Truth Discovery via Exploiting Implications from Multi-Source Data

Abstract

Data veracity is a grand challenge for various tasks on the Web. Since the web data sources are inherently unreliable and may provide conflicting information about the same real-world entities, truth discovery is emerging as a countermeasure of resolving the conflicts by discovering the truth, which conforms to the reality, from the multi-source data. A major challenge related to truth discovery is that different data items may have varying numbers of true values (or multi-truth), which counters the assumption of existing truth discovery methods that each data item should have exactly one true value. In this paper, we address this challenge by exploiting and leveraging the implications from multisource data. In particular, we exploit three types of implications, namely the implicit negative claims, the distribution of positive/negative claims, and the co-occurrence of values in sources' claims, to facilitate multi-truth discovery. We propose a probabilistic approach with improvement measures that incorporate the three implications in all stages of truth discovery process. In particular, incorporating the negative claims enables multi-truth discovery, considering the distribution of positive/negative claims relieves truth discovery from the impact of sources' behavioral features in the specific datasets, and considering values' co-occurrence relationship compensates the information lost from evaluating each value in the same claims individually. Experimental results on three real-world datasets demonstrate the effectiveness of our approach.

An Attention-Based Neural Network Using Human Semantic Knowledge and Its Application to Clickbait Detection

“…As a new topic of study, prior research works on clickbait detection extracted and applied hidden features from the data [2], [35], [1]. In such works, both non-textual information (e.g., images) and textual content were taken into account.…”

From Appearance to Essence

“…They measure source authority based on their links to the claimed values and estimate source reliability and value correctness based on the bipartite graph. The probabilistic graphical model based methods [41,49,50] apply probabilistic graphical models to jointly reason about source trustworthiness and value correctness. Finally, the optimization based methods [20,21] formulate the truth discovery problem as an optimization problem.…”

“…The manually collected ground truth only covers 7.91% of the objects. With truth discovery gaining growing popularity, considerable methods [5,10,14,15,20,25,29,31,35,40,41,43,44,[49][50][51] have been proposed to deal with various scenarios. Those works, however, have the common limitation that they either require labour-intensive labelling of data or use datasets with limited ground truth to conduct experiments.…”

Similarity-Aware Deep Attentive Model for Clickbait Detection

“…As an arguably new research topic, first attempts on this problem extract latent features [12,17]. For example, Chen et al [3] considered both content cues and non-text cues.…”