Eyal Sagi

Using analogical mapping to simulate time-course phenomena in perceptual similarity

We present a computational model of visual similarity. The model is based upon the idea that perceptual comparisons may utilize the same mapping processes as are used in analogy. We use the Structure Mapping Engine (SME), a model of Gentners structure-mapping theory of analogy, to perform comparison on representations that are automatically generated from visual input. By encoding visual scenes incrementally and sampling the output of SME at multiple stages in its processing, we are able to model not only the output of similarity judgments, but the time course of the comparison process. We demonstrate the models effectiveness by replicating the results from three psychological studies that bear on the time course of comparison.

Purity homophily in social networks.

Does sharing moral values encourage people to connect and form communities? The importance of moral homophily (love of same) has been recognized by social scientists, but the types of moral similarities that drive this phenomenon are still unknown. Using both large-scale, observational social-media analyses and behavioral lab experiments, the authors investigated which types of moral similarities influence tie formations. Analysis of a corpus of over 700,000 tweets revealed that the distance between 2 people in a social-network can be predicted based on differences in the moral purity content-but not other moral content-of their messages. The authors replicated this finding by experimentally manipulating perceived moral difference (Study 2) and similarity (Study 3) in the lab and demonstrating that purity differences play a significant role in social distancing. These results indicate that social network processes reflect moral selection, and both online and offline differences in moral purity concerns are particularly predictive of social distance. This research is an attempt to study morality indirectly using an observational big-data study complemented with 2 confirmatory behavioral experiments carried out using traditional social-psychology methodology.

Measuring Moral Rhetoric in Text

In this paper we present a computational text analysis technique for measuring the moral loading of concepts as they are used in a corpus. This method is especially useful for the study of online corpora as it allows for the rapid analysis of moral rhetoric in texts such as blogs and tweets as events unfold. We use latent semantic analysis to compute the semantic similarity between concepts and moral keywords taken from the “Moral foundation Dictionary”. This measure of semantic similarity represents the loading of these concepts on the five moral dimensions identified by moral foundation theory. We demonstrate the efficacy of this method using three different concepts and corpora.

Identifying Issue Frames in Text

Framing, the effect of context on cognitive processes, is a prominent topic of research in psychology and public opinion research. Research on framing has traditionally relied on controlled experiments and manually annotated document collections. In this paper we present a method that allows for quantifying the relative strengths of competing linguistic frames based on corpus analysis. This method requires little human intervention and can therefore be efficiently applied to large bodies of text. We demonstrate its effectiveness by tracking changes in the framing of terror over time and comparing the framing of abortion by Democrats and Republicans in the U.S.

Identity, Causality, and Pronoun Ambiguity

This article looks at the way people determine the antecedent of a pronoun in sentence pairs, such as: Albert invited Ron to dinner. He spent hours cleaning the house. The experiment reported here is motivated by the idea that such judgments depend on reasoning about identity (e.g., the identity of the he who cleaned the house). Because the identity of an individual over time depends on the causal-historical path connecting the stages of the individual, the correct antecedent will also depend on causal connections. The experiment varied how likely it is that the event of the first sentence (e.g., the invitation) would cause the event of the second (the house cleaning) for each of the two individuals (the likelihood that if Albert invited Ron to dinner, this would cause Albert to clean the house, versus cause Ron to clean the house). Decisions about the antecedent followed causal likelihood. A mathematical model of causal identity accounted for most of the key aspects of the data from the individual sentence pairs.

Language Use and Coalition Formation in Multiparty Negotiations

The alignment of bargaining positions is crucial to a successful negotiation. Prior research has shown that similarity in language use is indicative of the conceptual alignment of interlocutors. We use latent semantic analysis to explore how the similarity of language use between negotiating parties develops over the course of a three-party negotiation. Results show that parties that reach an agreement show a gradual increase in language similarity over the course of the negotiation. Furthermore, reaching the most financially efficient outcome is dependent on similarity in language use between the parties that have the most to gain from such an outcome.

The role of speaker beliefs in determining accent placement

In English and other languages, the distribution of nuclear pitch accents within a sentence usually reflects how the meaningful parts of the sentence relate to the context. Generally speaking, the nuclear pitch accent can only occur felicitously on focused parts of the sentence, corresponding to information that is not contextually retrievable or given. In most contemporary theories, focus is formally represented by an abstract syntactic feature ‘F’. Those parts of the sentence that are given tend to resist F-marking and thus nuclear accentuation. In short, there is a more or less tight coupling between (i) the contextual information status of parts of the sentence; (ii) the focus structure of the sentence (represented by the distribution of syntactic F-marking); and (iii) the actual accent placement in the phonological form.

Developing a New Method for Psychological Investigation Using Text as Data

Over the past few decades, the study of psychology has been undergoing a methodological transformation. The increasing availability and quantity of real-world big data have prompted researchers to look for new ways to use these data to gain psychological insights. In this case study, I trace the process of developing a new method for testing psychological theories using corpora. Instead of bringing participants to the lab, I analyze statistical patterns of cooccurrence in naturally occurring texts obtained from a variety of sources, including political speeches, literature, and the Internet. The basic assumption I make is that these patterns reflect the representations and cognitive processes of their author. I present three different applications of the method and use them to describe how such data can be analyzed and used to answer a range of questions in psychology and social science. The first application examines the linguistic question of the relationship between word form and meaning. The second application identifies cognitive frames as they are found in text. The final application uses texts to measure the style of moral reasoning individuals apply in particular contexts. This case study provides insight into the process of developing new methodologies for hypothesis testing, as well as demonstrating how to formulate hypotheses that can be tested using corpora. In addition, several key pitfalls in the process of adapting statistical methods to new uses are identified and discussed.

Correction: Identifying Issue Frames in Text

Several errors occurred in the DOI (Digital Object Identifier) links for the manuscripts listed in the Reference section. Please see an updated Reference list with the correct links below: