Parsing prices with weight and ingridients from sites of supermarkets or marketplaces could be an initial step, but for whatever reasons all that businesses try hard to prevent such activity, from showing prices as pictures to limiting number of queries per minute for IP. So you need some distributed network with OCR and other abilities to do that. Kind of interesting, but sophisticated task that needs large amount of supporters willing to run browser plugin or separate utility. :)
A saw few price aggregators, but they are mostyl outdated and does not provide necessary additional info.
IDK, if some marketplaces as Amason or local analogues could be used for data scraping. Limiting scraping to only one site will make things much easier, but there will be question of data revelance.
Groceries needs its own index, Utilties need its own index, Housing needs its own index at least.
I think it should be even finer division. Say, groceries should be divided into diary, flour products, vegetables and so on. More common indexes will be piece of cake.
Shortly, it should be some browser plugin to daily scrap data for single or several commodity from marketplace site and send it to some database, preferrably distributed one. Then, there should be sites or local apps to show data from that database as graphs, averages, indexes and so on. Technical problems are solvable, I see how that could be done, but I hardly could imagine how to convince many people to participate in such adventure.
Yeah getting people to participate will be hard part. It would need to be accessible and somewhat easy to add data otherwise it will be too much effort. But not too easy that it inspires manipulation through the data itself.
This could be a good start.
Parsing prices with weight and ingridients from sites of supermarkets or marketplaces could be an initial step, but for whatever reasons all that businesses try hard to prevent such activity, from showing prices as pictures to limiting number of queries per minute for IP. So you need some distributed network with OCR and other abilities to do that. Kind of interesting, but sophisticated task that needs large amount of supporters willing to run browser plugin or separate utility. :)
A saw few price aggregators, but they are mostyl outdated and does not provide necessary additional info.
IDK, if some marketplaces as Amason or local analogues could be used for data scraping. Limiting scraping to only one site will make things much easier, but there will be question of data revelance.
I think it should be even finer division. Say, groceries should be divided into diary, flour products, vegetables and so on. More common indexes will be piece of cake.
Shortly, it should be some browser plugin to daily scrap data for single or several commodity from marketplace site and send it to some database, preferrably distributed one. Then, there should be sites or local apps to show data from that database as graphs, averages, indexes and so on. Technical problems are solvable, I see how that could be done, but I hardly could imagine how to convince many people to participate in such adventure.
Yeah getting people to participate will be hard part. It would need to be accessible and somewhat easy to add data otherwise it will be too much effort. But not too easy that it inspires manipulation through the data itself.