LAIČR is a C++ template class library written for Linux. With LAIČR it is very easy to handle and process very large sets of text documents. It contains several modules and a web crawling and daemon framework.

In the following only a few use cases are listed that can be easily realized with LAIČR.

1. Building a full-text index of a collection of text documents

With LAIČR it is possible to build static and dynamic full-text indexes of text documents with only a few lines of code. A static full-text index is build once from scratch and does not change during its lifetime anymore. In contrast, a dynamic index additionally supports update and delete operations but requires more space.

2. Building an automatic text classification system

LAIČR contains machine learning algorithms like the Naive Bayes. With these algorithms automatic text classification systems can be realized which are trained by examples and which automatically classify documents into predefined categories.

3. Building a web crawler system

Building a high performance web crawler system with LAIČR is no problem anymore. LAIČR already comes with implementations of daemons for an html repository to store downloaded pages, url server which delivers the urls which have to be visited next and multithreaded robots with internal DNS caches.

Retrieval Solutions is a company with competence in artificial intelligence, text mining and information retrieval with an in-depth knowledge in learning algorithms, text retrieval, text processing and in analyzing documents within very large text collections. We are developing reliable algorithms and solutions which scale very well with the size of the problem.

LAIČR contains modules for the following tasks:

•	Bit encoding (e.g. elias delta, elias gamma, golomb)
•	Building full-text indexes (static and dynamic indexes)
•	Threading (e.g. threads, mutexes, conditions)
•	Network communication (e.g. tcp, socketserver)
•	External datastructures (e.g. btree and variants)
•	External string sorting
•	Clustering (e.g. k-mean)
•	Supervised learning (e.g. Naive Bayes)
•	Text processing (e.g. tokenizers, parsers)

LAIČR contains the following frameworks:

•	Web crawling framework
•	Daemon framework