Codifying commonsense knowledge in machines is a longstanding goal of artificial intelligence. Recently, much progress toward this goal has been made with automatic knowledge base (KB) construction techniques. However, such techniques focus primarily
on the acquisition of positive (true) KB statements, even though negative (false) statements are often also important for discriminative reasoning over commonsense KBs. As a first step toward the latter, this paper proposes NegatER, a framework that ranks potential negatives in commonsense KBs using a contextual language model (LM). Importantly, as most KBs do not contain negatives, NegatER relies only on the positive knowledge in the LM and does not require ground-truth negative examples. Experiments demonstrate that, compared to multiple contrastive data augmentation approaches, NegatER yields negatives that are more grammatical, coherent, and informative---leading to statistically significant accuracy improvements in a challenging KB completion task and confirming that the positive knowledge in LMs can be re-purposed'' to generate negative knowledge.
Knowledge Base Question Answering (KBQA) is to answer natural language questions posed over knowledge bases (KBs). This paper targets at empowering the IR-based KBQA models with the ability of numerical reasoning for answering ordinal constrained que
stions. A major challenge is the lack of explicit annotations about numerical properties. To address this challenge, we propose a pretraining numerical reasoning model consisting of NumGNN and NumTransformer, guided by explicit self-supervision signals. The two modules are pretrained to encode the magnitude and ordinal properties of numbers respectively and can serve as model-agnostic plugins for any IR-based KBQA model to enhance its numerical reasoning ability. Extensive experiments on two KBQA benchmarks verify the effectiveness of our method to enhance the numerical reasoning ability for IR-based KBQA models.
Automatic construction of relevant Knowledge Bases (KBs) from text, and generation of semantically meaningful text from KBs are both long-standing goals in Machine Learning. In this paper, we present ReGen, a bidirectional generation of text and grap
h leveraging Reinforcement Learning to improve performance. Graph linearization enables us to re-frame both tasks as a sequence to sequence generation problem regardless of the generative direction, which in turn allows the use of Reinforcement Learning for sequence training where the model itself is employed as its own critic leading to Self-Critical Sequence Training (SCST). We present an extensive investigation demonstrating that the use of RL via SCST benefits graph and text generation on WebNLG+ 2020 and TekGen datasets. Our system provides state-of-the-art results on WebNLG+ 2020 by significantly improving upon published results from the WebNLG 2020+ Challenge for both text-to-graph and graph-to-text generation tasks. More details at https://github.com/IBM/regen.
We propose a multi-task, probabilistic approach to facilitate distantly supervised relation extraction by bringing closer the representations of sentences that contain the same Knowledge Base pairs. To achieve this, we bias the latent space of senten
ces via a Variational Autoencoder (VAE) that is trained jointly with a relation classifier. The latent code guides the pair representations and influences sentence reconstruction. Experimental results on two datasets created via distant supervision indicate that multi-task learning results in performance benefits. Additional exploration of employing Knowledge Base priors into theVAE reveals that the sentence space can be shifted towards that of the Knowledge Base, offering interpretability and further improving results.
Through our research we develop an Expert System called
Transformer Fault Detection and abbreviation Exformer, to help
engineers and technical's in detecting and diagnosis of oiled power
transformer faults before it going out of service. We also u
se Fuzzy
Logic in ambiguous data cases about gas ratios in transformer oil,
which require use of fuzzy rules in knowledge base of expert
system. We also discuss basis of using Artificial Neural Networks
and choose number of layers, number of neurons and suitable neural
network for power transformers faults analysis and compare.
The purpose of this research is to detect, locate, and define the blood vessels
in the arm of any person who has a problem in taking samples of blood for
laboratory testing in order to make it easier, not dangerous nor harmful.
Depending on the pr
operties of the scattering wave from the blood and the
depth of the penetration, we have calculated the frequency which is
necessary to choose the suitable transducer including the Geometric
Dimension as well as the materials which is made from (we have taken the
5MHz Doppler –CW for 1.2 cm depth, 2.37 attenuation ratio and the 8MHz
Doppler –CW for 0.74 cm depth , with the same attenuation ratio).
Depending on the velocity's variation of the blood flow throughout the
Biodynamic studies for important arteries in the upper limb, we have found
the Doppler frequency which occurs when the acoustic wave passes across
the blood red cell. We have designed a suitable electronic instrument which
includes the transmitter circuit, receiver circuit, and the output unit - Audio
graph.
The objective of this study was to model the Blood Flow into human arm’s
arteries in order to define velocity profile. All steps were based on
computational fluid dynamics .Simplified model for arm’s most important
arteries were made, while primar
y data such as length, diameter, and
velocity were collected for a healthy 40 years old, male , weight 64 Kg with
pulse rate 62 bpm ,and his arteries ranges from 1.6 to 2.6 mm by using
Doppler measurement.Bio dynami
In Artificial Intelligence field, Knowledge Engineering phase is considered
the most crucial phase of the development life cycle of the Knowledge Base
Systems [1]. In fact, Formal Logic in general and Modus Ponens specifically has
been the dominan
t tools for structuring this knowledge [3]. This led for forming
a gap between the knowledge area and the information area, which depends
structurally on the Set Theory in general and on the Relational Algebra in
particular [1]. Thus, trying to introduce a bridge to pass this gap in structuring
and treating knowledge, we have conducted a new knowledge representation
model that depends structurally on (Classical and Fuzzy) Set Theory. Then we
used it as the base for conducting an inference model that attempt, using a set
of algebraic operations and by going through a series of stages, to reach a
solution of the problem under study, in a manner very close to the one that
humans usually use in treating their knowledge, taking into consideration the
speed and accuracy as much as the problem allows.
