goutam1511 / wikifinder

Given a topic name this project finds you the most suitable parent topics for the topic, searching the Wikipedia Category Network (WCN) related to that topic by the help of a statistical approach. It helps you form an autogenerated topic tree for a given topic.

WikiFinder

Given a topic name this project finds you the most suitable parent topics for the topic, searching the Wikipedia Category Network (WCN) related to that topic by the help of a statistical approach. It helps you form an autogenerated topic tree for a given topic.

Automated topic identification is an essential component in various information retrieval and knowledge representation tasks such as automated summary generation, categorization search and document indexing. In this project, we present the WikiFinder system to automatically generate topictrees from the input text by performing hierarchical classification using the Wikipedia Category Network (WCN).

A topic hierarchy can be defined as topic tags organised in tree structure, where every parent node is a conceptual generalisation of its children. For a given input text document, the problem of automatic topic recognition involves assigning topic tags from the topic hierarchy to an output topic tree such that, the tags are the best description of the topic of the text and are organised in the same way as they are in the topic hierarchy where each parent is a conceptual generalisation of it’s children.

Again, Each parent is a conceptual generalisation of it’s children. It is this assumption on the basis of which we applied the simple statistics to calculate how probable a Wikipedia category is to be a parent node of the given topic. The main idea is the category containing the most number of subpages relevant/close to the searched input topic can be counted as a parent node of the input topic.

Please go through the setup guide to setup this project before starting. Also please correct me if I'm wrong anywhere.