[HN Gopher] Mapping Hacker News to find who knows what in the HN...
       ___________________________________________________________________
        
       Mapping Hacker News to find who knows what in the HN community
        
       Author : robg
       Score  : 115 points
       Date   : 2024-07-25 15:04 UTC (8 hours ago)
        
 (HTM) web link (blog.wilsonl.in)
 (TXT) w3m dump (blog.wilsonl.in)
        
       | robg wrote:
       | As Wilson and I are working on this project, we'd love to hear
       | your thoughts!
        
         | ipnon wrote:
         | The visualization of the space is excellent.
        
         | thomasfromcdnjs wrote:
         | It's very fun to play around with, there _seems_ to be value
         | somewhere.
         | 
         | I wonder if you could make a video of how an individual could
         | use it effectively in a pragmatic sense e.g. networking,
         | research
         | 
         | (after more playing around I get the UX, very powerful indeed)
         | 
         | (feedback: add some controls for filtering (dates etc),
         | relevancy, color islands slightly, just general UX improvements
         | and I would probably pay a few dollars a month for this)
         | 
         | (feedback: when looking at a user, you have to zoom really far
         | in before results start appearing. it should probably show
         | labels/results if less than X amount of data points are in your
         | viewport.)
         | 
         | (feedback: Is there "Explore more users" section the more
         | related users to the current profile? It would seem so, but not
         | immediately clear. And if it is, none of my related users have
         | me as a related user in their profile.
         | 
         | ===
         | 
         | This is really inspiring work, hats off!
        
           | robg wrote:
           | Thanks so much and for the great feedback!
        
         | disqard wrote:
         | Thanks for making and sharing!
         | 
         | Could you share some info about how you generated the 2D space
         | embedding/visualization?
        
           | robg wrote:
           | [delayed]
        
         | JohnFen wrote:
         | OK, I'm going to just be upfront here and admit that I'm an
         | idiot. This is the third time I've seen this link come up on
         | HN, and each time I've checked it out, but I can't really
         | figure out how to interpret what I'm seeing there in a way that
         | is useful.
         | 
         | How can that map be used to determine who is knowledgeable
         | about what? Looking up myself, I can't connect what I see with
         | my own areas of knowledge.
         | 
         | I think I need an ELI5 for this. Again, this isn't a criticism
         | of the effort at all, it's a public admission of my own
         | ignorance.
        
           | djoldman wrote:
           | Take everything a user says and encode it into some vector.
           | That's "embedding."
           | 
           | Then you encode a search query the same way and return the
           | nearest N vectors/users. That's vector search.
           | 
           | It seems they did the above for HN comments.
        
             | JohnFen wrote:
             | Right, I understand all that. I just don't understand how
             | to interpret the resulting map (or how to formulate a
             | search) to provide useful information, specifically "who
             | knows what".
        
               | robg wrote:
               | Thank you for being so clear on where we are failing you.
               | We're basically computing _your_ semantic space as
               | compared to the whole community. The map shows the
               | relevant instances in that space. As you browse your
               | concepts the space changes and we link to the relevant
               | comments from the associated threads. Who knows what is
               | prioritizing the user and the semantics  / voice of that
               | user amidst the noise of the whole community speaking at
               | once.
               | 
               | We'd love to hear your feedback as you play with it more
               | and we improve all of the above.
        
         | jmuguy wrote:
         | Looking at my own profile the words (topics i know about?) seem
         | sort of generic. "issue", "company" - not sure what to make of
         | that other than I assume those are words I use a lot in
         | comments?
        
           | robg wrote:
           | We'll take a look at how to do better, thanks!
        
         | toast0 wrote:
         | I feel targeted when you call me a "system server user network
         | number service code issue" ... not that it's inaccurate ;)
        
           | robg wrote:
           | Ha, exactly!
        
       | wcedmisten wrote:
       | Wow I really like the topographic visualization of the data,
       | looks really slick!
        
         | robg wrote:
         | Thanks so much!
        
       | causal wrote:
       | This is really cool. Also a healthy reminder that anything we
       | post publicly is likely to be analyzed. With a little analysis
       | it's probably easy to know me better than I know myself.
        
         | robg wrote:
         | Maybe also who you'd tend to get along with and your unique
         | place in the community?
        
       | junon wrote:
       | Would be interested in a study, unfeasible as it may be, on the
       | social compatibility of people who have the strongest overlap
       | between themselves and the next closest user.
       | 
       | Cool visualization and analysis, really well made!
        
         | robg wrote:
         | Thanks so much! How do you see that study playing out in
         | metrics that we can measure and examine?
        
       | mattdesl wrote:
       | For the "terrain contours" is there something specific you've
       | done to make it feel more cartographic? Or is it basically just
       | marching cubes / iso lines on some data points?
       | 
       | Looks fantastic. Very cool project.
        
         | robg wrote:
         | [delayed]
        
       | batch12 wrote:
       | When I did this, I found it to also be an interesting way to
       | fingerprint users and find alternate accounts. I was able to
       | match an old account I used in a top 10 similarity match out of
       | all users.
        
       | rob wrote:
       | https://hn2.wilsonl.in/user/rob
       | 
       | Seems pretty accurate. I do love me some PHP and WordPress.
        
         | robg wrote:
         | [delayed]
        
       | hu3 wrote:
       | Searched for "dupe" and yes, it correctly pointed out the one
       | user that takes it a hobby:
       | 
       | https://hn2.wilsonl.in/search/dupe
        
       | 0xbadcafebee wrote:
       | HN and other forums perpetuate a fallacy (that I assume isn't
       | well known because nobody has published a research paper on
       | it)... but "the wisdom of the crowd" isn't.
       | 
       | "Trusted voices" on HN are often not experts, yet espouse views
       | which are not what the larger body of experts would consider
       | correct. In addition, actual expert voices are drowned out by
       | whatever the popular position is. HN also provides its own
       | cultural bias, in that everything spoken on HN has to follow a
       | rigorous cultural sieve set down by the guidelines, such that a
       | "negative" view, even if correct, is considered either wrong or
       | distasteful and buried. This is exacerbated by banning the use of
       | humor to disarm controversial or heated comments. And this is the
       | comments that HN does get; many opinions are never entered as
       | comments here, so there exists a large knowledge gap. Then
       | there's the "taboo" subjects like race, gender, religion,
       | politics, social justice, etc which get buried for fear of
       | controversy, so you're definitely not gonna find any expert
       | opinions on those, as the stories just aren't there for
       | discussion.
       | 
       | The end result is that often experts go unheeded or even
       | downvoted, popular shallow opinions get upvoted, and substantive
       | commentary based on evidence and experience is frequently
       | missing. The fact is that we have no idea who knows what, or
       | what's true or right. We just have "popularity" according to the
       | particular cultural quirks of this site. So you can definitely
       | find out "who _thinks_ what ", and who is considered to be more
       | trustworthy to a HNer, but it has absolutely nothing to do with
       | objective truth or the body of real knowledge that exists outside
       | the world of HN comments. This is an echo chamber, but it's not a
       | chamber of experts. It just seems that way because occasionally
       | you see a minor tech celebrity, and people talk with absolute
       | authority regardless of if they have any.
       | 
       | You want to find out who knows what? Look at their diplomas and
       | careers. If they've done 20 years in a single field, probably
       | they're an expert. If they have a degree (or multiple) in a
       | field, probably they're an expert. If they spent half their life
       | working on a single hobby, they're at least very knowledgeable in
       | that field. But you can't determine that just by looking at who's
       | talking about what or how many completely subjective "points"
       | they get for what they say. Determining real knowledge requires
       | analysis of specific criteria, filtered to get a higher quality
       | result.
        
         | saagarjha wrote:
         | Humor is not banned on Hacker News. Bad attempts at it are
         | penalized, though.
        
         | robg wrote:
         | IMHO if there's any take away from the last 20 years of the
         | internet, it's how hard it is to have decent communities of
         | people online, curated like a blooming garden of diversity.
         | What are some online communities that you appreciate?
         | 
         | I have a Ph.D. in neuroscience with exactly 20 years of
         | experience. I talk about what I know and try to be curious
         | about what I don't know. HN really helps me on both counts. How
         | does it help you?
        
       | wonger_ wrote:
       | Personally, I like how HN focuses on content and discussions
       | rather than individual users. If I wanted to follow experts, I'd
       | probably curate a selection on a social network like Mastodon, or
       | kludge together some RSS feeds.
       | 
       | Also, I feel like this tool selects for active commenters, not
       | for knowledgeable experts. Not to mention throwaway accounts.
       | 
       | Still a cool project.
        
         | robg wrote:
         | Thanks! Those are fair points. We're thinking we could uplevel
         | the social layer so you can connect with people of similar
         | interests for deeper connections. In this way we compute not
         | just your contributions but how they relate to others.
        
       | photochemsyn wrote:
       | > "Despite the intervening 16 years, we're amazed that social
       | networks, even Hacker News, don't compute and display the trusted
       | voices across topics. Instead of prioritizing pages based on
       | content, social networks could prioritize the people behind the
       | content."
       | 
       | Fallacy: appeal to authority. Practically, just because someone
       | generates great content on subject A doesn't mean their take on
       | subject B is any better than random. A well-reasoned self-
       | consistent argument informed by accurate data is far more valuble
       | than 'trust this expert opinion because this expert can be
       | trusted' approaches - although it may require more work on the
       | part of the reader. Don't get lazy.
        
         | robg wrote:
         | We're trying to stay away from authority and just compute based
         | on what you say then transparently show where that information
         | came from. How do you think we can do better?
        
       | simonw wrote:
       | https://hn2.wilsonl.in/user/simonw includes "Risk of COVID from
       | pianos" down at the bottom of the map. I'd love to know where
       | that came from!
        
         | wonger_ wrote:
         | Appears to be a summary of this comment:
         | https://news.ycombinator.com/item?id=30105274
         | 
         | If you zoom and pan the red cursor over that map location/text,
         | your source comment will appear underneath the map.
        
         | robg wrote:
         | Great call out! Would it help to make the points more
         | interactive?
        
       | DoreenMichele wrote:
       | Just curious what the mechanism is for adding a bio to individual
       | users. Your sample -- robg -- has the same bio on your stuff as
       | on HN but others have stuff on HN and "No bio provided" on your
       | thing.
        
         | robg wrote:
         | Right now if you add it to your HN profile we'll be able to
         | include it.
        
       | zamalek wrote:
       | Some sentiment analysis might help? Apple is one of my keywords,
       | but you'd pretty-much only get an "avoid" response from me (to a
       | degree, I have recommended their phones to a _very_ specific
       | demographic).
       | 
       | Awesome project, fantastic UI.
        
         | hu3 wrote:
         | Same. I think a sentiment analysis could be helpfull.
         | 
         | I find myself dragged to Apple discusions more than I would
         | like to admit, but they are generally negative in the last
         | years.
        
           | robg wrote:
           | It's a topic we keep coming back to, seems like we can do
           | better than sentiment analysis to show the semantic space of
           | the sentiment(s) and how those breakdown by users and groups.
           | For instance assuming a large discussion set of apples,
           | different types of apples, and other types of fruits, we'd
           | discover the patterns among users. Have you seen folks doing
           | that type of semantic sentiment analytics?
        
       | kaycebasques wrote:
       | Super cool! Happy to see that I show up under documentation-
       | related stuff, considering how frequently I mention that I'm a
       | technical writer in my comments. Also pretty excited to connect
       | with the other people that show up related to docs / technical
       | writing / etc.
        
         | nadermx wrote:
         | As opposed to me not showing up at all for things I comment on.
         | Guess I'm not knowledgeable enough.
        
           | layer8 wrote:
           | The results are biased towards users with high post count. It
           | doesn't matter how knowledgeable you are about a topic, only
           | how much you comment on it.
        
             | kaycebasques wrote:
             | And using the correct keywords. SEO is dead, long live SEO!
             | 
             | Also I imagine that my area is pretty niche (not many
             | people talking about documentation on HN) so I may have
             | less competition
        
           | robg wrote:
           | We'll take a look!
        
       | pcthrowaway wrote:
       | How does determine which comments are valuable ("trusted")
       | without access to the vote count of comments?
        
         | robg wrote:
         | We're trying to stay away from karma and focus on what people
         | are saying within the community, rank orders, overall
         | semantics, and more. How do you think we can do better?
        
       | dang wrote:
       | Recent and related:
       | 
       |  _Show HN: Exploring HN by mapping and analyzing 40M posts and
       | comments for fun_ - https://news.ycombinator.com/item?id=40307519
       | - May 2024 (159 comments)
        
         | robg wrote:
         | Thanks dang! Amazing to me how much HN brought Wilson and I
         | together after 16 years and across the world. We're hopeful
         | about the future of trust on the internet exactly because of
         | communities like HN.
        
       | neilv wrote:
       | > _Mapping Hacker News to find who knows what in the HN
       | community_
       | 
       | Would things like this create/increase incentives to game the
       | metrics?
        
         | gradyfps wrote:
         | In my limited experience, people tend to gamify any numeric
         | metric readily apparent to them.
         | 
         | Anecdotally, total game time (in hours, usually) is used to
         | convey experience in video games (WoW, CS:GO, PUBG). I've seen
         | people create & run 3rd-party software to artificially inflate
         | these types of metrics.
        
           | neilv wrote:
           | And some metrics aren't just a "game", but are gamed for very
           | real benefit (e.g., gaining admission to a college, getting a
           | promotion, being more employable) in ways that are
           | _counterproductive_ to pretty much all other goals.
           | 
           | So, when something like HN has goals, and you create, say,
           | something akin to a contest and link it to real-world
           | benefits (e.g., people expect a recruiter or hiring to use
           | that "who knows what" info), then many people will modify
           | their behavior, probably to the detriment of original goals.
        
             | robg wrote:
             | We're trying to stay away from karma for that reason and
             | focus on what people are saying within the community to
             | find trusted voices. How can we do better?
        
               | neilv wrote:
               | How can Web sites do better at producing valuable and
               | trustworthy content, when what everyone cares about now
               | comes down to SEO?
        
       | asp_hornet wrote:
       | I hate this. If i used my real name id have some pleb website on
       | the internet making chump guesses at what i know based on what i
       | was prepared to share. Thanks for reminding me to use anonymous
       | burner accounts on everything app i use.
        
         | robg wrote:
         | Thanks for the feedback. How can we better support authenticity
         | and trusted voices, where you fit in, and others like you?
        
       | motohagiography wrote:
       | Good God I am a bore. May track some people down in my niche
       | interest areas though, thank you for this.
        
         | SCUSKU wrote:
         | Oi! Watch it! Our distance in embedding space is small. Then
         | again, maybe I am a bore...
        
       | mmastrac wrote:
       | I found it a bit challenging to actually drill down into my own
       | username, but it doesn't seem to offer much other than throwing a
       | lot of dots all over the map. I'm trying to understand what the
       | overall clusters might be, but most of them are just
       | android/apple/google?
       | 
       | https://hn2.wilsonl.in/user/mmastrac
        
       | qdot76367 wrote:
       | Searched for "Buttplugs".
       | 
       | My name pops up.
       | 
       | Hell yeah. A++++ completely accurate would use again.
        
         | robg wrote:
         | [delayed]
        
       | gnicholas wrote:
       | Interesting. I have noticed that my most-upvoted comments relate
       | to legal questions, which I have relatively more expertise than
       | most HNers (I used to be a lawyer). Although I'm not among the
       | top commenters for law/legal/lawyer according to this tool, I
       | definitely recognize some of the top names and can recall seeing
       | their comments in legal threads. Pretty cool tool!
        
       | dmurray wrote:
       | Things I thought I had expertise in and have posted about
       | disproportionately on HN: chess, Python, HFT, Fermi estimation,
       | Ireland.
       | 
       | Keywords the site actually associates with me: language, English,
       | article, team, book. To be fair, at some zoom levels I do get
       | "chess".
        
       | rpmisms wrote:
       | I wonder how high I am on the guns knowledge list?
        
       ___________________________________________________________________
       (page generated 2024-07-25 23:05 UTC)