TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Show HN: Visualize how HN/Reddit talk about your company and products

74 pointsby jonathanbgnover 5 years ago

15 comments

kenover 5 years ago
I was at a company that tried &quot;sentiment analysis&quot; 10 or 15 years ago. It was impossible to get right. The results were useless. It&#x27;s funny how nothing seems to have changed. There&#x27;s always some executive who&#x27;s heard of &quot;sentiment&quot; and thinks they can spend a week shoving a bunch of text into a database, and then have it tell you whether the words are perceived as good or bad.<p>- I typed &quot;Chevrolet&quot; and it says &quot;mpg&quot; and &quot;venerable&quot; have negative sentiment, while &quot;underachieving&quot;, &quot;supposedly&quot;, &quot;kilometers&quot;, and &quot;cars&quot; have positive sentiment. &quot;Stunning&quot; and &quot;awkward&quot; are both neutral.<p>- I tried &quot;Mazda&quot; and it was only slightly better. &quot;Praise&quot; is neutral, &quot;regret&quot; is positive, and &quot;touring&quot; is negative.<p>- I tried &quot;Petzl&quot; and only shows 16 words, all neutral, and that includes stop words like &quot;had&quot; and &quot;and&quot; and &quot;etc&quot;.<p>Those are at least companies with unique names. For companies with less common names, you might be lucky to get keywords related to the right company at all.<p>I think there can be good uses for word clouds, but they are few and far between, and this isn&#x27;t one. Just make 3 lists, side by side, and title them &quot;positive&quot;, &quot;neutral&quot;, and &quot;negative&quot;. Instead of font size, put the more common words higher on their respective list. The only reason I can see to use a word cloud here is to hide how bad the analysis is.
评论 #21580893 未加载
swalshover 5 years ago
This is a really cool idea, there are other products out there that do the same thing, but there is definitely room to differentiate yourself.<p>One interesting issue, you might need to read the context a bit more to gauge if the post is actually talking about the Brand. For example Target is a common english word, and if you look at the results, only 1 usage was actually referring to the store Target.<p>One nice bonus that would be useful, I&#x27;d like to be able to compare sentiment by Subreddit, if i&#x27;m marketing on Reddit I would not target &quot;Reddit&quot; as a whole.
评论 #21574998 未加载
moralestapiaover 5 years ago
Good idea in principle, however, did you even try to run a test before releasing it to the public?<p>For a particular term I get the following words having &quot;positive&quot; sentiments: boycott, punishment, greed, backlash, protesters, damaging, upset.<p>How in hell is that positive?
评论 #21577695 未加载
评论 #21576885 未加载
评论 #21577165 未加载
vecplaneover 5 years ago
Getting an error on <a href="https:&#x2F;&#x2F;brandimage.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;brandimage.io&#x2F;</a><p>Over Quota<p>This application is temporarily over its serving quota. Please try again later.
评论 #21577576 未加载
lganzzzoover 5 years ago
Cool stuff!<p>I found an issue with &quot;+&quot; symbols in the brand name. Check out this URL - <a href="https:&#x2F;&#x2F;brandimage.io&#x2F;insight&#x2F;c++?source=hn" rel="nofollow">https:&#x2F;&#x2F;brandimage.io&#x2F;insight&#x2F;c++?source=hn</a><p>If &quot;+&quot; symbol is present, the UI won&#x27;t show anything.
kmcquadeover 5 years ago
Pretty awesome that it’s open source.<p>For large companies there are some proprietary solutions for this. Example:<p><a href="https:&#x2F;&#x2F;www.trendkite.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.trendkite.com&#x2F;</a><p>Disclaimer: family member works there so that’s the reason I’m aware that this niche exists.
criddellover 5 years ago
I kind of hate this idea.<p>It&#x27;s one thing if companies want to find out what people are saying about them so they can improve. However, I&#x27;d be surprised if that&#x27;s how it is used.<p>This feels more like a way to measure the effectiveness of their astroturfing or low-key marketing efforts.
RemingtonLakover 5 years ago
Good job. Certainly see some potential. Our company used Clara in the past. I actually wrote my own in xls if you can believe it but it was highly vertical and tight topic so it was relatively easy. Where are you physically located?
评论 #21582703 未加载
guillaumecover 5 years ago
Interesting project, and I like the simple UI. How does the Sentiment Analysis is done? What kind of back-end have you used?
评论 #21572838 未加载
blader_johnyover 5 years ago
We made something similar at a hackathon: <a href="https:&#x2F;&#x2F;spicy-lip.surge.sh&#x2F;features" rel="nofollow">https:&#x2F;&#x2F;spicy-lip.surge.sh&#x2F;features</a>
评论 #21576097 未加载
koryover 5 years ago
Interesting concept, but the models powering it need a lot of work. For example, searching &quot;Apple&quot; brings up &quot;cider&quot;.
detaroover 5 years ago
negative: &quot;employer&quot;, &quot;apps&quot;, &quot;perf&quot;, &quot;idioms&quot;, &quot;culture&quot;, &quot;humanity&quot;<p>positive: &quot;closes&quot;, &quot;binary&quot;, &quot;btw&quot;, &quot;disclaimer&quot;, &quot;bullshit&quot;<p>The word cloud seems fairly useless.
JamesQuigleyover 5 years ago
Getting a &quot;503 Over Quota&quot; when trying to access your site :(
sali0over 5 years ago
This is fantastic, good work. Love the idea.
mk74160over 5 years ago
Very cool.
评论 #21577760 未加载