Event box

Text, data mining og Visual programming med Orange

Text, data mining og Visual programming med Orange In-Person

 *** For English see below *** 

Orange er et nyt, sejt, open source dataminingsprogram, der giver en bred vifte af muligheder til den digitale humanist, der er til text og data mining uden at skulle kode algoritmerne fra bunden.

I Orange foregår databehandlingen ved, at du opbygger et workflow af "widgets", som er algoritmer, som bliver forbundet af "kanaler", som sørger for udveksliningen af data mellem algoritmerne.

Humanister og andre akademikere, der ikke er vant til at scripte i f.eks. Python eller R, oplever det som en fordel, at de ikke skal programmere, men kan behandle data ved at "dragge og droppe" algoritmer.

Det kan potentielt set frigive et overskud, der kan sætte ind på at forstå, hvad de forskellige algoritmer gør. I Orange kan brugerne både benytte egne data i form af tekst eller billeder, og de kan webscrape data fra for eksempel Pudmed, The Guardian, NY Times eller Wikipedia. 
  
Programmet er udviklet ved University of Ljubljana, Slovenien, og udover sin smarte og visulelle funktionalitet er det meget veldokumenteret sammenlignet med andre open source-produkter. Dokumentationen består dels i  mange korte, informative videoer, dels i læsestof, der findes på Oranges officielle hjemmeside, hvorfra man også kan downloade programmet.

Hvis der er engelsktalene blandt de tilmeldte, bliver kruset afholdt på engelsk.

----

Orange is a new, cool, open source data mining program that provides a wide range of options for the digital humanist who is into text and data mining without having to code the algorithms from scratch.

In Orange, the data processing takes place by building a workflow of "widgets", which are algorithms which are connected by "channels", which ensure the exchange of data between the algorithms.

Humanists and other academics who are not used to scripting in e.g. Python or R, find it an advantage that they don't have to program, but can process data by  "dragging and dropping" algorithms.

It can potentially release a surplus that can be used to understand what the various algorithms do. In Orange, users can both use their own data in the form of text or images, and they can webscrape data from, for example, Pudmed, The Guardian, NY Times or Wikipedia.

The program was developed at the University of Ljubljana, Slovenia, and besides its smart and visual functionality, it is very well documented compared to other open source products. The documentation consists partly of many short, informative videos, partly of reading material that can be found on Orange's official website, from which you can also download the program.

If there are English speakers among those registered, the mug will be held in English.

Related LibGuide: KUB Datalab by Christian Knudsen

Date:
25/10/2023
Time:
13:00 - 15:00
Time Zone:
Central European Time (change)
Location:
KUB Datalab - Sdr Campus
Campus:
KUB Søndre Campus, Karen Blixens Plads 7, 2300 Kbh. S
Categories:
  Analysis     Datalab     Harvesting     Visualisation  
Registration has closed.

Event Organizer

Profile photo of Lars Kjær
Lars Kjær