reality a large portion of the available information does not appear in structured databases but rather in collections of text articles drawn from various sources. Unstructured information refers to computerized information that either does not have a data model or the one that is not easily usable by a computer program. The term distinguishes such information from data stored in field form in databases or annotated in documents. However, data mining deals with structured data, whereas text presents special characteristics and is unstructured. The important task is how these documented data can be properly retrieved, presented and classified. Extraction, Integration and classification of electronic documents from different sources and knowledge information discovery from these documents are important. In data mining, Machine learning is often used for Prediction or Classification. Classification involves finding rule that partition the data into disjoint groups. The input for the classification is the training data set, whose class labels are already known. Classifications analyze the training data set and construct a model based on the class label. The goal of classification is to build a set of models that can correctly predict the class of the different objects. Machine learning is an area of artificial intelligence concerned with the development of techniques which allow computers to learn. More specifically, machine learning is a method for creating computer programs by the analysis of data sets since machine learning study the analysis of data. Some machine learning systems attempt to eliminate the need for human intuition in the analysis of the data, while others adopt a collaborative approach between human and machine. Human intuition cannot be entirely eliminated since the designer of the system must specify how the data are to be represented and what mechanisms will be used to search for a characterization of the data. Machine learning has a wide spectrum of applications including search engines, medical diagnosis, detecting credit card fraud, stock market analysis, classifying DNA sequences, speech and handwriting recognition, game playing and robot locomotion.

IKernel: Exact indexing for support vector machines

SVM (Support Vector Machine) is a well-established machine learning methodology popularly used for learning classification, regression, and ranking functions. Especially, SVM for rank learning has been applied to various applications including search engines or relevance feedback systems. A ranking function F learned by SVM becomes the query in some search engines: A relevance function F is learned from the user’s feedback which expresses the user’s search intention, and top-k results are found by evaluating the entire database by F. This paper proposes an exact indexing solution for the SVM function queries, which is to find top-k results without evaluating the entire database. Indexing for SVM faces new challenges, that is, an index must be built on the kernel space (SVM feature space) where (1) data points are invisible and (2) the distance function changes with queries. Because of that, existing top-k query processing algorithms, or existing metric-based or reference-based indexing methods are not applicable. We first propose key geometric properties of the kernel space – ranking instability and ordering stability – which is crucial for building indices in the kernel space. Based on them, we develop an index structure iKernel and processing algorithms. We then present clustering techniques in the kernel space to enhance the pruning effectiveness of the index. According to our experiments, iKernel is highly effective overall producing 1–5% of evaluation ratio on large data sets.

Internet-Based Recruitment to a Depression Prevention Intervention: Lessons From the Mood Memos Study

BackgroundRecruiting participants to randomized controlled trials of health interventions can be very difficult. Internet-based recruitment is becoming an increasingly important mode of recruitment, yet there are few detailed accounts of experiences recruiting participants to mental health interventions.ObjectiveTo report on our experience with Internet-based recruitment to an online depression prevention intervention and pass on lessons we learned.MethodsParticipants were recruited to the Mood Memos study, an online preventive depression intervention, purely through Internet-based sources. The study was targeted to adults with subthreshold depression symptoms from several English-speaking countries. A variety of online recruitment sources were trialed, including search engine advertising (Google, Yahoo!, Bing), Facebook advertising, posts in forums and online noticeboards, and promotion through relevant websites and email newsletters of mental health organizations.ResultsThe study website received visits from 94,808 individuals over the 14-month recruitment period. The recruitment target was reached with 1699 individuals signing up to the randomized controlled trial and 1326 fully enrolling. Most visitors arrived via Google advertising, which promoted a depression-screening questionnaire. Google advertising accounted for nearly half of the total participants who signed up to the study, at an average cost of AUD $12 per participant. Promoting the study through trustworthy organizations and websites known to participants was also effective. Recruitment techniques that were less effective were contacting forums, email groups, and community noticeboards.ConclusionsSeveral techniques, including Google advertising, were successful in recruiting participants to a trial evaluating an online depression intervention. Results suggest that Internet-based recruitment to mental health interventions is feasible and can be relatively affordable.Trial RegistrationACTRN12609000925246

Undergraduate students' interaction with online information resources in their academic tasks

PurposeThis paper aims to identify the opinions of undergraduate students on the importance of internet‐based information sources when they undertake academic tasks.Design/methodology/approachBased on a set of identified typical academic tasks for undergraduate students, three research questions were designed around the students' usage and views of information resources for completing these tasks. Web‐accessible questionnaires were used to collect data from participants in two universities in the USA and China, and the data were analyzed using quantitative methods, which included several statistic methods.FindingsThe results confirm that undergraduate students use different information resources for various academic tasks. In their tasks, online electronic resources including search engines are the most commonly used resources, particularly for complex academic tasks. Social networking sites are not used for the students' individual academic tasks, and traditional resources still play equal or more important roles in certain specific academic tasks. Students in collaborative tasks look for resources that make it easy to share documents. Participants from the two countries also exhibit interesting and important differences in their usage of information resources.Originality/valueThis study examines undergraduate students' usages and views of different information resources in their various academic tasks, and pays special attention to the impacts of being from their different countries. The study also considers both students' individual academic tasks and collaborative tasks. This study is an invaluable addition to the information seeking behaviour literature.

First Amendment Protection for Search Engine Search Results -- White Paper Commissioned by Google

Once, the leading sources to which people turned for useful information were newspapers, guidebooks, and encyclopedias. Today, these sources also include search engine results, which people use (along with other sources) to learn about news, local institutions, products, services, and many other matters. Then and now, the First Amendment has protected all these forms of speech from government attempts to regulate what they present or how they present it. And this First Amendment protection has applied even when the regulations were motivated by a concern about what some people see as “fairness.” Google, Microsoft’s Bing, Yahoo! Search, and other search engines are speakers. First, they sometimes convey information that the search engine company has itself prepared or compiled (such as information about places appearing in Google Places). Second, they direct users to material created by others, by referencing the titles of Web pages that the search engines judge to be most responsive to the query, coupled with short excerpts from each page. Such reporting about others’ speech is itself constitutionally protected speech.Third, and most valuably, search engines select and sort the results in a way that is aimed at giving users what the search engine companies see as the most helpful and useful information. (That is how each search engine company tries to keep users coming back to it rather than to its competitors.) This selection and sorting is a mix of science and art: It uses sophisticated computerized algorithms, but those algorithms themselves inherently incorporate the search engine company engineers’ judgments about what material users are most likely to find responsive to their queries. In this respect, each search engine’s editorial judgment is much like many other familiar editorial judgments made by newspapers, guidebooks, and Web sites. All these speakers must decide: Out of the thousands of possible items that could be included, which to include, and how to arrange those that are included? Such editorial judgments may differ in certain ways: For example, a newspaper also includes the materials that its editors have selected and arranged, while the speech of DrudgeReport.com or a search engine consists almost entirely of the selected and arranged links to others’ material. But the judgments are all, at their core, editorial judgments about what users are likely to find interesting and valuable. And all these exercises of editorial judgment are fully protected by the First Amendment. Note that this White Paper was commissioned by Google, rather than a traditional academic article.

Structure–semantics interplay in complex networks and its effects on the predictability of similarity in texts

The classification of texts has become a major endeavor with so much electronic material available, for it is an essential task in several applications, including search engines and information retrieval. There are different ways to define similarity for grouping similar texts into clusters, as the concept of similarity may depend on the purpose of the task. For instance, in topic extraction similar texts mean those within the same semantic field, whereas in author recognition stylistic features should be considered. In this study, we introduce ways to classify texts employing concepts of complex networks, which may be able to capture syntactic, semantic and even pragmatic features. The interplay between various metrics of the complex networks is analyzed with three applications, namely identification of machine translation (MT) systems, evaluation of quality of machine translated texts and authorship recognition. We shall show that topological features of the networks representing texts can enhance the ability to identify MT systems in particular cases. For evaluating the quality of MT texts, on the other hand, high correlation was obtained with methods capable of capturing the semantics. This was expected because the golden standards used are themselves based on word co-occurrence. Notwithstanding, the Katz similarity, which involves semantic and structure in the comparison of texts, achieved the highest correlation with the NIST measurement, indicating that in some cases the combination of both approaches can improve the ability to quantify quality in MT. In authorship recognition, again the topological features were relevant in some contexts, though for the books and authors analyzed good results were obtained with semantic features as well. Because hybrid approaches encompassing semantic and topological features have not been extensively used, we believe that the methodology proposed here may be useful to enhance text classification considerably, as it combines well-established strategies.

Effective Blogging for Libraries

Effective Blogging for Libraries

History Of Search Engines

As the number of sites on the Web increased in the mid-to-late 90s, search engines started appearing to help people find information quickly. Search engines developed business models to finance their services, such as pay per click programs offered by Open Text in 1996 and then Goto.com in 1998. Goto.com later changed its name to Overture in 2001, and was purchased by Yahoo! in 2003, and now offers paid search opportunities for advertisers through Yahoo! Search Marketing. Google also began to offer advertisements on search results pages in 2000 through the Google Ad Words program. By 2007, pay-per-click programs proved to be primary money-makers for search engines. In a market dominated by Google, in 2009 Yahoo! and Microsoft announced the intention to forge an alliance. The Yahoo! & Microsoft Search Alliance eventually received approval from regulators in the US and Europe in February 2010. Search engine optimization consultants expanded their offerings to help businesses learn about and use the advertising opportunities offered by search engines, and new agencies focusing primarily upon marketing and advertising through search engines emerged. The term "Search Engine Marketing" was proposed by Danny Sullivan in 2001 to cover the spectrum of activities involved in performing SEO, managing paid listings at the search engines, submitting sites to directories, and developing online marketing strategies for businesses, organizations, and individuals. Some of the latest theoretical advances include Search Engine Marketing Management (SEMM). SEMM relates to activities including SEO but focuses on return on investment (ROI) management instead of relevant traffic building (as is the case of mainstream SEO). SEMM also integrates organic SEO, trying to achieve top ranking without using paid means of achieving top in search engines, and PayPerClick SEO. For example some of the attention is placed on the web page layout design and how content and information is displayed to the website visitor.

“I did not realize so many options are available”: Cognitive authority, emerging adults, and e-mental health

Approximately one in five people in developed countries such as Canada and Australia will experience mental illness during their lifespan, and this statistic may be even higher for people between the ages of 18 and 25. Due to widespread stigma and other issues, access to mental health care is limited. However, given the heavy online use by people in this age range as well as the prevalence of existing online health information, it is possible that “e-mental health,” when delivered in an efficacious and engaging format, could be a viable dissemination option. A quantitative and qualitative online survey was distributed to university students in order to determine their opinions of currently available e-mental health resources. Within the sample ( n = 1308), text-based searching on Web sites such as Google as well as reading informational Web sites were the most highly preferred and utilized methods. However, a significant knowledge gap regarding the existence of other e-mental health formats was noted. Suggestions for future research include search engine optimization changes and continued development of e-mental health resources.

The Development and Implementation of an Outreach Program to Identify Acute and Recent HIV Infections in New York City

Introduction: Since 2004, the authors have been operating First Call NYU, an outreach program to identify acute and recent HIV infections, also called primary HIV infections, among targeted at-risk communities in the New York City (NYC) metropolitan area. Materials and Methodology:First Call NYU employed mass media advertising campaigns, outreach to healthcare providers in NYC, and Internet-based efforts including search engine optimization (SEO) and Internet-based advertising to achieve these goals.Results:Between October 2004 and October 2008, 571 individuals were screened through this program, leading to 446 unique, in-person screening visits. 47 primary HIV infections, including 14 acute and 33 recent HIV infections, were identified.Discussion:Internet and traditional recruitment methods can be used to increase self-referrals for screening following possible exposure to HIV. Conclusion:Community education of at-risk groups, with the goal of increased self-diagnosis of possible acute HIV infection, may be a useful addition to traditional efforts to identify such individuals.

The e-Rise and Fall of Social Problems: The Blogosphere as a Public Arena

Social problems theory has yet to fully address the impact that new communication technologies are having on the claims-making process. This article examines the emergence of the blogosphere as a cultural phenomenon that provides claims-makers with a powerful new public arena to advance social problem claims. Using Stephen Hilgartner and Charles Bosk’s (1988) public arenas model of social problem construction, blog-generated problem claims are examined to analyze how Internet driven social problems compete for public attention. Findings suggest that blogs make the claims-making process more efficient, offer expanded carrying capacity compared to traditional arenas, and provide outsider claims-makers with greater opportunity to have a voice in social problems construction. Still, only a small number of blogs have become recognized as claims-making arenas; they still rely on traditional principles of selection; and bloggers face the same competition for mainstream media attention as claims-makers using traditional arenas. Keywords: blog, blogosphere, Internet, new media, public arena. Mainstream news media maintain a gatekeeping function that serves to control the flow of information to audiences. Some claims find it harder to gain media access or to receive coverage (Jacobs 2000). In recent years, proponents of the Internet have proclaimed that new media technology will lead to a democratization of mass media (Rodman 2003). Since the deregulation of the Internet in 1995, users have quickly adapted to and become engaged in an online environment that can transmit large volumes of information in real time for relatively low cost (Plant 2004). The expansion of mass media into cyberspace has already created countless new sources for news: Web sites presented by the mainstream press and sites unique to the Internet, including search engines, message boards, and blogs, may have the potential to diffuse the gatekeeping function of traditional media, thereby altering their agenda-setting function (Williams and Delli Carpini 2004). In particular, the emergence of the blogosphere as an Internet-based claims-making arena may profoundly affect the process of social problems construction. This article expands Stephen Hilgartner and Charles L. Bosk’s (1988) public arenas model of social problems construction by exploring how the blogosphere increases the overall carrying capacity for problem claims, expands the opportunities for outsider claims-makers to promote social problems, and provides new avenues through which insider and secondary claims can be disseminated. Analysis of social problems constructed, in part, through claims made by bloggers also serves to verify the findings of Sheldon Ungar (1992) and Jerry Williams and R. S. Frey (1997) that dramatic real world events serve as focal issues that enhance audience receptiveness of problem claims. 1 Still, while blogs provide novel arenas where problem

Accessing Legal and Regulatory Information in Internet Resources and Documents

SUMMARY The quantity of legal and government information accessible through online delivery has increased tremendously. Information technology has revolutionized the way much of this information is disseminated by government and the way that researchers access it. The increase in publication can be attributed to two factors: the government's efforts to create a virtual depository of publications and the capability of Internet technology to provide personalized “feeds” of topic-specific information through Web logs and RSS news aggregators. In this article, finding tools, primary and secondary sources of legal and regulatory government information will be identified and described including search engines and guides.

Endometriosis

Endometriosis is a puzzling disease characterized by pelvic pain, infertility, allergies, fatigue, and bowel problems. It is a non-lethal medical condition that disables only women and frustrates physicians who are frequently limited in their treatment success. Recently endometriosis has been linked with endocrine, environmental, genetic, and immune factors. The public health impact of endometriosis is significant in terms of suffering, lost income, infertility, stress on families, and medical costs. Resources for endometriosis on the Internet include search engines, medical sources, advocacy sources, U.S. government sources, personal Web sites, public library databases, and social science sources.

Are webliographies still in use?

Information retrieval has emerged as one of the most important areas of study in the field of information science during the past decades. Since the invention of the World Wide Web, some solutions have been planned to solve the problem of digital information retrieval, including search engines and Web directories. A webliography is a special type of Web directory and consists of an enumerative list of hypertext links to the scientific sources of information on the Net. According to the last time of updating of the webliographies, this article tries to provide an answer to this question: are webliographies still in use? In order to do this, two searches on webliography and webography made using Google. Among 357 updated and retrieved hits, 303 items have been updated at the second period of webliographies lifetime. It concludes that doing some research projects on webliographies is not only possible, but essential.

Measurement of online visibility and its impact on Internet traffic

The Internet landscape is an increasingly crowded space where hundreds of thousands of companies are fighting for attention. Some of these companies are universally known (Yahoo!, Amazon) while others are more obscure (Literary Guild, Just Say Wow). Regardless of their level of Internet presence, all these companies are interested in improving their position, particularly now that managers are asked to justify their online activities. To help managers assess their online presence, and give them a way to compare their position relative to their competitors, this article aims to develop a measure of a company's online visibility. Moreover, this study seeks to understand what factors drive online visibility, and how. The visibility measure we develop captures the extent to which a user is likely to come across an online reference to a company's Web site. It is based on data collected from multiple sources that include search engine results, Web-site contents, and online directory listings. It is calibrated using a large-scale telephone survey, and is validated using data obtained from Web crawlers and Internet consumer-panel sources. In the latter part of the article, we show how managers can use the visibility measure to compare the online presence of their company to their competitors’. We also show how they can use this information to perform scenario analysis aimed at finding the best way to improve their online position.