Earlier quoted context omitted.
Looks like there are a few of us on HN trying to solve the news consumption problem. My site[1] tries to crowdsource the summaries by encouraging the readers to summarize a story themselves. In the meantime, we create most of the summaries in-house. Newser[2] tried to do the same thing but gave up and just focused on in-house created summaries. Regarding the summary process, it's good to see another group manually su…
For anyone working on the summary problem: Chapter 8 of O'Reilly's Mining the Social Web has a great description of how you can use Python NLP libraries to accurately break text into individual sentences, then analyze those sentences to pull out the most important. I don't know how well it works in practice, but their example is amazing.
TL;DR — Faster News
91–100 of 142 posts
Re: TL;DR — Faster News
#92Humans, no matter how altruistic, have inherent bias that will influence their selection of news to summarize, as well as the nature of their summarization.
Circa is another prime example of right idea, problematic solution. It gives me news and photos with no attribution... not even bylines of whomever created their shortened bits of things that they claim could be news. (intentional sarcasm) There's no mechanism whereby I can learn to trust them, or toolong-didntread.com.
Re: TL;DR — Faster News
#93Why don't people like to read anymore?
Re: TL;DR — Faster News
#94Priorities?
Re: TL;DR — Faster News
#95I don't want to step on these guys' PR, but I do have a similar personal project for anybody who is interested, http://newspaper23.com Initially it's just an aggregator that presents commentary in plain text. I plan on adding a summarizer one day. For a personal project, I've been using it daily for over a year, so I know I find a lot of value in this type of thing. As sites try to get more sticky, the signal-to-nois…
I'm guessing it has to do with column width and coloring.
Re: TL;DR — Faster News
#96Earlier quoted context omitted.
Looks like there are a few of us on HN trying to solve the news consumption problem. My site[1] tries to crowdsource the summaries by encouraging the readers to summarize a story themselves. In the meantime, we create most of the summaries in-house. Newser[2] tried to do the same thing but gave up and just focused on in-house created summaries. Regarding the summary process, it's good to see another group manually su…
I'll add my baby to the pot: http://skimfeed.com .
Re: TL;DR — Faster News
#97Common problem, problematic solution. Humans, no matter how altruistic, have inherent bias that will influence their selection of news to summarize, as well as the nature of their summarization. Circa is another prime example of right idea, problematic solution. It gives me news and photos with no attribution... not even bylines of whomever created their shortened bits of things that they claim could be news. (intent…
Re: TL;DR — Faster News
#98I always had thought it was TooLazy;Didn'tRead....I guess TooLong also works.
Re: TL;DR — Faster News
#99This is a actually a very nice effort.
The issues I have with growth and actual value, is with how the summaries are generated. Automatic generation is fast and inaccurate, manually curated is slow and very accurate. In a world where people no longer have the time to read newspapers, their not only looking for quick news on the run, they want current news. Something that happened today, everything that happened today. But quick, not the full story, "i'll read that later." Meeting in the middle between fast and slow approaches does not work here. You're too slow, and the headlines in my RSS feed and twitter have already informed me of the news, too fast and your summary becomes a failed attempt to make twitter and RSS better quality. I have no idea how your TL;DR are currently generated. But I would think you'd have an aggregation and be doing some manual curation. To me for this to really work, you'd have to have a large group of people that read the article generating the TL;DR, constantly iterating, until you end up with an extremely efficient 2 sentence summary. Or there needs to be a project that integrates TL;DR on large scale, the publishers, news papers of the world, blogs..They submit these directly.
I think there's still a lot of value in what you're doing. I just don't think it will take off as it is now. Away from the name/marketing/novelty/social aspect not really being there. Twitter, RSS feeds, and sites like http://skimfeed.com/ end up providing me with far more day to day value. If you took this and spoon-fed me the TL;DR via my phone, i'd consider being a repeat visitor a little bit more. But you'd then be competing with a whole 'nother slice of the pie.
Re: TL;DR — Faster News
#100Earlier quoted context omitted.
agreed. the current "hand curated" approach does not seem sustainable in the long term. it's already showing some quite old news as a result. the new jersey earthquake was several weeks ago, positioned as the third "world news" story. great concept and design, though.
>the current "hand curated" approach does not seem sustainable in the long term I disagree. I had the same idea for a news site and I'm disappointed they beat me to it. I think it would only take me maybe 30-minutes to and hour to summarize the news of the moment. Do it three times a day. How hard is that? You could easily hire extremely smart people to do this. What English major at a top college wouldn't love to pu…
That said, I know first hand most people are incapable of writing well much less condense into exact words. This was on my list of to do as well and I've interviewed some writers and those of relevant majors and I assure you, the stuff that comes back is not as good as you think it would be. In the long run, you'd need an army of decent writers that can condense content well.