scholarly journals Transductive Learning for Text Classification Using Explicit Knowledge Models

2006 ◽  
pp. 223-234 ◽  
Author(s):  
Georgiana Ifrim ◽  
Gerhard Weikum
Keyword(s):  
Text Classification ◽  
Explicit Knowledge ◽  
Transductive Learning ◽  
Knowledge Models

scholarly journals TextGTL: Graph-based Transductive Learning for Semi-supervised Text Classification via Structure-Sensitive Interpolation

2021 ◽  
Author(s):  
Chen Li ◽  
Xutan Peng ◽  
Hao Peng ◽  
Jianxin Li ◽  
Lihong Wang
Keyword(s):  
Text Classification ◽  
Data Augmentation ◽  
Oriented Graph ◽  
Representation Learning ◽  
Free Text ◽  
Transductive Learning ◽  
Benchmark Datasets ◽  
Significant Performance ◽  
Performance Gains ◽  
Global And Local

Compared with traditional sequential learning models, graph-based neural networks exhibit excellent properties when encoding text, such as the capacity of capturing global and local information simultaneously. Especially in the semi-supervised scenario, propagating information along the edge can effectively alleviate the sparsity of labeled data. In this paper, beyond the existing architecture of heterogeneous word-document graphs, for the first time, we investigate how to construct lightweight non-heterogeneous graphs based on different linguistic information to better serve free text representation learning. Then, a novel semi-supervised framework for text classification that refines graph topology under theoretical guidance and shares information across different text graphs, namely Text-oriented Graph-based Transductive Learning (TextGTL), is proposed. TextGTL also performs attribute space interpolation based on dense substructure in graphs to predict low-entropy labels with high-quality feature nodes for data augmentation. To verify the effectiveness of TextGTL, we conduct extensive experiments on various benchmark datasets, observing significant performance gains over conventional heterogeneous graphs. In addition, we also design ablation studies to dive deep into the validity of components in TextTGL.

A Framework for Knowledge Models Transformation: A Step Towards Knowledge Integration and Warehousing

2019 ◽  
Vol 18 (02) ◽  
pp. 1950025
Author(s):  
Rim Ayadi ◽  
Yasser Hachaichi ◽  
Jamel Feki
Keyword(s):  
Knowledge Integration ◽  
Explicit Knowledge ◽  
Graphical Language ◽  
Intelligent Decision Support System ◽  
Transformation Rules ◽  
Knowledge Models ◽  
Intelligent Decision ◽  
Building Process ◽  
Three Layer Architecture ◽  
Intelligent Decision Support

An intelligent decision support system should based on a knowledge warehouse (KW). A KW gathers knowledge initially expressed in different formalisms and therefore heterogeneous. Consequently, the KW building process requires knowledge homogenisation. This paper deals with this main issue; it introduces a three-layer architecture for a KW; more precisely, it focuses on the first layer architecture called Knowledge Acquisition and Transformation. This layer aims to transform heterogeneous knowledge models into the MOT (Modeling with Object Types) semi-formal language [Paquette, G (2002). Knowledge and Skills Modeling: A Graphical Language for Designing and Learning. Sainte-Foy: University of Quebec Press (in French).] that we have selected as a pivot knowledge model. For this transformation step, first, we design four meta-models; one for MOT and one for each of the three explicit knowledge models, namely, decision tree, association rules and clustering. Secondly, we define 15 transformation rules that we formalise in ATL (Atlas Transformation Language). Finally, we exemplify the knowledge transformation in order to show its usefulness for the KW building process.

TRANSDUCTIVE LEARNING FOR SHORT-TEXT CLASSIFICATION PROBLEMS USING LATENT SEMANTIC INDEXING

2005 ◽  
Vol 19 (02) ◽  
pp. 143-163 ◽  
Author(s):  
SARAH ZELIKOVITZ ◽  
FINELLA MARQUEZ
Keyword(s):  
Text Classification ◽  
Latent Semantic Indexing ◽  
Training Data ◽  
Data Sets ◽  
Semantic Indexing ◽  
Classification Problems ◽  
Transductive Learning ◽  
Short Text ◽  
Series Of Experiments ◽  
Value Decomposition

This paper presents work that uses Transductive Latent Semantic Indexing (LSI) for text classification. In addition to relying on labeled training data, we improve classification accuracy by incorporating the set of test examples in the classification process. Rather than performing LSI's singular value decomposition (SVD) process solely on the training data, we instead use an expanded term-by-document matrix that includes both the labeled data as well as any available test examples. We report the performance of LSI on data sets both with and without the inclusion of the test examples, and we show that tailoring the SVD process to the test examples can be even more useful than adding additional training data. This method can be especially useful to combat possible inclusion of unrelated data in the original corpus, and to compensate for limited amounts of data. Additionally, we evaluate the vocabulary of the training and test sets and present the results of a series of experiments to illustrate how the test set is used in an advantageous way.

Deficits in Explicit Language Problem Solving Rather Than in Implicit Learning in Specific Language Impairment: Evidence From Learning an Artificial Morphological Rule

2019 ◽  
Vol 62 (10) ◽  
pp. 3790-3807 ◽  
Author(s):  
Sara Ferman ◽  
Liat Kishon-Rabin ◽  
Hila Ganot-Budaga ◽  
Avi Karni
Keyword(s):  
Language Learning ◽  
Language Impairment ◽  
Specific Language Impairment ◽  
Explicit Knowledge ◽  
Procedural Memory ◽  
Procedural Learning ◽  
Specific Language ◽  
Language Problems ◽  
Morphological Rule ◽  
Semantic Distinction

Purpose The purpose of this study was to delineate differences between children with specific language impairment (SLI), typical age–matched (TAM) children, and typical younger (TY) children in learning and mastering an undisclosed artificial morphological rule (AMR) through exposure and usage. Method Twenty-six participants (eight 10-year-old children with SLI, 8 TAM children, and ten 8-year-old TY children) were trained to master an AMR across multiple training sessions. The AMR required a phonological transformation of verbs depending on a semantic distinction: whether the preceding noun was animate or inanimate. All participants practiced the application of the AMR to repeated and new (generalization) items, via judgment and production tasks. Results The children with SLI derived significantly less benefit from practice than their peers in learning most aspects of the AMR, even exhibiting smaller gains compared to the TY group in some aspects. Children with SLI benefited less than TAM and even TY children from training to judge and produce repeated items of the AMR. Nevertheless, despite a significant disadvantage in baseline performance, the rate at which they mastered the task-specific phonological regularities was as robust as that of their peers. On the other hand, like 8-year-olds, only half of the SLI group succeeded in uncovering the nature of the AMR and, consequently, in generalizing it to new items. Conclusions Children with SLI were able to learn language aspects that rely on implicit, procedural learning, but experienced difficulties in learning aspects that relied on the explicit uncovering of the semantic principle of the AMR. The results suggest that some of the difficulties experienced by children with SLI when learning a complex language regularity cannot be accounted for by a broad, language-related, procedural memory disability. Rather, a deficit—perhaps a developmental delay in the ability to recruit and solve language problems and establish explicit knowledge regarding a language task—can better explain their difficulties in language learning.

Patterns of Implicit Learning Below the Level of Conscious Knowledge

2010 ◽  
Vol 24 (2) ◽  
pp. 91-101 ◽  
Author(s):  
Juliana Yordanova ◽  
Rolf Verleger ◽  
Ullrich Wagner ◽  
Vasil Kolev
Keyword(s):  
Implicit Learning ◽  
Explicit Knowledge ◽  
Implicit Knowledge ◽  
Knowledge Generation ◽  
Sleep Stages ◽  
Implicit Processing ◽  
Implicit Processes ◽  
Late Night ◽  
On Line ◽  
On Line Learning

The objective of the present study was to evaluate patterns of implicit processing in a task where the acquisition of explicit and implicit knowledge occurs simultaneously. The number reduction task (NRT) was used as having two levels of organization, overt and covert, where the covert level of processing is associated with implicit associative and implicit procedural learning. One aim was to compare these two types of implicit processes in the NRT when sleep was or was not introduced between initial formation of task representations and subsequent NRT processing. To assess the effects of different sleep stages, two sleep groups (early- and late-night groups) were used where initial training of the task was separated from subsequent retest by 3 h full of predominantly slow wave sleep (SWS) or rapid eye movement (REM) sleep. In two no-sleep groups, no interval was introduced between initial and subsequent NRT performance. A second aim was to evaluate the interaction between procedural and associative implicit learning in the NRT. Implicit associative learning was measured by the difference between the speed of responses that could or could not be predicted by the covert abstract regularity of the task. Implicit procedural on-line learning was measured by the practice-based increased speed of performance with time on task. Major results indicated that late-night sleep produced a substantial facilitation of implicit associations without modifying individual ability for explicit knowledge generation or for procedural on-line learning. This was evidenced by the higher rate of subjects who gained implicit knowledge of abstract task structure in the late-night group relative to the early-night and no-sleep groups. Independently of sleep, gain of implicit associative knowledge was accompanied by a relative slowing of responses to unpredictable items suggesting reciprocal interactions between associative and motor procedural processes within the implicit system. These observations provide evidence for the separability and interactions of different patterns of processing within implicit memory.

Tacit vs. Explicit Knowledge of the University's Mission

2007 ◽  
Author(s):  
Katherine J. VanGiffen ◽  
Rachel Couvrey ◽  
Erica Agustin ◽  
Diane Johnson ◽  
Ryan Smith
Keyword(s):  
Explicit Knowledge

Competenza ad agire insieme e senso di iniziativa: il tirocinio per attraversare i confini tra studio e lavoro

2020 ◽  
pp. 93-102
Author(s):  
Fabrizio d'Aniello
Keyword(s):  
Professional Growth ◽  
Capability Approach ◽  
Boundary Crossing ◽  
Methodological Framework ◽  
Knowledge Models ◽  
Internship Experience ◽  
Cultural Historical Activity ◽  
The University ◽  
And Training ◽  
The Capability Approach

The pre-eminent motivation behind this contribution lies in the intention to offer students of three-year degree course in education and training sciences and master's degree in pedagogical sciences of the University of Macerata a further support than those already existing, aimed at expanding the educational meaningfulness of the internship experience. The main criticality of such experience is connected with the difficulty in translating knowledge, models, ideas into appropriate activities. This notably refers to the conceptual and educational core of the sense of initiative and entrepreneurship and, consistently, to the skill to act. Therefore, after a deepening of the sense of initiative and entrepreneurship, followed by related pedagogical reflections based on the capability approach, the paper presents an operative proposal aimed at increasing young people's possibilities of action and supporting their personal and professional growth. With regard to this training proposal, the theoretical and methodological framework refers to the third generation cultural historical activity theory and to the tool of the boundary crossing laboratory, variant of the change laboratory

Implicit and explicit corrective feedback and feedback exposure conditions

2020 ◽  
Vol 4 (1) ◽  
pp. 3-48
Author(s):  
Takehiro Iizuka ◽  
Kimi Nakatsukasa
Keyword(s):  
Explicit Knowledge ◽  
Corrective Feedback ◽  
Significant Difference ◽  
Oral Corrective Feedback ◽  
Post Test ◽  
Picture Description ◽  
Potential Factors ◽  
Implicit And Explicit ◽  
Instructional Settings ◽  
The Impact

This exploratory study examined the impact of implicit and explicit oral corrective feedback (CF) on the development of implicit and explicit knowledge of Japanese locative particles (activity de, movement ni and location ni) for those who directly received CF and those who observed CF in the classroom. Thirty-six college students in a beginning Japanese language course received either recast (implicit), metalinguistic (explicit) or no feedback during an information-gap picture description activity, and completed a timed picture description test (implicit knowledge) and an untimed grammaticality judgement test (explicit knowledge) in a pre-test, immediate post-test and delayed post-test. The results showed that overall there was no significant difference between CF types, and that CF benefited direct and indirect recipients similarly. Potential factors that might influence the effectiveness of CF, such as instructional settings, complexity of target structures and pedagogy styles, are discussed.

Sign in / Sign up

Export Citation Format

Share Document