Topic Correlations over Time
نویسندگان
چکیده
Topic models have proved useful for analyzing large clusters of documents. Most models developed, however, have paid little attention to the analysis of the latent topics themselves, particularly with regards to change in their correlation over time. We present a novel, probabilistically well-founded extension to Latent Dirichlet Allocation (LDA) which can explicitly model topic drift over time. Using this extension, we analyze the correlations of topics over time in a corpus of ACL papers.
منابع مشابه
A Continuous-Time Model of Topic Co-occurrence Trends
Recent work in statistical topic models has investigated richer structures to capture either temporal or inter-topic correlations. This paper introduces a topic model that combines the advantages of two recently proposed models: (1) The Pachinko Allocation model (PAM), which captures arbitrary topic correlations with a directed acyclic graph (DAG), and (2) the Topics over Time model (TOT), whic...
متن کاملTopic Modeling of Research Fields: An Interdisciplinary Perspective
This paper addresses the problem of scientific research analysis. We present various novel topic models which classify research papers based on topic and language. Moreover, we show various insightful statistics and correlations within and across three research fields: Linguistics, Computational Linguistics, and Education. In particular, we show how topics change over time within each field, wh...
متن کاملExploiting Temporal Authors Interests via Temporal-Author-Topic Modeling
This paper addresses the problem of discovering temporal authors interests. Traditionally some approaches used stylistic features or graph connectivity and ignored semantics-based intrinsic structure of words present between documents, while previous topic modeling approaches considered semantics without time factor, which is against the spirits of writing. We present Temporal-Author-Topic (TAT...
متن کاملCorrelations in Economic Time Series
The correlation function of a financial index of the New York stock exchange, the S&P 500, is analyzed at 1min intervals over the 13-year period, Jan 84 – Dec 96. We quantify the correlations of the absolute values of the index increment. We find that these correlations can be described by two different power laws with a crossover time t× ≈ 600min. Detrended fluctuation analysis gives exponents...
متن کاملScientific Research Analysis Across Multiple Fields
This paper addresses the problem of scientific research analysis across multiple fields. We use Latent Dirichlet Allocation to model research topics in papers and show various insightful statistics and correlations within and across five research fields: Linguistics, Computational Linguistics, Psychology, Education, and Marketing. In particular, we introduce three new measures of coherence, imp...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2007