Articles

We propose a normalized-cut model for the problem of aligning a known hierarchical browsing structure, e.g., electronic slides of lecture recordings, with the sequential tran- scripts of the corresponding spoken docu- ments, with the aim to help index and access the latter. This model optimizes a normalized- cut graph-partitioning criterion …...

In this paper we present methods for automat- ically acquiring training examples for the task of entity extraction. Experimental evidence show that: (1) our methods compete with a current heavily supervised state-of-the-art sys- tem, within 0.04 absolute mean average pre- cision; and (2) our model significantly out- performs other supervised …...

The first step in graph-based semi-supervised classification is to construct a graph from in- put data. While the k-nearest neighbor graphs have been the de facto standard method of graph construction, this paper advocates using the less well-known mutual k-nearest neigh- bor graphs for high-dimensional natural lan- guage data. To …...

In this paper, we introduce a knowledge-based method to disambiguate biomedical acronyms using second-order co-occurrence vectors. We create these vectors using information about a long-form obtained from the Unified Medical Language System and Medline. We evaluate this method on a dataset of 18 acronyms found in biomedical text. Our method …...

Keyphrase extraction aims to select a set of terms from a document as a short summary of the document. Most methods extract keyphrases according to their statistical prop- erties in the given document. Appropriate keyphrases, however, are not always statis- tically significant or even do not appear in the given …...

We investigate the use of Semi-Supervised Learning (SSL) in opinion detection both in sparse data situations and for domain adapta- tion. We show that co-training reaches the best results in an in-domain setting with small la- beled data sets, with a maximum absolute gain of 33.5%. For domain transfer, we …...

Traditional text similarity measures consider each term similar only to itself and do not model semantic relatedness of terms. We pro- pose a novel discriminative training method that projects the raw term vectors into a com- mon, low-dimensional vector space. Our ap- proach operates by finding the optimal matrix to …...

Finding the right representation for words is critical for building accurate NLP systems when domain-specific labeled data for the task is scarce. This paper investigates lan- guage model representations, in which lan- guage models trained on unlabeled corpora are used to generate real-valued feature vec- tors for words. We investigate …...

. This paper presents first the CasEN transducer cascade to recognize French Named Entities. CasEN is implemented with the CasSys software of the Unitex plateform and is put at user free disposal (LGPL-LR license). We discuss about Named Entity typology used and we describe the cascade, before reporting its evaluation …...

. Our study applies statistical methods to French and Italian corpora to examine the phenomenon of multi-word term reduction in specialty languages. There are two kinds of re- duction: anaphoric and lexical. We show that anaphoric reduction depends on the discourse type (vulgarization, pedagogical, specialized) but is independent of both …...

. Semantic lexical resources are a mainstay of various NLP applications. However, comprehensive and reliable resources rarely exist or are often not freely available. We discuss in this paper the context of lexical resources building and the problems of evaluation. We present Wiktionary, a freely available and collaboratively built multilingual …...

Since the present-day Japanese use of voiced consonant mark had established in the Meiji Era, modern Japanese lit- erary text written in the Meiji Era of- ten lacks compulsory voiced consonant marks. This deteriorates the performance of morphological analyzers using ordi- nary dictionary. In this paper, we pro- pose an …...

Sentiment analysis is the problem of de- termining the polarity of a text with re- spect to a particular topic. For most ap- plications, however, it is not only nec- essary to derive the polarity of a text as a whole but also to extract negative and positive utterances on …...

Annotating Named Entity Recognition (NER) training corpora is a costly process but necessary for supervised NER systems. This paper presents an ap- proach to generate large-scale Chinese NER training data from an English- Chinese discourse level aligned parallel corpus. Difficulty of NER is different among languages due to their unique …...

There has been substantial recent interest in aligning mentions of named entities in unstructured texts to knowledge base de- scriptors, a task commonly called entity linking. This technology is crucial for applications in knowledge discovery and text data mining. This paper presents ex- periments in the new problem of cross- …...

We describe “treeblazing”, a method of us- ing annotations from the GENIA treebank to constrain a parse forest from an HPSG parser. Combining this with self-training, we show significant dependency score im- provements in a task of adaptation to the biomedical domain, reducing error rate by 9% compared to out-of-domain …...

This paper presents a system for Natural Language Programming using Class Se- quential Rules (CSR). The system recog- nizes a number of procedural primitives and operators. The domain of the system is presently limited to the world of num- bers and operations on numbers. We eval- uate the effectiveness of …...

Context-free grammars with latent anno- tations (PCFG-LA) have been found to be effective for parsing many languages; however, currently their lexical model may be subject to over-fitting and requires language engineering to handle out-of- vocabulary (OOV) words. Inspired by pre- vious studies that have incorporated rich features into generative models, …...

In this paper, we present a simple and effective fine-grained feature generation scheme for dependency parsing. We focus on the problem of grammar representa- tion, introducing fine-grained features by splitting various POS tags to different de- grees using HowNet hierarchical semantic knowledge. To prevent the oversplitting, we adopt a threshold-constrained …...

We describe a method for training a se- mantic role labeler for CCG in the absence of gold-standard syntax derivations. Tra- ditionally, semantic role labeling is per- formed by placing human-annotated se- mantic roles on gold-standard syntactic parses, identifying patterns in the syntax- semantics relationship, and then predict- ing roles …...