Show simple item record

dc.contributor.author Ehrenfeld, Steven Emil en
dc.date.accessioned 2011-05-27T20:24:09Z en
dc.date.available 2011-05-27T20:24:09Z en
dc.date.issued 5/27/11 en
dc.identifier.uri http://hdl.handle.net/10211.10/1073 en
dc.description.abstract Analysis of language posted publicly on the internet has given us a new way of observing consumers, and natural language processing methods allow us to analyze this large amount of text as it is being posted online. In this study we build and analyze corpora of internet message board discussions and use this analysis to build a model that attempts to predict videogame sales figures. Weekly corpora are built by downloading and processing text consisting of the discussions of a large community focused on the topic of videogames. This text is then analyzed to determine which videogame titles generate the most discussion within the community for each week. We use support vector regression to create a model that is able to make predictions about future sales figures. Similar methods have been successful in predicting success in other areas such as box office ticket sales for movies and book sales. By tracking and analyzing consumer interest within online communities, these methods can provide a number of industries with the ability to predict the success of products. en
dc.publisher Arts and Letters en
dc.title Predicting video game sales using an analysis of internet message board discussions en
dc.type Thesis en
dc.contributor.department Linguistics and Asian/Middle Eastern Languages en
dc.description.degree Master of Arts (M.A.) San Diego State University, 2011 en
dc.description.discipline Linguistics en


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record

Search DSpace


My Account

RSS Feeds