Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Wishminer – Mining social media for app ideas, my weekend project (wishminer.com)
164 points by julvo on March 25, 2018 | hide | past | favorite | 59 comments

Dear HN,

Currently, I am stuck in a long-term project and really had the desire to build and finish a small app - I guess some of you can relate.

So this weekend, I built Wishminer, a little tool which searches social media for app ideas that people wish for. I hope it can be helpful to some of you. It's fairly basic and nowhere close to perfect, but I would appreciate your advice on how to improve it.

Two ideas:

1) clustering similar ideas, for example when I looked at it there were multiple about finding friends. Related, you could build an reddit-style ranking instead of a purely temporal one

2) if you want to turn it into a product, maybe a way to share “hey, this is exactly my app”... saw some similar comments here. Probably you’ll have to work on anti-spam too :)

Good luck, great project so far!

Thanks for your ideas. 1) would be a really cool feature, but also technically quite challenging.

2) Could potentially also be helpful as email alerts, so that an app maker gets notified when someone wishes for his app so that he can reach out on Twitter. Commenting on Wishminer itself will probably not be seen by the 'wisher'.

1) could probably be achieved reasonably by filtering out keywords from the sources and simplify them ("app find friend" or "app sell item") and generate tags based on that.

1) you could also try to use word embeddings for distance. Afaik they work better than bag of words when few words are present in one observation.

Cool! Interestingly, I'd seen a product on indiegogo that kinda answered one person's wish, so I tweeted it to them (so that idea is out :-) And no, I'm not gonna say what it was, I'd feel like I was product-spamming here.

I'd love to learn more about what consumer products people are interested in. I'm always looking for consumer trends and new product ideas. I analyze a lot of products for my Amazon business especially now that I've started getting into manufacturing my own products, so I could definitely use a service that can help gauge consumer interest in social media and quantify it.

That's great! This way I guess both sides could benefit, people looking for product ideas as well as people searching for solutions :)

Super cool weekend project! This is a great example of an MVP. The design and functionality are all attractive, intuitive and consistent, but you kept the scope very narrow and didn't get distracted by different filter methods, search, etc.

Nicely done!

I see a lot of 'a laptop for school' wishes from teenagers. I'm curious whether the current generation of cheap laptops just aren't cheap enough, or is it a status thing where they actually mean they want an expensive Macbook?

Having been poor, for a lot of people, you don't just "go out and buy a laptop." You have to save up for it. You might be working two jobs while going to school just to cover the bare costs of tuition, housing, food and possibly transportation.

I saw one instance where it sounded like they had a computer, but not a laptop. (Or they were stuck in the computer lab.) Another said they wish they had a computer to help them jump-start a career for their family.

And like my sibling post points out, sometimes the laptops that are available for $150-200 really are pretty terrible. When you have a limited budget, it isn't appealing to save up for something, and then you buy it, and it's buggy or painfully slow, or unable to run the software you'd need it to for your classes.

And finally, when you don't have a lot of money, and you don't already have a good computer yourself, and you maybe haven't learned all the tricks, you don't know how to find a "good" laptop on the internet. (Not that it's wildly hard - Amazon has some decent ones, depending on price point... but you won't be coming from a place of confidence in your buying skills.)

The next step up seems to be - how many compromises can I avoid at my price point? (This is super common on SlickDeals: you see $600-1000 laptops, and people pick apart the CPU, the GPU, the weight, the screen technology, the HDD vs SSD, the amount of RAM, the build quality. None of them get all of those right because if they did, they would be $1200+ laptops, not deals!)

Having been in the same place, I can second all of this. Even I’m my (40% below market) current job I still take my time weeding through comparisOne to consider the most weight for my money. The city is expensive.

That said I know of programs in this city where disadvantaged people can apply to purchase refurbished laptops for between 30 and 100 CAD. They won’t get you far in the work world but they might get you through a boot camp or part time management of an EBay account or job search and some online courses, etc. (Toronto)

There could be more of that, yeah? I haven’t had to go to that route. But I can attest to the ability to get to that next rung being strengthened by having the correct tool.

Or it could be that they want something that's not bottom-of-the-barrel cheap while also not being as expensive as a macbook.

There are people who know they don't need a $1200 computer, but also don't want a $100 one that crashes when you open the sixth browser tab.

The middle ground of markets like this is always a crap-fest, it's the same with phones: when you're not solely competing on price (performance is ~no object) or performance (price is ~no object), you end up trying and cram in as much crap as possible to differentiate your product. Look at any mid-range laptop or phone's sales pitch and it'll scream features and specs, and consumers are drawn into trying to compare by these metrics, and the game becomes all about these pointless numbers, rather than how well it'll actually just handle what you want it to do. As a manufacturer, you care more about having a processor with a faster clock speed, or a hard disk with more space than a setup that'll launch launch Chrome and get to Facebook in n seconds.

Chromebooks are perfectly positioned to totally occupy this middle-ground with a fast-ish machine that just works, I don't know why they haven't pulled it off yet.

Just from the first page I found this one, "A restaurant that does food for when you feel too sick to eat real food" which is interesting because normally restaurants are not eager to cater to sick people, but if you combine it with home delivery there might actually be a market.

Once I was out to dinner with my boss and his wife was sick at home. He asked the waiter for a to-go order of “something very mild and bland”—I think the result was steamed vegetables and rice.

I found this really interesting at the time because most restaurants wouldn’t like to describe any of their food as “bland”! But sometimes that’s exactly what you want—if you’re sick, hung over, a picky eater, or very sensitive to textures/flavours (e.g. with ASD)—and you may not have the energy or time to prepare something for yourself.

So, a chicken soup delivery service?

Delivery people in hazmat

My favorite is "An app that reminds me to turn my phone on".

I really miss alarms from the old time phones, the ones that didn't require your phone to be open already

If you keep your phone in flight mode regularly, why not?

Wishminer is a lot of fun, partly because some of the suggestions are so crazy. I enjoy scrolling through them.

Good work! This could prove to be useful for both the wishers and wish-granters. It looks great, but I had a couple of issues on my first impression:

1. Not much info at the top of the page, could use a description, especially mentioning the goal of trying to help people fulfill these wishes. (I initially thought this was more of a selfish goal of mining/stealing ideas)

2. All the wishes seem to be from Twitter, apart from one Reddit item. Maybe a filter to categorize the different sources, or a total from each source at the top of the page.

Very cool though, thanks for sharing and I hope you continue to work on it!

Thanks for taking the time. The missing description is a good point. I will probably add an about section, to avoid the misunderstanding.

Twitter is the dominant source at the moment. Yeah, maybe filtering by source could be useful, just wondering if reddit in this case would be too sparse. I'll try to add more sources, but couldn't find any other good ones so far.

An app for planning holidays with 1 click

What's the closest to this? Is there an app where you can specify a destination/price/date and have a 1 click checkout with your bookings and tickets?

By definition that app would have to hold on to your payment information, passport numbers and so on; I also think that all the evidence points out to people being willing to click a lot more to save money on holidays by exploring the option space.

I vaguely remember a service posted here where you’d sign up and they’d find a random trip for you and you wouldn’t know where you were going until the day of.

Google flights is probably the closest to one click though. Super fast to view destinations and prices without much effort.

Maybe you are talking about Srprs.me[0]? They are only available in a few countries within Europe, and they only let you travel to to other countries in Europe but it is exactly what you are talking about.

[0] https://srprs.me

Edit: changed link to the correct one.

There’s always going to be limitations because you can’t get visa or vaccinations.

A relationship with a travel agent where it has been established that you want to be very hands off.

That's ofcourse the issue; travel agents have been doing this since forever, but most people now want easy and cheap and travel agents are expensive. So expensive that many people don't use them and go through the pain of arranging all themselves. Very much not 1-click, but much cheaper generally.

Are they? I was under the impression that they mostly work on referral fees/commission. They get paid by the hotels, airlines, cruise companies, etc instead of booking.com or wherever you would look things up yourself

With that, I thought people stopped using them because most find some value in planning their vacations themselves, and the barrier to doing so got easy enough for them to do it. There's a huge subjective component to a lot of choices involved in a holiday. A travel agent seems ideal for someone willing to give up control of those specifics, that just wants to be able to send an email that reads something like "15-28th August. Portugal. I'm thinking something involving forests" and be done with planning.

> Are they? I was under the impression that they mostly work on referral fees/commission.

I like the idea of a travel agent and have tried many times, also for company travel, tried them; I have never had quotes less than 2x higher than finding myself. And others have similar results. Experiences may differ but this is what we generally find in NL / UK.

Heh. An app-interlock: "An app that only allows you to call taxis, your parents or 999 when drunk and not any other numbers"

This is genius (the site, not the above idea).

Weird but saw at least 3 tweets validating my existing app, likely to be frequency illusion.

Was definitely addicted in moving down to see more.

Im sharing this with my friend right now.

Brilliant mvp, impressive cares for details bravo!

Note that you're rewriting html entities to their readable counterpart, eg:

``` Loved & Found ```

Thanks you, good catch. Will fix it.

someone wishing for myspace app made me laugh

I just saw one which said "a show that's all about you", This sounds like a very creepy idea which should be turned into a black mirror episode.

And as always with this idea, people have no idea how IT and app development works. I also want an app that does my dishes but that’s not how it works..

Bad example...


Not too far off. If you get in the habit of putting dirty dishes in the washer there’s no reason it can’t detect a full load and start a cycle. Could even integrate a soap reservoir so you don’t have to even do that every day.

> Not too far off

There’s a big difference between an app that does my dishes and an app that controls an expensive and space-consuming machine that does my dishes.

Yes, you can’t have a $1 digital phone app that does the dishes on its own, but I think that was OP’s point.

> I also want an app that does my dishes

How about an app that let’s you hire people to come to your house and wash your dishes? All you need (after initial setup) is a single tap.

My startup, DshWshr[1], is close to launching and we’ll revolutionise the dish washing space.

[1]: I’m joking.

Sooo you ignore the dumb ones and maybe one of the good ones will get you excited? A lot of people are looking for side project ideas and this could be a great resource for them.

This simply parses social media for people saying they want things. It isn't really a list of app requests, nor are all the sources serious.

Decent idea, but super hard to filter spam it seems.

Best of luck!

First three or so pages, four different versions of Tinder-but-for-finding-friends.

Seems very good for a weekend -- for an year, even.

How do you do it?

Thank you. I'm using Firebase as a database and for hosting, which makes development really fast.

For getting the data, I'm using a worker written in Python, which queries the Twitter and Reddit APIs and does some regex filtering before writing it to the database. In the future, it could be interesting to experiment with more advanced NLP techniques for filtering/grouping/classifying.

Very interesting, thanks for sharing. Im impressed by what you've done with just regex for the filtering. Which makes it possible to curate a dataset of wishes to do some unsupervised clustering and fitering.

Filtering could benefit from knowing what other syntax ( phrases ) would map to wishes, broadening the coverage.

Would love to work on this --let me know if you're open to that.

Either way, this is a useful project

Re: finding other syntax, have you seen https://answerthepublic.com ?

That site lets you type in some key words and then it finds what questions people are searching for on Google based on those keywords (with prefixes like "who", "what", "where", and search suggestions).

Then it makes a nice big graph of the questions.

That's an amazing resource, thanks for sharing. Using autosuggestions might actually be really useful.

Hey guys, we're working on a platform that literally collects wishes from any touchpoint and it has quite some traction, especially in the gaming industry. Could you send me an email to hn{@}stomt.com? Would love to catch up on this. We also do a lot of research in the fields of NLP and clustering.

Unsupervised learning could be really interesting for this. Let's have a chat about how we could work together. Would you mind dropping me an email either via the website or to julian{at}wishminer.com?

Hey, I'm thinking about doing something similar with baseball JSON apis of all things. Would you mind sharing your firebase/python implementation details with me? Thanks!

I am curious to know how you did this filtering with regex. Can you share an example or any tutorial for reference?

Search for tweets containing "I wish there was". Remove the "I wish there was" part. (Profit?)

Pretty much what undai says. It then comes down to building a dictionary of search queries and a 'blacklist'. The regex itself is relatively simple.

neat. thanks. will probably waste a lot of time on this :)

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact