Commonsense Properties from Query Logs and Question Answering Forums

Abstract

Commonsense knowledge about object properties, human behavior and general concepts is crucial for robust AI applications. However, automatic acquisition of this knowledge is challenging because of sparseness and bias in online sources. This paper presents Quasimodo, a methodology and tool suite for distilling commonsense properties from non-standard web sources. We devise novel ways of tapping into search-engine query logs and QA forums, and combining the resulting candidate assertions with statistical cues from encyclopedias, books and image tags in a corroboration step. Unlike prior work on commonsense knowledge bases, Quasimodo focuses on salient properties that are typically associated with certain objects or concepts. Extensive evaluations, including extrinsic use-case studies, show that Quasimodo provides better coverage than state-of-the-art baselines with comparable quality.

Confidences for Commonsense Reasoning

“…We assume that the confidence in a fact or rule in our common sense knowledge base (KB in the following) typically arises from a large number of human users via crowd-sourcing like in ConceptNet [35,7], NLP-analyzed scraped text from the web like NELL [27], and/or combining different knowledge bases with weights like in [14] and [7] or assigned to the equivalence of name pairs in the vocabulary like in [28] and [19]. There is recent progress towards making knowledge bases for common sense reasoning where the relation strengths (typicality, saliency) have been empirically evaluated [7,33].…”

Inside Quasimodo

“…The extraction pipeline is detailed in the original paper [10] and is summarized in Figure 1. We recall here the main steps.…”

“…Resulting Data As of May 2020, Quasimodo is composed of approximately four million statements for over 95.000 different subjects, making it more than ten times bigger than ConceptNet. In [10], it was shown that its precision is significantly higher than Webchild, and that its recall significantly exceeds all other commonsense knowledge bases.…”

“…Each module in this workflow is itself composed of several sub-modules for performing specialized tasks. The reader can find a list of these components in the original paper [10] and in the code provided with it. 7 This extraction pipeline, however, is very dense, and it can be not easy to understand the effect of each part.…”

“…Recently, we [10] introduced the commonsense knowledge base Quasimodo. The system leverage human curiosity expressed in query logs and question answering forums to extract commonsense, and crowdsourcing experiments showed that it significantly outperformed existing resources like ConceptNet in terms of coverage.…”