{"pk":28901,"title":"Using Big Data to Understand Memory and Future Thinking","subtitle":null,"abstract":"Imagining the future and remembering the past both involve\nmental time travel. This commonality could indicate shared\nmental processes, as held by the Constructive Episodic\nSimulation Hypothesis (Schacter & Addis, 2008), or else\ninteractive processes that complement one another, a\npossibility we call the Complementarity Hypothesis.\nAccording to the Complementarity Hypothesis, future thoughts\nare constructed from schemas making them episodically poor,\nwhereas past thoughts are constructed from schemas and direct\nretrieval of memory traces, making them relatively\nepisodically rich. We tested these hypotheses using machine\nlearning to data mine mental operations in language, much as\na geologist can recover physical processes from the geological\nrecord. People’s natural, unprompted talk on web blogs was\nautomatically analyzed for past, present, and future references\nusing a temporal orientation classifier. In Study 1, we found\nthat perceptual details were mentioned more often in past than\nfuture talk, implying greater use of episodic processing in past\nthan future thinking. In Study 2, a neural network using\nschemas generated from Latent Dirichlet Allocation better\npredicted the content of references to the future than the past,\nimplying that constructive processes are more common in\nfuture than past thinking. In Study 3, we used the results from\nthe two prior studies to construct an episodic-by-constructive\nprocess space. We adapted techniques from fMRI analysis to\nanalyze this space for clusters of activity, as if the frequency of\npast and future thinking were BOLD responses in cortical\nspace. We found that past and future thinking occupy highly\nseparable regions of processing space, supporting the\nComplementarity Hypothesis.","language":"eng","license":{"name":"","short_name":"","text":null,"url":""},"keywords":[{"word":"Prospection; Memory; Future Thinking; Big\nData; Naturally Occurring Datasets"}],"section":"Papers with Poster Presentations","is_remote":true,"remote_url":"https://escholarship.org/uc/item/7fq047v4","frozenauthors":[{"first_name":"Robert","middle_name":"","last_name":"Thorstad","name_suffix":"","institution":"Emory University","department":""},{"first_name":"Phillip","middle_name":"","last_name":"Wolff","name_suffix":"","institution":"Emory University","department":""}],"date_submitted":null,"date_accepted":null,"date_published":"2019-01-01T13:00:00-05:00","render_galley":null,"galleys":[{"label":"PDF","type":"pdf","path":"https://journalpub.escholarship.org/cognitivesciencesociety/article/28901/galley/18772/download/"}]}