20.500.12556/RUL-91122 Postavitev skalabilne arhitekture za napovedovanje razpoložljivosti postaj sistema BicikeLj Scalable architecture for availability prediction of BicikeLj stations Dandanes smo priča pravi digitalni revoluciji v vseh panogah industrije, ki se ukvarjajo z mobilnostjo ali s transportom. Kot zgled lahko izpostavimo ameriški start up Uber, ki je v nekaj letih po vsem svetu digitaliziral mobilnost. Njegov celotni poslovni model sloni na podatkovnem rudarjenju, saj z metodami poskuša odgovarjati na vprašanje: "Od kod in kam bodo ljudje potovali?" ter za to potrebovali njihove storitve. Bolje kot lahko predvidijo gibanje populacije skozi neko področje, bolj uspešno in optimizirano je lahko njihovo poslovanje na trgu. V tej diplomski nalogi bomo sprva zbrali podatke o povpraševanju ljudi po mobilnosti v določenih predelih mesta in nato postavili učinkovit ter horizontalno razširljiv sistem, ki bo omogočal njihovo shranjevanje in obdelavo. Najbolj primeren za tovrstni projekt je ljubljanski sistem za izposojo koles, poimenovan BicikeLj. Vprašanje, na katerega bomo iskali odgovor, se glasi, kako zasedene bodo postaje BicikeLj ob izbranem času v prihodnosti in kako natančno znamo to napovedati iz preteklih podatkov. Recently, we are witnessing a true digital revolution in every sector of the industry, which is involved in mobility and transportation. As an example, we can expose American start-up Uber, which has drastically digitalized mobility in the recent years. Their business strategy is data driven; with the data mining methodologies they are trying to answer a question "From and where people will travel?", and for that they use their services. The better they can predict the movement of people through the area, the more successful and optimised their actions on the market are. In this bachelor thesis, first we will collect the data of movement of people through the city of Ljubljana, then we will form horizontally scalable system, which will enable us persistent storage and processing of the collected data. The most suitable system for this kind of project is bike sharing system located in Ljubljana, which operates under the brand BicikeLj. The question, on which we will try to answer, is how empty or full BicikeLj stations will be in a certain time in the future, and how precisely we can predict that based on the previous data. apache hadoop apache spark podatkovno rudarjenje apache hadoop apache spark data mining true false false Slovenski jezik Angleški jezik Diplomsko delo/naloga 2017-03-21 13:05:01 2017-03-21 13:05:03 2022-08-09 21:13:54 0000-00-00 00:00:00 2017 0 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 1970-01-01 18154 Kozelj_Klemen_-_Postavitev_skalabilne_arhitekture_za_napovedovanje_razpolozljivosti_postaj_siste.pdf Kozelj_Klemen_-_Postavitev_skalabilne_arhitekture_za_napovedovanje_razpolozljivosti_postaj_siste.pdf 1 5BD8804827BCE1760C57AC7026787231 bef7b3f2267989312cb14823b429e1d02f43e0367919a3731a2f82dc60a90e76 df3c346c-a1b1-11eb-a523-00155dcfd717 20.500.12556/rul/d3039d0a-c044-4181-9882-b0c5df3e948d https://repozitorij.uni-lj.si/Dokument.php?lang=slv&id=94736 Fakulteta za računalništvo in informatiko 0 0 0