ABSTRACT. A keyword query is the representation of the information need of a user, and is the result of a complex cognitive process which often results in under-specification. We propose an unsupervised method namely Latent Concept Modeling (LCM) for mining and modeling latent search concepts in order to recreate the conceptual view of the original information need. We use Latent Dirichlet Allocation (LDA) to exhibit highly-specific query-related topics from pseudo-relevant feedback documents. We define these topics as the latent concepts of the user query. We perform a thorough evaluation of our approach over two large ad-hoc TREC collections. Our findings reveal that the proposed method accurately models latent concepts, while being very effective in a query expansion retrieval setting.RÉSUMÉ. Une requête est la représentation du besoin d'information d'un utilisateur, et est le résultat d'un processus cognitif complexe qui mène souvent à un mauvais choix de mots-clés. Nous proposons une méthode non supervisée pour la modélisation de concepts implicites d'une requête, dans le but de recréer la représentation conceptuelle du besoin d'information initial. Nous utilisons l'allocation de Dirichlet latente (LDA) pour détecter les concepts implicites de la requête en utilisant des documents pseudo-pertinents. Nous évaluons cette méthode en profondeur en utilisant deux collections de test de TREC. Nous trouvons notamment que notre approche permet de modéliser précisément les concepts implicites de la requête, tout en obtenant de bonnes performances dans le cadre d'une recherche de documents.

Understanding the nature and dynamics of conflicting opinions is a profound and challenging issue. In this paper we address several aspects of the issue through a study of more than 3,000 Amazon customer reviews of the controversial bestseller The Da Vinci Code, including 1,738 positive and 918 negative reviews. The study is motivated by critical questions such as: What are the differences between positive and negative reviews? What is the origin of a particular opinion? How do these opinions change over time? To what extent can differentiating features be identified from unstructured text? How accurately can these features predict the category of a review? We first analyze terminology variations in these reviews in terms of syntactic, semantic, and statistic associations identified by TermWatch and use term variation patterns to depict underlying topics. We then select the most predictive terms based on log likelihood tests and demonstrate that this small set of terms classifies over 70% of the conflicting reviews correctly. This feature selection process reduces the dimensionality of the feature space from more than 20,000 dimensions to a couple of hundreds. We utilize automatically generated decision trees to facilitate the understanding of conflicting opinions in terms of these highly predictive terms. This study also uses a number of visualization and modeling tools to identify not only what positive and negative reviews have in common, but also they differ and evolve over time.

We study a new content-based method for the evaluation of text summarization systems without human models which is used to produce system rankings. The research is carried out using a new content-based evaluation framework called FRESA to compute a variety of divergences among probability distributions. We apply our comparison framework to various well-established content-based evaluation measures in text summarization such as COVERAGE, RESPONSIVENESS, PYRAMIDS and ROUGE studying their associations in various text summarization tasks including generic multi-document summarization in English and French, focus-based multi-document summarization in English and generic single-document summarization in French and Spanish.

En este artı́culo abordamos el tema de la generación automática de frases literarias, que es una parte importante de los estudios relacionados al área de la Creatividad Computacional (CC). Proponemos tres modelos de generación textual guiados por un contexto, basados principalmente en algoritmos estadı́sticos y análisis sintáctico superficial. Los textos generados fueron evaluados por siete personas a partir de 4 criterios: gramaticalidad, coherencia, relación con el contexto y una adaptación del test de Turing, en donde se pidio a los evaluadores clasificar los textos en: textos generados automáticamente y textos generados por humanos. Los resultados obtenidos son bastante alentadores.

No abstract

Abstract. INEX investigates focused retrieval from structured documents by providing large test collections of structured documents, uniform evaluation measures, and a forum for organizations to compare their results. This paper reports on the INEX 2014 evaluation campaign, which consisted of three tracks: The Interactive Social Book Search Track investigated user information seeking behavior when interacting with various sources of information, for realistic task scenarios, and how the user interface impacts search and the search experience. The Social Book Search Track investigated the relative value of authoritative metadata and usergenerated content for search and recommendation using a test collection with data from Amazon and LibraryThing, and user profiles and personal catalogues. The Tweet Contextualization Track investigated tweet contextualization, helping a user to understand a tweet by providing him with a short background summary generated from relevant Wikipedia passages aggregated into a coherent summary. INEX 2014 was an exciting year for INEX in which we for the third time ran our workshop as part of the CLEF labs in order to facilitate knowledge transfer between the evaluation forums. This paper gives an overview of all the INEX 2014 tracks, their aims and task, the built test-collections, the participants, and gives an initial analysis of the results.

The INEX QA track aimed to evaluate complex questionanswering tasks where answers are short texts generated from the Wikipedia by extraction of relevant short passages and aggregation into a coherent summary. In such a task, Question-answering, XML/passage retrieval and automatic summarization are combined in order to get closer to real information needs. Based on the groundwork carried out in 2009-2010 edition to determine the sub-tasks and a novel evaluation methodology, the 2011 edition experimented contextualizing tweets using a recent cleaned dump of the Wikipedia. Participants had to contextualize 132 tweets from the New York Times (NYT). Informativeness of answers has been evaluated, as well as their readability. 13 teams from 6 countries actively participated to this track. This tweet contextualization task will continue in 2012 as part of the CLEF INEX lab with same methodology and baseline but on a much wider range of tweet types.

Microblogging platforms such as Twitter are increasingly used for on-line client and market analysis. This motivated the proposal of a new track at CLEF INEX lab of Tweet Contextualization. The objective of this task was to help a user to understand a tweet by providing him with a short explanatory summary (500 words). This summary should be built automatically using resources like Wikipedia and generated by extracting relevant passages and aggregating them into a coherent summary. Running for four years, results show that the best systems combine NLP techniques with more traditional methods. More precisely the best performing systems combine passage retrieval, sentence segmentation and scoring, named entity recognition, text part-of-speech (POS) analysis, anaphora detection, diversity content measure as well as sentence reordering. This paper provides a full summary report on the four-year long task. While yearly overviews focused on system results, in this paper we provide a detailed report on the approaches proposed by the participants and which can be considered as the state of the art for this task. As an important result from the 4 years competition, we also describe the open access resources that have been built and collected. The evaluation measures for automatic summarization designed in DUC or MUC were not appropriate to evaluate tweet contextualization, we explain why and depict in detailed the LogSim measure used to evaluate informativeness of produced contexts or summaries. Finally, we also mention the lessons we learned and that it is worth considering when designing a task.