Andrés Montoyo

UA-ZBSA: A Headline Emotion Classification through Web Information

This paper presents a headline emotion classification approach based on frequency and co-occurrence information collected from the World Wide Web. The content words of a headline (nouns, verbs, adverbs and adjectives) are extracted in order to form different bag of word pairs with the joy, disgust, fear, anger, sadness and surprise emotions. For each pair, we compute the Mutual Information Score which is obtained from the web occurrences of an emotion and the content words. Our approach is based on the hypothesis that group of words which co-occur together across many documents with a given emotion are highly probable to express the same emotion.

Combining knowledge- and corpus-based word-sense-disambiguation methods

In this paper we concentrate on the resolution of the lexical ambiguity that arises when a given word has several different meanings. This specific task is commonly referred to as word sense disambiguation (WSD). The task of WSD consists of assigning the correct sense to words using an electronic dictionary as the source of word definitions. We present two WSD methods based on two main methodological approaches in this research area: a knowledge-based method and a corpus-based method. Our hypothesis is that word-sense disambiguation requires several knowledge sources in order to solve the semantic ambiguity of the words. These sources can be of different kinds-- for example, syntagmatic, paradigmatic or statistical information. Our approach combines various sources of knowledge, through combinations of the two WSD methods mentioned above. Mainly, the paper concentrates on how to combine these methods and sources of information in order to achieve good results in the disambiguation. Finally, this paper presents a comprehensive study and experimental work on evaluation of the methods and their combinations.

Detecting implicit expressions of emotion in text: A comparative analysis

Sentiment analysis is one of the recent, highly dynamic fields in Natural Language Processing. Most existing approaches are based on word-level analysis of texts and are mostly able to detect only explicit expressions of sentiment. However, in many cases, emotions are not expressed by using words with an affective meaning (e.g. happy), but by describing real-life situations, which readers (based on their commonsense knowledge) detect as being related to a specific emotion. Given the challenges of detecting emotions from contexts in which no lexical clue is present, in this article we present a comparative analysis between the performance of well-established methods for emotion detection (supervised and lexical knowledge-based) and a method we propose and extend, which is based on commonsense knowledge stored in the EmotiNet knowledge base. Our extensive evaluations show that, in the context of this task, the approach based on EmotiNet is the most appropriate.

Paraphrase identification on the basis of supervised machine learning techniques

This paper presents a machine learning approach for paraphrase identification which uses lexical and semantic similarity information. In the experimental studies, we examine the limitations of the designed attributes and the behavior of three machine learning classifiers. With the objective to increase the final performance of the system, we scrutinize the influence of the combination of lexical and semantic information, as well as techniques for classifier combination.

Building and Exploiting EmotiNet, a Knowledge Base for Emotion Detection Based on the Appraisal Theory Model

The task of automatically detecting emotion in text is challenging. This is due to the fact that most of the times, textual expressions of affect are not direct-using emotion words-but result from the interpretation and assessment of the meaning of the concepts and interaction of concepts described in the text. This paper presents the core of EmotiNet, a new knowledge base (KB) for representing and storing affective reaction to real-life contexts, and the methodology employed in designing, populating, and evaluating it. The basis of the design process is given by a set of self-reported affective situations in the International Survey on Emotion Antecedents and Reactions (ISEAR) corpus. We cluster the examples and extract triples using Semantic Roles. We subsequently extend our model using other resources, such as VerbOcean, ConceptNet, and SentiWordNet, with the aim of generalizing the knowledge contained. Finally, we evaluate the approach using the representations of other examples in the ISEAR corpus. We conclude that EmotiNet, although limited by the domain and small quantity of knowledge it presently contains, represents a semantic resource appropriate for capturing and storing the structure and the semantics of real events and predicting the emotional responses triggered by chains of actions.

Preface: Computational approaches to subjectivity and sentiment analysis: Present and envisaged methods and applications

Recent years have witnessed a surge of interest in computational methods for affect, ranging from opinion mining, to subjectivity detection, to sentiment and emotion analysis. This article presents a brief overview of the latest trends in the field and describes the manner in which the articles contained in the special issue contribute to the advancement of the area. Finally, we comment on the current challenges and envisaged developments of the subjectivity and sentiment analysis fields, as well as their application to other Natural Language Processing tasks and related domains.

Combining data-driven systems for improving Named Entity Recognition

The increasing flow of digital information requires the extraction, filtering and classification of pertinent information from large volumes of texts. All these tasks greatly benefit from involving a Named Entity Recognizer (NER) in the preprocessing stage. This paper proposes a completely automatic NER system. The NER task involves not only the identification of proper names (Named Entities) in natural language text, but also their classification into a set of predefined categories, such as names of persons, organizations (companies, government organizations, committees, etc.), locations (cities, countries, rivers, etc.) and miscellaneous (movie titles, sport events, etc.). Throughout the paper, we examine the differences between language models learned by different data-driven classifiers confronted with the same NLP task, as well as ways to exploit these differences to yield a higher accuracy than the best individual classifier. Three machine learning classifiers (Hidden Markov Model, Maximum Entropy and Memory Based Learning) are trained on the same corpus in order to resolve the NE task. After comparison, their output is combined using voting strategies. A comprehensive study and experimental work on the evaluation of our system, as well as a comparison with other systems has been carried out within the framework of two specialized scientific competitions for NER, CoNLL-2002 and HAREM-2005. Finally, this paper describes the integration of our NER system in different NLP applications, in concrete Geographic Information Retrieval and Conceptual Modelling.

A feature dependent method for opinion mining and classification

Mining the web for customer opinion on different products is both a useful, as well as challenging task. Previous approaches to customer review classification included document level, sentence and clause level sentiment analysis and feature based opinion summarization. In this paper, we present a feature driven opinion summarization method, where the term ldquodrivenrdquo is employed to describe the concept-to-detail (product class to product-specific characteristics) approach we took. For each product class we first automatically extract general features (characteristics describing any product, such as price, size, design), for each product we then extract specific features (as picture resolution in the case of a digital camera) and feature attributes (adjectives grading the characteristics, as for example high or low for price, small or big for size and modern or faddy for design). Further on, we assign a polarity (positive or negative) to each of the feature attributes using a previously annotated corpus and Support Vector Machines Sequential Minimal Optimization machine learning with the Normalized Google Distance. We show how the method presented is employed to build a feature-driven opinion summarization system that is presently working in English and Spanish. In order to detect the product category, we use a modified system for person names classification. The raw review text is split into sentences and depending on the product class detected, only the phrases containing the specific product features are selected for further processing. The phrases extracted undergo a process of anaphora resolution, Named Entity Recognition and syntactic parsing. Applying syntactic dependency and part of speech patterns, we extract pairs containing the feature and the polarity of the feature attribute the customer associates to the feature in the review. Eventually, we statistically summarize the polarity of the opinions different customers expressed about the product on the web as percentages of positive and negative opinions about each of the product features. We show the results and improvements over baseline, together with a discussion on the strong and weak points of the method and the directions for future work.

Determining the Polarity and Source of Opinions Expressed in Political Debates

In this paper we investigate different approaches we developed in order to classify opinion and discover opinion sources from text, using affect, opinion and attitude lexicon. We apply these approaches on the discussion topics contained in a corpus of American Congressional speech data. We propose three approaches to classifying opinion at the speech segment level, firstly using similarity measures to the affect, opinion and attitude lexicon, secondly dependency analysis and thirdly SVM machine learning. Further, we study the impact of taking into consideration the source of opinion and the consistency in the opinion expressed, and propose three methods to classify opinion at the speaker intervention level, showing improvements over the classification of individual text segments. Finally, we propose a method to identify the party the opinion belongs to, through the identification of specific affective and non-affective lexicon used in the argumentations. We present the results obtained when evaluating the different methods we developed, together with a discussion on the issues encountered and some possible solutions. We conclude that, even at a more general level, our approach performs better than trained classifiers on specific data.

Towards Building a Competitive Opinion Summarization System: Challenges and Keys

This paper presents an overview of our participation in the TAC 2008 Opinion Pilot Summarization task, as well as the proposed and evaluated post-competition improvements. We first describe our opinion summarization system and the results obtained. Further on, we identify the systems weak points and suggest several improvements, focused both on information content, as well as linguistic and readability aspects. We obtain encouraging results, especially as far as F-measure is concerned, outperforming the competition results by approximately 80%.