Semantically speaking

51 views Leave a comment

A psychology highbrow and an engineering highbrow accommodate during a party.

It sounds like a commencement of a joke, though during this sold probability encounter, a dual professors detected they have identical interests in language, persuasion, and a origination of meaning.

“You don’t mostly accommodate people during parties who are doing textual analysis,” says Chris Westbury, a highbrow in a Faculty of Science’s Department of Psychology who specializes in denunciation and semantics.

“Most people run divided when we start articulate about co-occurrence modelling,” echoes Lianne Lefsrud, a triple U of A alumna (in environmental engineering and vital supervision and organization) and new partner highbrow in a Faculty of Engineering.

Now Westbury and Lefsrud are operative together on their common investigate interests to beam a Alberta Chamber of Resources (ACR) in improved bargain open conversations by examining how a use and multiple of certain difference affects notice of meaning. (Think “oilsands” contra “tarsands” and a connotations compared with any term.) The ACR—representing 174 member companies from a mining, appetite and forestry sectors—is focused on Alberta’s “triple future”: not usually economics, though also environmental and amicable impacts—the latter dual of that Westbury and Lefsrud see as closely tied to their research.

Listening in on a tellurian conversation

“How can we listen to vast open conversations?” asks Lefsrud. “You can reason a concentration group, though we usually listen to a people in that concentration group. Or we can do a survey, though we usually hear from a people you’re surveying. Conversations about Alberta’s apparatus growth have stretched over niche conversations in Alberta to Toronto, Washington, DC, London, and worldwide. How can we know what is romantic tongue contra ubiquitous open sentiment?”

To that end, Westbury and Lefsrud teamed adult this summer with their co-worker Geoff Hollis and dual psychology students—Janel Comeau and Jeff Keith—to investigate scarcely half a million singular journal articles from 2004 to present, regulating custom-written software. The organisation is also formulation to inspect a use of Facebook and Twitter by looking during how word frequencies change over time on amicable media.

“We are examining how mostly difference seem in a same essay and how this changes opposite time,” says Westbury, adding that a organisation is focusing on 200 keywords subsequent in conference with attention experts. “For example, how mostly when we see a word ‘oil’ do we also see a word ‘dirty’?”

‘Oilsands’ and a spillover effect

“Oil and connect were deliberate detailed synonyms until a 1970s and ‘80s,” says Lefsrud. “As it became some-more valuable, companies and a supervision really consciously started job them ‘oilsands’ rather than ‘tarsands.’ When connect was reintroduced in 2006–2007, ‘tarsands’ took on a newly irreverent meaning. It was unwashed and gummy and black and death.

“Our formula uncover that ‘tarsands’ is most some-more emotionally disastrous than a some-more neutral ‘oilsands,’” says Lefsrud. “Our investigate demonstrates that these terms started out sincerely tighten on what we call a ‘negativity index,’ a magnitude of disastrous romantic sentiment. Then they split. The engaging thing is that it looks like there’s a spillover outcome within about an eight-month window. As ‘tarsands’ acquires disastrous meaning, a negativity spills over, and ‘oilsands’ becomes some-more negative.”

Predicting how a review will change

Lefsrud explains that a idea is to yield collection for companies to improved know open conversation. “The universe has changed,” she says. “It’s no longer exclusively an economics discussion, so it’s not usually changing how these organizations speak about what they’re doing, though also how they emanate their vital priorities. This is not a PR campaign. It’s some-more elemental than that.”

Westbury and Lefsrud see a probability of their investigate being practical over apparatus growth to other argumentative issues—energy, meridian change, and branch dungeon research, for example. “We can lane argumentative issues to see how and where they’ve altered by time,” says Lefsrud.

They also wish to take their investigate one step serve to envision stirring trends in open notice formed on denunciation being used. “Our finish idea is to try to be means to envision in allege what changes competence be about to start or to find signs that will tell us that things are about to change,” says Westbury. “Can we envision that people are going to start vocalization negatively about a sold topic?

“We’ve tangible a lot of measures that tell us about a characteristics of text, how things change opposite time and within opposite seductiveness groups. We have ways of characterizing vast blocks of content now. This is not simply automating media monitoring. It characterizes a conversations in a suggestive approach to promote deeper understanding.”

To arrive during that understanding, a researchers use a computational indication that characterizes a romantic tinge of a word from examining a context of use (co-occurrence modelling), rather like how we competence come to know a definition of an unknown word from regularly conference how it is used. They hunt by this and their other measures to impersonate changes opposite time in a open review about Alberta’s resources.

Source: University of Alberta