Pranam Kolari

Time is of the essence: improving recency ranking using Twitter data

Realtime web search refers to the retrieval of very fresh content which is in high demand. An effective portal web search engine must support a variety of search needs, including realtime web search. However, supporting realtime web search introduces two challenges not encountered in non-realtime web search: quickly crawling relevant content and ranking documents with impoverished link and click information. In this paper, we advocate the use of realtime micro-blogging data for addressing both of these problems. We propose a method to use the micro-blogging data stream to detect fresh URLs. We also use micro-blogging data to compute novel and effective features for ranking fresh URLs. We demonstrate these methods improve effective of the portal web search engine for realtime web search.

Using Ontologies in the Semantic Web: A Survey

The Semantic Web is well recognized as an effective infrastructure to enhance visibility of knowledge on the Web. The core of the Semantic Web is “ontology”, which is used to explicitly represent our conceptualizations. Ontology engineering in the Semantic Web is primarily supported by languages such as RDF, RDFS and OWL. This chapter discusses the requirements of ontology in the context of the Web, compares the above three languages with existing knowledge representation formalisms, and surveys tools for managing and applying ontology. Advantages of using ontology in both knowledge-base-style and database-style applications are demonstrated using three real world applications.

Web mining: research and practice

Web mining techniques seek to extract knowledge from Web data. This article provides an overview of past and current work in the three main areas of Web mining research - content, structure, and usage - as well as emerging work in semantic Web mining.

Enhancing Web privacy protection through declarative policies

The platform for privacy preferences (P3P) is a W3C framework for Web privacy management. It provides a standard vocabulary that Websites can use to describe their privacy practices. The presence of Web site published P3P policies enable users to configure Web browsers to allow, block or warn users during access and data exchange with Websites. Its a good idea that unfortunately is rarely used. We identify three primary reasons: (i) the languages available to describe user privacy preferences are not sufficiently expressive, (ii) P3P policies published by Web sites are not trusted by users and (iii) P3P framework does not provide a coherent view of available privacy protection mechanisms to the user towards addressing these issues; we present enhancements to the P3P framework. We use a more expressive policy language based on deontic concepts to describe user privacy-related policies, constraints and preferences. We introduce a new trust model for Websites and describe its use in user privacy preferences. Finally, we present sample policies to demonstrate the relevance of our work and offer it as an effective starting point towards enhancing Web privacy management.

Improving recency ranking using twitter data

In Web search and vertical search, recency ranking refers to retrieving and ranking documents by both relevance and freshness. As impoverished in-links and click information is the the biggest challenge for recency ranking, we advocate the use of Twitter data to address the challenge in this article. We propose a method to utilize Twitter TinyURL to detect fresh and high-quality documents, and leverage Twitter data to generate novel and effective features for ranking. The empirical experiments demonstrate that the proposed approach effectively improves a commercial search engine for both Web search ranking and tweet vertical ranking.

Ensembles in adversarial classification for spam

The standard method for combating spam, either in email or on the web, is to train a classifier on manually labeled instances. As the spammers change their tactics, the performance of such classifiers tends to decrease over time. Gathering and labeling more data to periodically retrain the classifier is expensive. We present a method based on an ensemble of classifiers that can detect when its performance might be degrading and retrain itself, all without manual intervention. Experiments with a real-world dataset from the blog domain show that our methods can significantly reduce the number of times classifiers are retrained when compared to a fixed retraining schedule, and they maintain classification accuracy even in the absence of manually labeled examples.

Policy management of enterprise systems: a requirements study

Policy enabled applications are being increasingly employed to support responsive information technology services. In competitive business environments, such services increase adaptability of both software and the processes they implement through externalized business and security logic. Over the last decade this has driven both industry and academia to contribute to policy research and engineering, by developing specification languages, frameworks and toolkits. Since this work has typically been applied to and evaluated using new enterprise solutions, policy management for existing applications has been less well studied. In this paper we share our experiences on policy enabling an existing Web based solution, together with identifying new policy enabling requirements from a specific class of enterprise systems

Temporal query log profiling to improve web search ranking

Temporal information can be leveraged and incorporated to improve web search ranking. In this work, we propose a method to improve the ranking of search results by identifying the fundamental properties of temporal behavior of low-quality hosts and spam-prone queries in search logs and modeling those properties as quantifiable features. In particular, we introduce the concepts of host churn, a measure of changes in host visibility for user queries, and query volatility, a measure of semantic instability of query results, and propose the methods for construction of temporal profiles from search query logs that can be used for estimation of a set of features based on the introduced concepts. The utility of the proposed concepts has been experimentally demonstrated for two language-independent search tasks: the regression-based ranking of search results and a novel classification problem of detecting spam-prone queries introduced in this work.