| Title |
Miglotojo kontekstinio panašumo mato modelio sudarymas ir tyrimas / |
| Translation of Title |
Fuzzy context similarity measure model creation and analysis. |
| Authors |
Čenytė, Justina |
| Full Text |
|
| Pages |
56 |
| Keywords [eng] |
context ; similarity ; distance |
| Abstract [eng] |
Similarity is one of the most important aspects in the area of information retrieval and data mining. A new trend in similarity judgment assumes that similarity between two objects can not be expressed as a fixed value and depends on the context in which the similarity is measured. However no formal definition of context has yet been specified. In this thesis a new formal context definition is proposed and a method to measure contextual similarity is developed. The main idea behind this method is to extract context information from distinct groups of data. This method is tested with synthetic and real-world data sets comparing context similarity measure with distance based similarity. Results indicate that taking the similarity into account may cause significant changes in object similarity. This work shows the importance of context when measuring similarity. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2013 |
