marvinkennis

I just did a project on this, but for Bitcoin instead of stocks where I examined news, Reddit, forum (Bitcointalk.org) and IRC sentiment using some simple ML algos. The goal was to determine whether this data has any predictive causality.

I scraped the above sources over a full year (2015) and then had the data annotated on positive, negative and neutral sentiment.

The problem with labeling sentiment data is that there might not be a single 'true' label due to varying interpretations and ambiguity. So at best you'll get to 80-85% accuracy there. The less formal (News > Reddit/Forum > IRC), the lower your accuracy due to lack of context.

I Then matched the annotated sentiment to market data and did some causality analysis. What I found is that interestingly, you can't just say positive news = price/volume goes up. It is way more fine grained than that. For example negative Reddit sentiment leads price movements, but price movements lead positive sentiment. For news its the reverse.

All in all I didn't incorporate this into any trading strategies, but found it interesting to see the differences between online sentiment channels.

I created something similar yesterday afternoon [0]. Instead of opening a new tab, it just requests pages through a hidden iFrame and drops the headers so the request goes through.

It doesn't click on anything, because it would be awkward if this by chance started sharing things on logged-in Facebook profiles etc. I plan on adding sequential requests over the weekend so the traffic is more realistic.

It's open source and on Github, so you can download, install and modify it from there if you wish [1].

[0] https://chrome.google.com/webstore/detail/decoy-requests/aeh...

[1] https://github.com/marvinkennis/Decoy-Requests

I think the idea has potential. Maybe you should rethink the way feedback is provided. If I understand the idea correctly, it gives the 'customers' a choice between two products. When doing customer research you should never ask suggestive questions, which is what you do when you ask "Which one of these do you like?". Maybe they don't like it at all, but they still need to make a choice. This leads you to believe that one might be better than the other, while both ideas/products could be bad.

Thanks for mentioning. Should be back up now.

Well said. It does indeed promote buying, but only of products that are well designed (both in form and function), products that hopefully delight their purchasers and make their lifes a little easier.

There's something to be said for both sides, although giving attention to people and suffering is more important. I will definitely try to do more of both in another sideproject.

Thanks! You are right, I should add some kind of description. A title and an image don't quite cut it when trying to give a good impression of the product. And more importantly for this kind of site, explain why the product has been curated.

Uncrate is a great site. I'm trying to differentiate by keeping it to one product a day. I hope I can keep the quality of products that I curate high that way. I'll also try to keep it at products that are more or less affordable.

I think you misunderstood (which I figure is a fault on my part). The product on the frontpage is what has been curated today. But this made me think of something else; I should probably add product descriptions.

SEEKING WORK - Remote, from the Netherlands.

UI designer for web, mobile and desktop applications. http://www.dribbble.com/marvin

My email address can be found in my profile, my twitter handle is @MarvinKennis

It's nice to experiment with these techniques, but in my opinion it fails as a practical way to quickly provide the visitor with the information they need. From a visual point of view, something like http://pitch.csspiffle.com/ provides a more pleasant viewing experience.

SEEKING WORK - Amsterdam, remote

Visual UI and UX designer for mobile, web and desktop. Available for work from the 14th of this month. I do more than making your apps look pretty ;)

http://www.dribbble.com/marvin My email is in my profile and I can also be contacted on twitter @marvinkennis

Ha, I know. I wasn't stating my own opinion there though. Adding in a server-side check seems to be the only way to solve this (Note: My knowledge on this subject is limited), but doing so removes the advantage they have, which is the easy implementation. This would be cool if it worked flawlessly, but then again, it doesn't (yet).

Last time I checked this actually worked in production as well, as the values are changed before the form is submitted and no data is pulled from a server to validate. So SimpleCart should be used for low traffic websites, where orders are manually inspected, as they said on their github page https://github.com/wojodesign/simplecart-js/issues/150

You could look around the design communitites such as Forrst and Dribbble to find someone whose design you like. I have no insights in their rates or your budget though.

Hire a well skilled designer?

Criteria like age, location, education, interests etc. You can see the entire list when creating a Facebook ad: http://www.facebook.com/ads/create/. But the advertisers don't get to see what belongs to who, unless someone clicks on an ad, buys something and they backtrace all of that. They only target it.

I don't believe that people will stop using social networks, it may just be in a different form than it is now. People have the need to stay connected and see what other people are doing with their lives, and share pictures or information with others. And the internet with its social networks has been a great way to do so. And that is not going to change in the near future. If you come up with something innovative and there's a need for it, why not try it? If you never do it you can never win.

Grey background, not enough shine on the phone, English apps on a Swiss page, slider doesn't have the right effect, I believe there is also no mention of 3G or 4G on the specs page. And last but not least, the first second of the video shows a non-apple url: Users/peter/desktop/iphone%205...

Other than that it's been made nicely and had its round of exposure around the web :).

Why is that?

But for the users, that wouldn't really matter?

I know of someone who is currently developing something like that. Should be done by mid september.

I had this idea last year but dismissed it, since I though search engines already used this. Makes me want to hit myself.

First off I wanted to say that you have a great site, and is very easy to use, nice work! It has always been though to get media attention overseas (atleast for me). But why not build your empire in Europe first and get satisfied customers there? As the growth graph starts flattening out you can move on to the US. If it becomes a success in the EU it is easier to expand to the US (already lots of EU press, happy customers etc.). Worked for Spotify.