HN user
One aspect I didn't see mentioned in the article was whether brand name is reflected in final purchase v.s. initial search.
As a consumer, I often am open to one of several common brands but will search generically in order to see all available options. However, I'm unlikely to buy from an extremely generic brand in most cases.
Not sure if this behavior is common, but I expect it is not uncommon.
How are the data points assigned to their locations in the geometric space? I didn't notice any mention of that. What makes something "more like" a thriller and "less like" a comedy, for example? We can all tell the intuitive difference, but what markers serve best to precisely assign a mathematical distance?
I know very little about this problem space, so these are actual questions, not at all rhetorical.
Is the code for that project on GitHub / would you be willing to share the source? I'd be interested in seeing that take on a blog implementation.
I'm curious about your semantic analysis process. Do you employ a particular threshold for neutrality, or does the team provide input?
No doubt neutral reporting is advantageous; personally I would find it helpful to also see both a highly positive and negative take. Coming at a topic from both high and low may, imho, improve accuracy of comprehension.