Abstract

The subject of this study is the content description of fictional works, especially novels. Study is divided into two sections. In the first part of it the aim is to investigate the communicative process of fiction, the previous studies on the content description of fiction and the making of thesauruses for fiction. In the second part the aim is to create means for the analysis of fictional content description (based on the first part's theoretical discussi-on) and to make a general model for fiction search and retrieval system. The material for the empirical part of the study was gathered in Finnish public libraries. The aim of the empirical part of the study is to find out how the clients and the library professionals of public libraries describe novels by indexing and abstacting them - what differences they have and what is the consistency between them.

The basic theoretical approach in this study is qualitative and specifically as seen within the grounded theory. The corpus of this study was analyzed both with qualitative and quantitative methods. Qualitative methods were mainly used when analyzing the abstracts and the indexings were analyzed with the the basic statistical methods as well as with the calculated consistency values for indexers.

The main finding was that the abstracts and indexings were very unconsistent. One could also typify the abstracts in four categories: plot/thematical abstracts, cultural/ historical abstracts, abstracts that describe the reading experience and critical abstracts. Also, with the aid of statistics one could make a typical indexing string of each novel that consisted about 10-15 indexing terms which described the basic contents of each novel. In the end of the study a model for search and retrieval system for fiction is presented.

Methodologically, the triangulative approach with different kinds of methods turned out to be fruitful. Especially when studying the human behaviour, both quantitative and qualitative methods are needed. When used together they test the findings and results and thus give more validity to the final results.