Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Data Mining Practical Machine Learning Tools and Techniques

The automated categorization (or classification) of texts into predefined categories has witnessed a booming interest in the last 10 years, due to the increased availability of documents in digital form and the ensuing need to organize them. In the research community the dominant approach to this problem is based on machine learning techniques: a general inductive process automatically builds a classifier by learning, from a set of preclassified documents, the characteristics of the categories. The advantages of this approach over the knowledge engineering approach (consisting in the manual definition of a classifier by domain experts) are a very good effectiveness, considerable savings in terms of expert labor power, and straightforward portability to different domains. This survey discusses the main approaches to text categorization that fall within the machine learning paradigm. We will discuss in detail issues pertaining to three different problems, namely, document representation, classifier construction, and classifier evaluation.

/pdf/machine-learning-in-automated-text-categorization-e4pkjhyj3r.pdf

Machine learning in automated text categorization

An important part of our information-gathering behavior has always been to find out what other people think. With the growing availability and popularity of opinion-rich resources such as online review sites and personal blogs, new opportunities and challenges arise as people now can, and do, actively use information technologies to seek out and understand the opinions of others. The sudden eruption of activity in the area of opinion mining and sentiment analysis, which deals with the computational treatment of opinion, sentiment, and subjectivity in text, has thus occurred at least in part as a direct response to the surge of interest in new systems that deal directly with opinions as a first-class object.

This survey covers techniques and approaches that promise to directly enable opinion-oriented information-seeking systems. Our focus is on methods that seek to address the new challenges raised by sentiment-aware applications, as compared to those that are already present in more traditional fact-based analysis. We include material on summarization of evaluative text and on broader issues regarding privacy, manipulation, and economic impact that the development of opinion-oriented information-access services gives rise to. To facilitate future work, a discussion of available resources, benchmark datasets, and evaluation campaigns is also provided.

/pdf/opinion-mining-and-sentiment-analysis-3v5ya3qqiw.pdf

Opinion Mining and Sentiment Analysis

Part I: AN INTRODUCTION TO INQUIRY. 1. Human Inquiry and Science. 2. Paradigms, Theory, and Research. 3. The Ethics and Politics of Social Research. Part II: THE STRUCTURING OF INQUIRY: QUANTITATIVE AND QUALITATIVE. 4. Research Design. 5. Conceptualization, Operationalization, and Measurement. 6. Indexes, Scales, and Typologies. 7. The Logic of Sampling. Part III: MODES OF OBSERVATION: QUANTITATIVE AND QUALITATIVE. 8. Experiments. 9. Survey Research. 10. Qualitative Field Research. 11. Unobtrusive Research. 12. Evaluation Research. Part IV: ANALYSIS OF DATA:QUANTITATIVE AND QUALITATIVE . 13. Qualitative Data Analysis. 14. Quantitative Data Analysis. 15. Reading and Writing Social Research. Appendix A. Using the Library. Appendix B. Random Numbers. Appendix C. Distribution of Chi Square. Appendix D. Normal Curve Areas. Appendix E. Estimated Sampling Error.

社会研究方法基础 = The Basics of Social Research

Faceted blog distillation aims at finding blogs with recurring interest to a topic while satisfying a specific facet of interest. In this paper we focus on the personal facet and propose a method that uses opinion features as indicators of personal content. Experimental results on TREC BLOG08 data-set confirm our intuition that personal blogs are more opinionated.

Personal blog retrieval using opinion features

Recent technological improvements allow us to capture an increasing share of our everyday experiences, e.g. holidays, shopping routines, or sports activities, and store them in a digital format. An interesting avenue to explore in this context is how reviewing such captured content can improve one's memories of the original events. In this position paper, we describe a planned experiment to investigate the impact of such captured recordings (and their subsequent review) on supporting work meetings. We provide the planned study procedure, explain the envisioned apparatus and metric, and describe the technology used to support the review activity.

Measuring the effect of cued recall on work meetings

Invited Papers.- Wittgenstein, Language and Information: "Back to the Rough Ground!".- Text, Co-text, Context and the Documentary Continuum.- Representing Context.- The Sense of Information: Understanding the Cognitive Conditional Information Concept in Relation to Information Acquisition.- Practical Implications of Handling Multiple Contexts in the Principle of Polyrepresentation.- Information Sharing and Timing: Findings from Two Finnish Organizations.- Context and Relevance in Information Seeking.- Contexts of Relevance for Information Retrieval System Design.- Searching for Relevance in the Relevance of Search.- Information Searching Behavior: Between Two Principles.- Context and Information.- Bradford's Law of Scattering: Ambiguities in the Concept of "Subject".- The Instrumentality of Information Needs and Relevance.- Lifeworld and Meaning - Information in Relation to Context.- Contextualised Information Seeking.- Personometrics: Mapping and Visualizing Communication Patterns in R&D Projects.- Annotations as Context for Searching Documents.- Conceptual Indexing Based on Document Content Representation.- Agendas for Context.- What's the Deal with the Web/Blogs/the Next Big Technology: A Key Role for Information Science in e-Social Science Research?.- Assessing the Roles That a Small Specialist Library Plays to Guide the Development of a Hybrid Digital Library.- Power Is Information: South Africa's Promotion of Access to Information Act in Context.- Context and Documents.- A Bibliometric-Based Semi-automatic Approach to Identification of Candidate Thesaurus Terms: Parsing and Filtering of Noun Phrases from Citation Contexts.- Context Matters: An Analysis of Assessments of XML Documents.- Workshops.- Developing a Metadata Lifecycle Model.- Evaluating User Studies in Information Access.

/pdf/context-nature-impact-and-role-3zuurjgomu.pdf

Context: nature, impact, and role

Structured document retrieval requires different user graphical interfaces from standard Information Retrieval. An Information Retrieval system dealing with structured documents has to enable a user to query, browse retrieved documents, provide query refinement and relevance feedback based not only on full documents, but also on specific document structural parts. In this paper, we present a new graphical user interface for structured document retrieval that provides the user with an intuitive and powerful set of tools for structured document searching, retrieved list navigation, and search refinement. We also present the results of a preliminary evaluation of the interface which highlights strengths and weaknesses of the current implementation and suggests directions of future work.

A Graphical User Interface for Structured Document Retrieval

In distributed information retrieval systems, document overlaps occur frequently across results from different resources. This is especially the case for meta-search engines which merge results from several web search engines. This paper addresses the problem of merging results exploiting overlaps in order to achieve better performance. New algorithms for merging results are proposed, which take advantage of the use of duplicate documents in two ways: one correlates scores from different results; the other regards duplicates as increasing evidence of being relevant to the given query. An extensive experimentation has demonstrated that these methods are effective.

Fabio Crestani

Papers

Personal blog retrieval using opinion features

Measuring the effect of cued recall on work meetings

Context: nature, impact, and role

A Graphical User Interface for Structured Document Retrieval

New Methods of Results Merging for Distributed Information Retrieval