Neural Decipherment via Minimum-Cost Flow: From Ugaritic to Linear B

Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 3146–3155 Florence, Italy, July 28 - August 2, 2019. c⃝2019 Association for Computational Linguistics 3146 Neural Decipherment via Minimum-Cost Flow: from Ugaritic to Linear B Jiaming Luo CSAIL, MIT j luo@csail.mit.edu Yuan Cao Google Brain yuancao@google.com Regina Barzilay CSAIL, MIT regina@csail.mit.edu Abstract In this paper we propose a novel neural ap- proach for automatic decipherment of lost lan- guages. To compensate for the lack of strong supervision signal, our model design is in- formed by patterns in language change doc- umented in historical linguistics. The model utilizes an expressive sequence-to-sequence model to capture character-level correspon- dences between cognates. To effectively train the model in an unsupervised manner, we in- novate the training procedure by formalizing it as a minimum-cost flow problem. When applied to the decipherment of Ugaritic, we achieve a 5.5% absolute improvement over state-of-the-art results. We also report the first automatic results in deciphering Linear B, a syllabic language related to ancient Greek, where our model correctly translates 67.3% of cognates.1 1 Introduction Decipherment is an ultimate low-resource chal- lenge for both humans and machines. The lack of parallel data and scarce quantities of ancient text complicate the adoption of neural methods that dominate modern machine translation. Even for human experts this translation scenario proved to be onerous: a typical decipherment spans over decades and requires encyclopedic domain knowledge, prohibitive manual effort and sheer luck (Robinson, 2002). Moreover, techniques ap- plied for the decipherment of one lost language are rarely reusable for another language. As a result, every significant human decipherment is consid- ered to be one of a kind, “the rarest category of achievement” (Pope, 1975). Prior work has demonstrated the feasibility of automatic decipherment. Snyder et al. (2010) 1Code and all datasets are hosted in https:// github.com/j-luo93/NeuroDecipher. translated the ancient Semitic language Ugaritic into Hebrew. Since both languages are derived from the same proto-Semitic origin, the translation involved matching their alphabets at the character level and mapping cognates at the word level. The effectiveness of their approach stemmed from its ability to incorporate expansive linguistic knowl- edge, including expected morphological corre- spondences, the nature of alphabet-level align- ment, etc. As with human decipherment, this ap- proach is highly customized for a given language pair and does not generalize to other lost lan- guages. In this paper, we introduce a neural decipher- ment algorithm that delivers strong performances across several languages with distinct linguistic characteristics. As in prior work, our input con- sists of text in a lost language and a non-parallel corpus in a known related language. The model is evaluated on the accuracy of aligning words from the lost language to their counterparts in the known language. To maintain the language-independent nature of the approach, we want to build the model around the most basic decipherment principles ap- plicable across multiple languages. These princi- ples are informed by known patterns in language change extensively documented in historical lin- guistics (Campbell, 2013). At …

Content truncated to 500 words