Enrich Web Entity Schema Based on Integrated Annotation

Web integration systems (WIS) need to collect web objects belong to a specific domain from different websites effectively. Most WIS defines entity schemas beforehand by domain experts. Due to the essence of diversity and variability of web, it is hard to model the web entity comprehensively beforehand, furthermore, wrong annotations happen when align object values from different websites into the WIS. In order to avoid the limitations, we propose an integrated annotating method combining the matching strategy and machine learning technology to dynamically discover synonyms for predefined attribute labels and new attribute labels for a specified type of web entity. Experimental results using real-world data in book and job domains show that the proposed approach is effective in enriching web entity schema to enhance the performance of data collection process in a WIS.