The quality of web sources has been traditionally evaluated using exogenous signals such as the hyperlink structure of the graph. We propose a new approach that relies on endogenous signals, namely, the correctness of factual information provided by the source. A source that has few false facts is considered to be trustworthy. The facts are automatically extracted from each source by information extraction methods commonly used to construct knowledge bases. We propose a way to distinguish errors made in the extraction process from factual errors in the web source per se, by using joint inference in a novel multi-layer probabilistic model. We call the trustworthiness score we computed Knowledge-Based Trust. On synthetic data, we show that our method can reliably compute the true trustworthiness levels of the sources. We then apply it to a database of 2.8B facts extracted from the web, and thereby estimate the trustworthiness of 119M webpages. Manual evaluation of a subset of the results confirms the effectiveness of the method.
Giorgio BertiniResearch on society, culture, art, neuroscience, cognition, critical thinking, intelligence, creativity, autopoiesis, self-organization, rhizomes, complexity, systems, networks, leadership, sustainability, thinkers, futures ++
Academic SupportThe Learning Change Project is a personal not for profit and without sponsors multidisciplinary initiative to support academic activities. Use the files freely for your Courses or Research. To prepare Reading Lists explore the Category List or Search for the topic of your interest. If you need any support, contact me.
710 Posts in this BlogFollow my Networks for recent Posts. For authors, date, publishers +metadata, view the source.
- Follow Learning Technologies of Change on WordPress.com