1. A hierarchical knowledge graph embedding framework for link predictionShuang Liu, Chengwang Hou, Jiana Meng, Peng Chen, Simon Kolmanič, 2024, izvirni znanstveni članek Ključne besede: knowledge graph embedding, knowledge graph completion, negative sampling, link prediction Objavljeno v DKUM: 04.02.2025; Ogledov: 0; Prenosov: 3
Celotno besedilo (1,76 MB) |
2. Efficient proper embedding of a daisy cubeAleksander Vesel, 2021, izvirni znanstveni članek Opis: For a set ▫$X$▫ of binary words of length ▫$h$▫ the daisy cube ▫$Q_h(X)$▫ is defined as the subgraph of the hypercube ▫$Q_h$▫ induced by the set of all vertices on shortest paths that connect vertices of ▫$X$▫ with the vertex ▫$0^h$▫. A vertex in the intersection of all of these paths is a minimal vertex of a daisy cube. A graph ▫$G$▫ isomorphic to a daisy cube admits several isometric embeddings into a hypercube. We show that an isometric embedding is proper if and only if the label ▫$0^h$▫ is assigned to a minimal vertex of ▫$G$▫. This result allows us to devise an algorithm which finds a proper embedding of a graph isomorphic to a daisy cube into a hypercube in linear time. Ključne besede: daisy cube, partial cube, isometric embedding, proper embedding Objavljeno v DKUM: 29.11.2024; Ogledov: 0; Prenosov: 3
Celotno besedilo (293,91 KB) Gradivo ima več datotek! Več... |
3. An asymptotic relation between the wirelength of an embedding and the Wiener indexK. Jagadeesh Kumar, Sandi Klavžar, R. Sundara Rajan, Indra Rajasingh, T. M. Rajalaxmi, 2021, izvirni znanstveni članek Opis: Wirelength is an important criterion to validate the quality of an embedding of a graph into a host graph and is used in particular in VLSI (Very-Large-Scale Integration) layout designs. Wiener index plays a significant role in mathematical chemistry, cheminformatics, and elsewhere. In this note these two concepts are related by proving that the Wiener index of a host graph is an upper bound for the wirelength of a given embedding. The wirelength of embedding complete ▫$2^p$▫-partite graphs into Cartesian products of paths and/or cycles as the function of the Wiener index is determined. The result is an asymptotic approximation of the general upper bound. Ključne besede: Wiener index, embedding, wirelength, complete 2p-partite graph, Cartesian product of graphs, integer labeling Objavljeno v DKUM: 23.09.2024; Ogledov: 0; Prenosov: 2
Celotno besedilo (365,63 KB) Gradivo ima več datotek! Več... |
4. Commit-level software change intent classification using a pre-trained transformer-based code modelTjaša Heričko, Boštjan Šumak, Sašo Karakatič, 2024, izvirni znanstveni članek Opis: Software evolution is driven by changes made during software development and maintenance. While source control systems effectively manage these changes at the commit level, the intent behind them are often inadequately documented, making understanding their rationale challenging. Existing commit intent classification approaches, largely reliant on commit messages, only partially capture the underlying intent, predominantly due to the messages’ inadequate content and neglect of the semantic nuances in code changes. This paper presents a novel method for extracting semantic features from commits based on modifications in the source code, where each commit is represented by one or more fine-grained conjoint code changes, e.g., file-level or hunk-level changes. To address the unstructured nature of code, the method leverages a pre-trained transformer-based code model, further trained through task-adaptive pre-training and fine-tuning on the downstream task of intent classification. This fine-tuned task-adapted pre-trained code model is then utilized to embed fine-grained conjoint changes in a commit, which are aggregated into a unified commit-level vector representation. The proposed method was evaluated using two BERT-based code models, i.e., CodeBERT and GraphCodeBERT, and various aggregation techniques on data from open-source Java software projects. The results show that the proposed method can be used to effectively extract commit embeddings as features for commit intent classification and outperform current state-of-the-art methods of code commit representation for intent categorization in terms of software maintenance activities undertaken by commits. Ključne besede: software maintenance, code commit, mining software repositories, adaptive pre-training, fine-tuning, semantic code embedding, CodeBERT, GraphCodeBERT, classification, code intelligence Objavljeno v DKUM: 14.08.2024; Ogledov: 87; Prenosov: 13
Celotno besedilo (1,65 MB) |
5. Reduction of Neural Machine Translation Failures by Incorporating Statistical Machine TranslationJani Dugonik, Mirjam Sepesy Maučec, Domen Verber, Janez Brest, 2023, izvirni znanstveni članek Opis: This paper proposes a hybrid machine translation (HMT) system that improves the quality of neural machine translation (NMT) by incorporating statistical machine translation (SMT). Therefore, two NMT systems and two SMT systems were built for the Slovenian-English language pair, each for translation in one direction. We used a multilingual language model to embed the source sentence and translations into the same vector space. From each vector, we extracted features based on the distances and similarities calculated between the source sentence and the NMT translation, and between the source sentence and the SMT translation. To select the best possible translation, we used several well-known classifiers to predict which translation system generated a better translation of the source sentence. The proposed method of combining SMT and NMT in the hybrid system is novel. Our framework is language-independent and can be applied to other languages supported by the multilingual language model. Our experiment involved empirical applications. We compared the performance of the classifiers, and the results demonstrate that our proposed HMT system achieved notable improvements in the BLEU score, with an increase of 1.5 points and 10.9 points for both translation directions, respectively. Ključne besede: neural machine translation, statistical machine translation, sentence embedding, similarity, classification, hybrid machine translation Objavljeno v DKUM: 20.02.2024; Ogledov: 322; Prenosov: 35
Celotno besedilo (400,40 KB) Gradivo ima več datotek! Več... |
6. Lattice embeddings of treesWilfried Imrich, Matjaž Kovše, 2009, izvirni znanstveni članek Opis: Predstavljen je algoritem časovno linearne zahtevnosti, ki na izometričen način vloži dano drevo ▫$T$▫ v celoštevilsko mrežo najmanjše možne dimenzije in omogoča izračun mrežnih koordinat vozlišč drevesa ▫$T$▫ v optimalnem času. Ključne besede: matematika, teorija grafov, drevo, izometrična vložitev, mrežna vložitev, delna kocka, mathematics, graph theory, lattice embedding, isometric embedding, partial cube, tree Objavljeno v DKUM: 10.07.2015; Ogledov: 1003; Prenosov: 168
Povezava na celotno besedilo |
7. Partial cubes are distance graphsMelita Gorše Pihler, Janez Žerovnik, 2008, objavljeni znanstveni prispevek na konferenci Opis: Chatrand, Kubicki in Schultz [Aequationes Math. 55 (1998) 129-145] so postavili domnevo, da so vsi dvodelni grafi razdaljni grafi. V tem članku pokažemo, da so vsi grafi podmnožice dvodelnih grafov, in sicer delne kocke, razdaljni grafi. Ključne besede: matematika, teorija grafov, razdaljni graf, delna kocka, hiperkocka, izometrični podgraf, vložitev, mathematics, graph theory, distance graph, partial cube, hypercube, isometric subgraph, embedding Objavljeno v DKUM: 10.07.2015; Ogledov: 1432; Prenosov: 159
Povezava na celotno besedilo |
8. Tribes of cubic partial cubesSandi Klavžar, Sergey Shpectorov, 2007, izvirni znanstveni članek Opis: Partial cubes are graphs isometrically embeddable into hypercubes. Three infinite families and a few sporadic examples of cubic partial cubes are known. The concept of a tribe is introduced as means to systematize the known examples and establish relations among them. Efficient methods of computation of tribes are developed and several concrete tribes, that include known, as well as new cubic partial cubes, are computed by hand and with the use of a computer. Ključne besede: mathematics, graph theory, partial cube, hypercube, isometric embedding, tribe, algorithm Objavljeno v DKUM: 10.07.2015; Ogledov: 1088; Prenosov: 182
Celotno besedilo (688,41 KB) Gradivo ima več datotek! Več... |
9. Closed embeddings into Lipscomb's universal spaceUroš Milutinović, 2007, objavljeni povzetek znanstvenega prispevka na konferenci Ključne besede: matematika, topologija, dimenzija pokrivanja, posplošena krivulja Sierpińskega, univerzalni prostor, Lipscombov univerzalni prostor, vložitev, razširitev, poln metrični prostor, zaprta vložitev, mathematics, topology, covering dimension, embedding, closed embedding, generalized Sierpiński curve, universal space, Lipscomb universal space, complete metric space, extension Objavljeno v DKUM: 10.07.2015; Ogledov: 1401; Prenosov: 29
Povezava na celotno besedilo |
10. Brešar, Boštjan: Partial Hamming graphs and expansion procedures. - Discrete Math. 237 (2001), no. 1-3, 13-27Sandi Klavžar, 2002, recenzija, prikaz knjige, kritika Ključne besede: matematika, teorija grafov, izometrična vložitev, delni Hammingovi grafi, mathematics, graph theory, isometric embedding, partial Hamming graphs Objavljeno v DKUM: 10.07.2015; Ogledov: 1929; Prenosov: 46
Povezava na celotno besedilo |