Ready-made solutions for text classification?

Question

Machine learning

Required: we give the text, in response we get the categories to which the text belongs.
Example

Turbine Mazda CX-7, Mazda MPS, Mazda CX7 original.

Define categories

Transport/Parts and Accessories/Spare Parts/For Cars/Engine.

Could you tell me, please, where can I quickly create such a solution? Like Microsoft Azure Machine Learning? Then what "grows up" can then be used at home, without going anywhere - that is, locally?
Fast - means without deep immersion in the subject.
The example is taken from an ad on Avito, let's say there is a comparison between the name of the ad and the category to which the ad belongs. So I understand it is necessary for learning.

Answer 1

https://github.com/loguntsov/bayes

Answer 2

Two options (choose any):
1. Name in the search engine and parsing of all categories in which this detail is found on other sites.
2. Take the TecDoc software base.
Then, you create a neural network based on Kohonen maps and run the obtained data to create clusters: mechanoid.kiev.ua/neural-net-kohonen-clusterizatio...
After clustering all the data, you will get what you are looking for.

Answer 3

Classification of catalogs according to product ontology, including spare parts.
Detailing on spare parts is not deep, but it is suitable as a baseline, for example, for clustering texts.
datamaster.incon.ru