Morfessor 2.0: Python Implementation and Extensions for Morfessor Baseline

By Sami Virpioja, Peter Smit, Stig-Arne Grönroos, Mikko Kurimo. 2013

240 citations recorded by the source corpus.

How researchers see this paper

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

In Natural Language Processing

Citations in the corpus

Neural Machine Translation for English–Kazakh with Morphological Segmentation and Synthetic Data (2019, methodology): “Unsupervised morphological segmentation is performed with LMVR (Ataman et al., 2017), 6 a variant of Morfessor (Virpioja et al., 2013) that allows a ﬁxed vocabulary size to be deﬁned.”
Findings of the 2018 Conference on Machine Translation (WMT18) (2018, methodology): “The main contribution is a novel cross-lingual Morfessor (Virpioja et al., 2013) segmentation using cognates extracted from the parallel data.”
Montreal Neural Machine Translation Systems for WMT’15 (2015, positive): “For Finnish (Fi), we used Morfessor 2.0 for morpheme segmentation (Virpioja et al., 2013) by using the default parameters.”
AI4Bharat-IndicNLP Corpus: Monolingual Corpora and Word Embeddings for Indic Languages (2020, unknown): “(Virpioja et al., 2013).”
Orthographic Syllable as basic unit for SMT between Related Languages (2016, methodology): “We used un-supervised morphological segmenters trained with Morfessor (Virpioja et al., 2013) for obtaining mor-pheme representations.”
Unsupervised Learning by Program Synthesis (2015, background): “These include systems that learn the phonological form of morphological rules [1, 13, 24], systems that induce morphemes from corpora [17, 25], and systems that learn the productivity of different rules [26].”
Abu-MaTran at WMT 2015 Translation Task: Morphological Segmentation and Web Crawling (2015, methodology): “We use two different algorithms: M ORFESSOR Baseline 2.0 (Virpioja et al., 2013) and F LATCAT (Gr¨onroos et al., 2014).”
A Density Ratio Approach to Language Model Fusion in End-to-End Automatic Speech Recognition (2019, unknown): “• Output classes: 10,000 sub-word “morph” units [30], input via a 512-dimensional embedding.”
Monotonic Recurrent Neural Network Transducer and Decoding Strategies (2019, methodology): “[7] Peter; Grnroos Stig-Arne; Kurimo Mikko Virpioja, Sami; Smit, “Morfessor 2.0: Python implementation and extensions for morfessor baseline,” D4 julkaistu kehittmis- tai tutkimusraportti tai -selvitys, 2013.”
Knowledge-Powered Deep Learning for Word Embedding (2014, methodology): “For each word in the Wikipedia dataset as described above, we collect corresponding morphological, syntactic, and semantic knowledge from four data sources: Morfessor [23], Longman Dictionaries2, WordNet [25], and Freebase3.”
Hybrid Morphological Segmentation for Phrase-Based Machine Translation (2016, methodology): “In order to force the Morfessor method to follow the linguistic morphs produced by Omorfi, we added some new features to the Morfessor Baseline implementation by Virpioja et al. (2013).”
The JHU Machine Translation Systems for WMT 2016 (2016, methodology): “Byte Pair Encoding (Gage, 1994) To create our training data for the Morfessor and ChipMunk experiments, we augment the original training data with a second copy that has been segmented.”
MetaMind Neural Machine Translation System for WMT 2016 (2016, methodology): “The latter techniques allow the decoder to construct words it has not previously encountered out of known characters or morphemes; we apply the subword splitting strategy using Morfessor 2.0, an unsupervised morpheme segmentation model (Virpioja et al., 2013).”
Grammar or Crammer? The Role of Morphology in Distinguishing Orthographically Similar but Semantically Unrelated Words (2025, unknown): “Our experiments confirm this scenario, where Char-gram[3-6] segmentation consistently yields higher scores for every word-pair than expected while getting similar results from wordsim task ( ρ rel = 0 .”