Syntactic Topic Models
نویسندگان
چکیده
We develop the syntactic topic model (STM), a nonparametric Bayesian model of parsed documents. The STM generates words that are both thematically and syntactically constrained, which combines the semantic insights of topic models with the syntactic information available from parse trees. Each word of a sentence is generated by a distribution that combines document-specific topic weights and parse-tree-specific syntactic transitions. Words are assumed to be generated in an order that respects the parse tree. We derive an approximate posterior inference method based on variational methods for hierarchical Dirichlet processes, and we report qualitative and quantitative results on both synthetic data and hand-parsed documents.
منابع مشابه
Syntactic Structures and Rhetorical Functions of Electrical Engineering, Psychiatry, and Linguistics Research Article Titles in English and Persian: A Cross-linguistic and Cross-disciplinary Study
A research article (RA) title is the first and foremost feature that attracts the reader's attention, the feature from which she/he may decide whether the whole article is worth reading. The present study attempted to investigate syntactic structures and rhetorical functions of RA titles written in English and Persian and published in journals in three disciplines of Electrical Engineering, Psy...
متن کاملSyntactic Topic Models for Language Generation
Since topic models’ inception as probabilistic generative models, it has only been natural to imagine actually applying the generative process to create documents. However, most topic models consist of a generative process that only provides a bag of words which is one critical step short of creating a readable text. With the recent introduction of syntactically sound topic models and structure...
متن کامل“IS That What You Mean?” Exploratory Study of Syntactic Pattern in Complement Responses
Complimenting behavior, as a common speech act of human beings, has become an intriguing topic in linguistics and its sub-branches. Compliment responses can be seen as solutions for maintaining a balance between (1) a preference to avoid self-praise and (2) a preference to accept or agree with the compliment (Pomerantz 1978). In the present study, the definition of a compliment draws on the wor...
متن کاملMaximum Entropy Language Modeling with Non-Local and Syntactic Dependencies
Standard N -gram language models exploit information only from the immediate past to predict the future word. To improve the performance of a language model, two di erent kinds of long-range dependence, the syntactic structure and the topic of sentences are taken into consideration. The likelihood of many words varies greatly with the topic of discussion and topics capture this di erence. Synta...
متن کاملThe Impact of Recasts on the Syntactic Accuracy of Iranian EFL University Students’ Oral Discourse
Among the major issues raised by classroom SLA researchers is the debate on the degree to which teacher’s or learner’s attention should be directed to linguistic features. However, one of the relevant variables in corrective feedback studies which seem to be less operationalized is the differential impact of different types of feedback on the accuracy of the oral performance of the participants...
متن کامل