Kontrollerat och okontrollerat språk En litteraturstudie i informationsåtervinning i databaser.
This thesis deals with one of the main questions within information science: The question about controlled and uncontrolled vocabulary and which of the two that are most effective when it comes to retrieve relevant information from electronic databases. The method used here is literature studies. I have described and examined five empirical studies which has been done in purpose to compare the effectiveness in retrieval with controlled and uncontrolled vocabulary, in the way they are used in retrieval of information in databases. The studies I have examined have also used recall and precision as a value to measure the effectiveness of retrieval. The studies which I have chosen is also limited to deal only with studies that have been done in bibliographic databases. They also cover three topics, humanities, social science and science and they have all been executed during the last decade. One thing that affects how effective one can retrieve information has to do with how the language is constructed. A language that has a lot of synonyms in it makes it more difficult to retrieve information, while a language that is more standardised makes retrieval more easily. Traditionally, the language that is used in science has considered being more suitable for information retrieval, because it is more standardised than the language of the humanities. In the thesis I discuss this item. Other factors that affect the result of the retrieval has to do with who seeks for the information, and what strategies that they use for that. The structure of the database is also important for the retrieval result. In the five studies that I have chosen, all these factors mentioned above have affected the result of the retrieval process.