Commonsense Knowledge Base Completion

By Xiang Lorraine Li, Aynaz Taheri, Lifu Tu, Kevin Gimpel. 2016

180 citations recorded by the source corpus.

We enrich a curated resource of commonsense knowledge by formulating the problem as one of knowledge base completion (KBC). Most work in KBC focuses on knowledge bases like Freebase that relate entities drawn from a fixed set. However, the tuples in ConceptNet (Speer and Havasi, 2012) define relations between an unbounded set of phrases. We develop neural network models for scoring tuples on arbitrary phrases and evaluate them by their ability to distinguish true held-out tuples from false ones. We find strong performance from a bilinear model using a simple additive architecture to model phrases. We manually evaluate our trained model’s ability to assign quality scores to novel tuples, finding that it can propose tuples at the same quality level as mediumconfidence tuples from ConceptNet.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inference (2018, background): “Physical knowledge Prior work has studied learning grounded knowledge about objects and verbs: from knowledge bases (Li et al., 2016), syntax parses (Forbes and Choi, 2017), word embed-dings (Lucy and Gauthier, 2017), and images and dictionary deﬁnitions (Zellers and Choi, 2017).”
Commonsense Knowledge Mining from Pretrained Models (2019, methodology): “Several approaches have been proposed for training models for commonsense knowledge base completion (Li et al., 2016; Jastrz˛ebski et al., 2018).”
CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning (2020, background): “There are many emerging datasets for testing machine commonsense from different angles, such as commonsense extraction (Xu et al., 2018; Li et al., 2016), next situation prediction (SWAG (Zellers et al., 2018), CODAH (Chen et al., 2019), Hel-laSWAG (Zellers et al., 2019b)), cultural and social…”
COMET: Commonsense Transformers for Automatic Knowledge Graph Construction (2019, methodology): “C L ] 1 4 Ju n 20 19 to model “entities" as natural language phrases and relations as any concept that can link them (Li et al., 2016; Sap et al., 2019).”
Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Fly (2019, neutral): “To make sure the misprime is misleading, we manually remove correct primes for SQuAD and ConceptNet and automatically remove primes that are the correct filler for a different instance of the same relation for T-REx and ConceptNet.”
Lawyers are Dishonest? Quantifying Representational Harms in Commonsense Knowledge Resources (2021, methodology): “As a popular downstream application, we first consider the task of commonsense knowledge base completion which looks to automatically augment a CSKB with generated facts (Li et al., 2016).”
Negated LAMA: Birds cannot fly (2019, background): “ConceptNet (Li et al., 2016) combines 16 commonsense relationships be-tween words and/or phrases.”
SafeText: A Benchmark for Exploring Physical Safety in Language Models (2022, background): “These are framed in tasks such as knowledge base completion (Li et al., 2016), question-answering (Talmor et al., 2019), and natu-ral language inference (Zellers et al., 2019b).”
Event2Mind: Commonsense Inference on Events, Intents, and Reactions (2018, background): “This complements prior work on more general commonsense inference (Speer and Havasi, 2012; Li et al., 2016; Zhang et al., 2017), by focusing on the causal relations between events and people’s mental states, which are not well covered by most existing resources.”
TransOMCS: From Linguistic Graphs to Commonsense Knowledge (2020, methodology): “Recently, several attempts [Li et al. , 2016; Davison et al. , 2019] have been made to enrich the existing commonsense knowledge by learning to predict new relations between concepts.”
KagNet: Knowledge-Aware Graph Networks for Commonsense Reasoning (2019, unknown): “Furthermore, a knowledge-centric approach is extensible through commonsense knowledge acquisition techniques (Li et al., 2016; Xu et al., 2018).”
Connecting the Dots: A Knowledgeable Path Generator for Commonsense Question Answering (2020, methodology): “Meanwhile, the paths are of high quality since any knowledge fact with a score over 0.5 is classiﬁed as positive by Bilinear AVG , which is indeed the case on both datasets.”
A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPT (2023, background): “[165] treat text-to-KG construction as a process of knowledge graph completion (KGC), where missing terms are progressively covered by inference.”
Understanding by Understanding Not: Modeling Negation in Language Models (2021, methodology): “…“date of birth” and “place of death”; 2) T-REx, a subset of Wikidata triples with 41 relations (ElSa-har et al., 2018); 3) ConceptNet with 16 relations (Li et al., 2016); 4) SQuAD, a subset of 305 context-insensitive questions manually rephrased as cloze-style questions (Rajpurkar et al., 2016).”
ExplaGraphs: An Explanation Graph Generation Task for Structured Commonsense Reasoning (2021, background): “Commonsense Reasoning Benchmarks: A large variety of CSR tasks have been developed recently, including commonsense extraction (Li et al., 2016; Xu et al., 2018), next situation prediction (Zellers et al., 2018, 2019), cultural, social, and physical commonsense understanding (Lin et al., 2018; Sap…”
COMET-ATOMIC 2020: On Symbolic and Neural Commonsense Knowledge Graphs (2020, methodology): “In this work, we compare our new ATOMIC2020 knowledge graph to three other prominent CSKGs: ATOMIC (Sap et al. 2019), CONCEPTNET6 (Li et al. 2016), and TRANSOMCS (Zhang et al. 2020a).”
A Knowledge-Enhanced Pretraining Model for Commonsense Story Generation (2020, methodology): “For example, when our model learns car accident causes injury from ConceptNet, it will agree with car accident leads to injury and denies car accident is driven by injury if our model can identify the speciﬁc relation between the head ( car accident ) and tail ( injury ).”
When large language models meet personalization: perspectives of challenges and opportunities (2023, unknown): “The multi-modal generation has become an essential aspect of AIGC, which learns the multimodal connection and interaction, typically including vision language generation [259], text audio generation [260], text graph generation [261], text Code Generation [262].”