[{"data":1,"prerenderedAt":285},["ShallowReactive",2],{"project-27664020":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":27,"sentiment":28,"stance_mix":29,"scored_comments":34,"aspects":35,"strip":42,"post_text":43,"what":44,"said":83,"maker_in_thread":169,"shape":179,"grid":182,"attempts":246,"same_week":247,"maker_shows":170,"maker_hidden":277,"reader_version":278,"card":279},"27664020","The Sample","newsletters curated for you with machine learning","title","Show HN: The Sample – newsletters curated for you with machine learning","https:\u002F\u002Fsample.findka.com\u002F?ref=showhn","sample.findka.com","jacobobryant",1624897964,2021,67,32,710,2533,29,"notable","productivity","newsletter discovery","web_app",[24,25,26,21],"newsletter recommendations","collaborative filtering","recommender system","debated",0.123,{"positive":30,"neutral":31,"mixed":32,"critical":33},6,5,0,4,15,{"idea":36,"design":39,"performance":41},{"positive":37,"critical":38},3,2,{"positive":32,"critical":40},1,{"positive":32,"critical":40},[32,40,32,32,32,37,32,37,32,37,32,31,32,32,31,32,40,32,37,32,37,32,40,32,40,32,31,32,31,32,31,31],null,{"problem":45,"audience":56,"pricing":43,"motivation":63,"time_to_build":43,"stack":69,"openness":81,"stage":82,"team_size":81},{"value":46,"source":47,"evidence":48},"Discovering newsletters worth subscribing to","inferred",[49,52],{"id":50,"by":11,"text":51},"27664021","This is the latest of several recommender systems[1] that I've attempted to grow over the past couple years. It launched in February and there are about 800 subscribers. The…",{"id":53,"by":54,"text":55},"27666545","giansegato","Very interesting use of reco engines! I always thought that the current state of newsletter discovery is very poor. At the same time, I would not be ok with a product owning the…",{"value":57,"source":47,"evidence":58},"Readers looking for newsletters, and newsletter authors seeking subscribers",[59,60],{"id":50,"by":11,"text":51},{"id":61,"by":11,"text":62},"27677757","(FYI looks like Needgap is down right now; Cloudflare is showing me a 521 response) There are a few reasons. The biggest one is distribution. I came up with the idea for The…",{"value":64,"source":65,"evidence":66},"Build a general-purpose recommender system, starting with newsletters for fast distribution","maker",[67,68],{"id":50,"by":11,"text":51},{"id":61,"by":11,"text":62},[70,76],{"value":71,"source":65,"evidence":72},"Surprise Python library",[73],{"id":74,"by":11,"text":75},"27667018","I'm planning to build a business on this, so probably won't open-source it--but I'm always looking for interesting things to write about! I write a weekly newsletter called…",{"value":77,"source":65,"evidence":78},"fastText",[79,80],{"id":50,"by":11,"text":51},{"id":74,"by":11,"text":75},"unknown","launched",{"praise":84,"objections":105,"requests":131,"questions":152,"alternatives":164},[85,98],{"value":86,"source":87,"evidence":88},"Novel use of recommender systems for newsletter discovery","commenter",[89,90,94],{"id":53,"by":54,"text":55},{"id":91,"by":92,"text":93},"27666091","beders","The sample newsletter looked interesting, so I signed up. Thanks for providing a sneak-peek.",{"id":95,"by":96,"text":97},"27667128","dmje","Interesting, thanks! Have signed up.",{"value":99,"source":87,"evidence":100},"Human curation combined with ML is a good approach",[101],{"id":102,"by":103,"text":104},"27667077","visarga","I like that, so it's helping humans find the news but deferring the final selection.",[106,113,124],{"value":107,"source":87,"evidence":108},"Recommendations feel poor for some users; many low-rated newsletters",[109],{"id":110,"by":111,"text":112},"27669875","CallMeJim","I've been signed up to The Sample for a few months now, getting one newsletter a day. Each newsletter includes a link to rate it from 1 star to 5 stars. I consistently rate the…",{"value":114,"source":87,"evidence":115},"Risk of echo chambers from ML filtering",[116,120],{"id":117,"by":118,"text":119},"27665879","desine","Is there any mechanism to prevent you from forming your own super-echo-chamber? ML filtering only what you and your similar profiles like seems possibly have problematic…",{"id":121,"by":122,"text":123},"27666350","thegginthesky","Thank you for sharing your approach, I find it very interesting. I've worked in some systems that also optimize for \"serendipity\"[1], which is just optimizing not only for…",{"value":125,"source":87,"evidence":126},"Categories are too broad",[127],{"id":128,"by":129,"text":130},"27668605","PeterWhittaker","I'm definitely interested! But.... NOTE: These are not criticisms so much as release 1.1 (or even 2.0) use cases.... The categories are too broad (and, for some use cases, the 21…",[132,139,146],{"value":133,"source":87,"evidence":134},"Diversity of viewpoints, not just topics",[135],{"id":136,"by":137,"text":138},"27670536","atlgator","Instead of 20% being completely random, could you instead present some newsletters that are completely opposite to what they prefer? For example, if someone actively reads…",{"value":140,"source":87,"evidence":141},"Write-up or open-source of how the recommender works",[142],{"id":143,"by":144,"text":145},"27666859","qwerty456127","Thanks. I'll try that. > a surprisingly difficult problem for people not familiar with recommender systems Do you feel like open-sourcing and\u002For writing a post on how it works? I…",{"value":147,"source":65,"evidence":148},"Ability to set priorities across subscribed newsletters",[149],{"id":150,"by":11,"text":151},"27679219","> Moreover investment of time on user side for newsletter delivering news increases with each newsletter vs One newsletter delivering what I want. You might be interested in this…",[153,157],{"value":154,"source":87,"evidence":155},"Can the system avoid forming an echo chamber?",[156],{"id":117,"by":118,"text":119},{"value":158,"source":87,"evidence":159},"Why newsletters instead of news articles directly?",[160],{"id":161,"by":162,"text":163},"27673188","Abishek_Muthian","Hey Jacob, Why Newsletters instead of News directly? (I read [2]) i.e. Personal curated news as that would apply to everyone who consumes news and just not those who consume…",[165],{"value":166,"source":65,"evidence":167},"The Factual",[168],{"id":150,"by":11,"text":151},{"comments":34,"answered":170,"key":171},7,[172,173,176],{"id":50,"text":51},{"id":174,"text":175},"27666142","Yes, the algorithm accounts for the exploration-vs-exploitation problem.[1] Right now it uses a simple epsilon-greedy strategy: 20% of the newsletters you receive are picked…",{"id":177,"text":178},"27666711","> I disagree that it's overblown. Fair enough. To be clear, I'm not saying ML bias isn't a significant problem; and certainly I'll continue to address it as the project grows.…",{"commenters":180,"max_depth":30,"maker_replies":34,"first_comment_mins":32,"lifespan_hours":181},14,33,[183,185,187,188,190,191,194,197,198,200,201,202,205,206,208,212,214,216,219,222,224,226,227,229,232,234,236,237,239,240,242,244],[50,32,32,184,11,40],"project\u002Fneutral\u002Fidea",[117,32,40,186,118,32],"project\u002Fcritical\u002Fidea",[174,32,32,184,11,40],[121,32,32,189,122,32],"commenter\u002Fneutral\u002Fidea",[177,32,32,189,11,40],[192,32,37,184,193,32],"27669863","Jarwain",[195,32,32,196,11,40],"27669931","maker\u002Fneutral\u002Fidea",[136,32,37,184,137,32],[199,32,32,196,11,40],"27670659",[161,32,37,184,162,32],[61,32,32,184,11,40],[203,32,31,204,162,32],"27678062","project\u002Fpositive\u002Fidea",[150,32,32,196,11,40],[207,32,32,189,162,32],"27684038",[209,40,31,210,211,32],"27668037","maker\u002Fpositive\u002Fnone","text_exch",[213,40,32,210,11,40],"27668186",[128,38,40,215,129,32],"project\u002Fcritical\u002Fdesign",[217,38,32,218,11,40],"27668714","maker\u002Fneutral\u002Fdesign",[220,37,37,221,144,32],"27666527","project\u002Fneutral\u002Fdesign",[223,37,32,221,11,40],"27666763",[143,37,37,225,144,32],"maker\u002Fneutral\u002Fnone",[74,37,32,225,11,40],[110,33,40,228,111,32],"project\u002Fcritical\u002Fperformance",[230,33,32,231,11,40],"27670001","maker\u002Fneutral\u002Fperformance",[233,31,40,186,103,32],"27666996",[235,31,32,204,11,40],"27667044",[102,31,31,204,103,32],[238,31,32,218,11,40],"27667298",[53,30,31,204,54,32],[241,30,32,184,11,40],"27666829",[91,170,31,243,92,32],"project\u002Fpositive\u002Fnone",[95,245,31,243,96,32],8,[],[248,253,257,262,267,272],{"id":249,"name":250,"points":251,"category":252},"27614874","3dasd",248,"hardware",{"id":254,"name":255,"points":256,"category":20},"27642977","I’m an introvert",169,{"id":258,"name":259,"points":260,"category":261},"27689664","Git history visualization",147,"dev_tools",{"id":263,"name":264,"points":265,"category":266},"27662266","21st Night",123,"learning_science",{"id":268,"name":269,"points":270,"category":271},"27662114","Optery",111,"security_privacy",{"id":273,"name":274,"points":275,"category":276},"27657164","Whirlog",110,"data",false,"v2",{"id":280,"name":5,"line":6,"ai":277,"cat":20,"year":13,"points":14,"comments":15,"rec":27,"s":281,"top":18,"kw":282,"maker":11,"domain":10,"tier":19,"attempts":32,"strip":283,"url":9,"mix":284,"praise":86,"objection":107,"sub":21},27664020,0.12,[24,25,26,21],"01000303030500501030301010505055",[30,31,32,33],1794036608494]