The strength with which a statement is made can have a significant impact on the audience. For example, international relations can be strained by how the media in one country describes an event in another; and papers can be rejected because they overstate or understate their findings. It is thus important to understand the effects of statement strength. A first step is to be able to distinguish between strong and weak statements. However, even this problem is understudied, partly due to a lack of data. Since strength is inherently relative, revisions of texts that make claims are a natural source of data on strength differences. In this paper, we introduce a corpus of sentence-level revisions from academic writing. We also describe insights gained from our annotation efforts for this task.
Consider a person trying to spread an important message on a social network. He/she can spend hours trying to craft the message. Does it actually matter? While there has been extensive prior work looking into predicting popularity of social-media content, the effect of wording per se has rarely been studied since it is often confounded with the popularity of the author and the topic. To control for these confounding factors, we take advantage of the surprising fact that there are many pairs of tweets containing the same url and written by the same user but employing different wording. Given such pairs, we ask: which version attracts more retweets? This turns out to be a more difficult task than predicting popular topics. Still, humans can answer this question better than chance (but far from perfectly), and the computational methods we develop can do better than both an average human and a strong competing method trained on non-controlled data.
We investigate whether one can determine from the transcripts of U.S. Congressional floor debates whether the speeches represent support of or opposition to proposed legislation. To address this problem, we exploit the fact that these speeches occur as part of a discussion; this allows us to use sources of information regarding relationships between discourse segments, such as whether a given utterance indicates agreement with the opinion expressed by another. We find that the incorporation of such information yields substantial improvements over classifying speeches in isolation.
Understanding the ways in which participants in public discussions frame their arguments is important in understanding how public opinion is formed. In this paper, we adopt the position that it is time for more computationally-oriented research on problems involving framing. In the interests of furthering that goal, we propose the following specific, interesting and, we believe, relatively accessible question: In the controversy regarding the use of genetically-modified organisms (GMOs) in agriculture, do pro- and anti-GMO articles differ in whether they choose to adopt a "scientific" tone? Prior work on the rhetoric and sociology of science suggests that hedging may distinguish popular-science text from text written by professional scientists for their colleagues. We propose a detailed approach to studying whether hedge detection can be used to understanding scientific framing in the GMO debates, and provide corpora to facilitate this study. Some of our preliminary analyses suggest that hedges occur less frequently in scientific discourse than in popular text, a finding that contradicts prior assertions in the literature. We hope that our initial work and data will encourage others to pursue this promising line of inquiry.
Understanding the ways in which information achieves widespread public awareness is a research question of significant interest. We consider whether, and how, the way in which the information is phrased --- the choice of words and sentence structure --- can affect this process. To this end, we develop an analysis framework and build a corpus of movie quotes, annotated with memorability information, in which we are able to control for both the speaker and the setting of the quotes. We find that there are significant differences between memorable and non-memorable quotes in several key dimensions, even after controlling for situational and contextual factors. One is lexical distinctiveness: in aggregate, memorable quotes use less common word choices, but at the same time are built upon a scaffolding of common syntactic patterns. Another is that memorable quotes tend to be more general in ways that make them easy to apply in new contexts --- that is, more portable. We also show how the concept of "memorable language" can be extended across domains.
Understanding social interaction within groups is key to analyzing online communities. Most current work focuses on structural properties: who talks to whom, and how such interactions form larger network structures. The interactions themselves, however, generally take place in the form of natural language --- either spoken or written --- and one could reasonably suppose that signals manifested in language might also provide information about roles, status, and other aspects of the group's dynamics. To date, however, finding such domain-independent language-based signals has been a challenge. Here, we show that in group discussions power differentials between participants are subtly revealed by how much one individual immediately echoes the linguistic style of the person they are responding to. Starting from this observation, we propose an analysis framework based on linguistic coordination that can be used to shed light on power relationships and that works consistently across multiple types of power --- including a more "static" form of power based on status differences, and a more "situational" form of power in which one individual experiences a type of dependence on another. Using this framework, we study how conversational behavior can reveal power relationships in two very different settings: discussions among Wikipedians and arguments before the U.S. Supreme Court.
We show that information about social relationships can be used to improve user-level sentiment analysis. The main motivation behind our approach is that users that are somehow "connected" may be more likely to hold similar opinions; therefore, relationship information can complement what we can extract about a user's viewpoints from their utterances. Employing Twitter as a source for our experimental data, and working within a semi-supervised framework, we propose models that are induced either from the Twitter follower/followee network or from the network in Twitter formed by users referring to each other using "@" mentions. Our transductive learning results reveal that incorporating social-network information can indeed lead to statistically significant sentiment-classification improvements over the performance of an approach based on Support Vector Machines having access only to textual features.
Conversational participants tend to immediately and unconsciously adapt to each other's language styles: a speaker will even adjust the number of articles and other function words in their next utterance in response to the number in their partner's immediately preceding utterance. This striking level of coordination is thought to have arisen as a way to achieve social goals, such as gaining approval or emphasizing difference in status. But has the adaptation mechanism become so deeply embedded in the language-generation process as to become a reflex? We argue that fictional dialogs offer a way to study this question, since authors create the conversations but don't receive the social benefits (rather, the imagined characters do). Indeed, we find significant coordination across many families of function words in our large movie-script corpus. We also report suggestive preliminary findings on the effects of gender and other features; e.g., surprisingly, for articles, on average, characters adapt more to females than to males.
Researchers in textual entailment have begun to consider inferences involving 'downward-entailing operators', an interesting and important class of lexical items that change the way inferences are made. Recent work proposed a method for learning English downward-entailing operators that requires access to a high-quality collection of 'negative polarity items' (NPIs). However, English is one of the very few languages for which such a list exists. We propose the first approach that can be applied to the many languages for which there is no pre-existing high-precision database of NPIs. As a case study, we apply our method to Romanian and show that our method yields good results. Also, we perform a cross-linguistic analysis that suggests interesting connections to some findings in linguistic typology.
We report on work in progress on extracting lexical simplifications (e.g., "collaborate" -> "work together"), focusing on utilizing edit histories in Simple English Wikipedia for this task. We consider two main approaches: (1) deriving simplification probabilities via an edit model that accounts for a mixture of different operations, and (2) using metadata to focus on edits that are more likely to be simplification operations. We find our methods to outperform a reasonable baseline and yield many high-quality lexical simplifications not included in an independently-created manually prepared list.