CSS-LM

1
[{"sentiment": "FEATURE-OF", "sentence": "We propose a family of non-uniform sampling strategies to provably speed up a class of << stochastic optimization algorithms >> with [[ linear convergence ]] including Stochastic Variance Reduced Gradient -LRB- SVRG -RRB- and Stochastic Dual Coordinate Ascent -LRB- SDCA -RRB- .", "aspect": "scii"}, {"sentiment": "FEATURE-OF", "sentence": "However most of the works found in the literature have focused on identifying and understanding << temporal expressions >> in [[ newswire texts ]] .", "aspect": "scii"}, {"sentiment": "FEATURE-OF", "sentence": "Contrary to existing greedy techniques , these << tasks >> are posed in the form of a [[ discrete global optimization problem ]] with a well defined objective function .", "aspect": "scii"}, {"sentiment": "FEATURE-OF", "sentence": "A unique characteristic of the system is its ability to cope with long-duration and complete occlusion without a [[ prior knowledge ]] about the shape or << motion of objects >> .", "aspect": "scii"}, {"sentiment": "FEATURE-OF", "sentence": "In comparison with earlier work , the proposed method covers a much wider range of verb entailment types and learns the << mapping between verbs >> with [[ highly varied argument structures ]] .", "aspect": "scii"}, {"sentiment": "FEATURE-OF", "sentence": "To obtain a more complete list of MWEs we propose and use a technique exploiting the << Word Sketch Engine >> , which allows us to work with [[ statistical parameters ]] such as frequency of MWEs and their components as well as with the salience for the whole MWEs .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "<< Automatic evaluation metrics >> for Machine Translation -LRB- MT -RRB- systems , such as [[ BLEU ]] or NIST , are now well established .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "We show that adding these conditions to Gib-son 's assumptions is not sufficient to ensure global computability with one hidden layer , by exhibiting a new << non-local configuration >> , the [[ `` critical cycle '' ]] , which implies that f is not computable with one hidden layer .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "We also estimate bounds on the Bayes classification error to quantify the distinction between two classes of << HFOs >> -LRB- those occurring during seizures and [[ those ]] occurring due to other processes -RRB- .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "We tested the clustering and filtering processes on electronic newsgroup discussions , and evaluated their performance by means of two << experiments >> : coarse-level clustering simple [[ information retrieval ]] .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "The main feature of this model is to view [[ parsing ]] and generation as two strongly interleaved << tasks >> performed by a single parametrized deduction process .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "We propose a family of non-uniform sampling strategies to provably speed up a class of << stochastic optimization algorithms >> with linear convergence including [[ Stochastic Variance Reduced Gradient -LRB- SVRG -RRB- ]] and Stochastic Dual Coordinate Ascent -LRB- SDCA -RRB- .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "Experiments were done for two << ag-glutinative and morphologically rich languages >> : [[ Finnish ]] and Turk-ish .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "We show that the model trained on a certain type of << data >> , e.g. , [[ RGB and depth images ]] , generalizes well for other modalities , e.g. , Flash/Non-Flash and RGB/NIR images .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "As evidence of its usefulness and usability , it has been used successfully in a research context to uncover relationships between language and behavioral patterns in two distinct << domains >> : [[ tutorial dialogue ]] -LRB- Kumar et al. , submitted -RRB- and on-line communities -LRB- Arguello et al. , 2006 -RRB- .", "aspect": "scii"}, {"sentiment": "HYPONYM-OF", "sentence": "Extensive experiments in common << applications >> such as 2D/3D image segmentations and [[ 3D surface fitting ]] demonstrate the effectiveness of our approach .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "A Bayesian framework is used to probabilistically model : [[ people 's trajectories and intents ]] , << constraint map of the scene >> , and locations of functional objects .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "First , images are partitioned into regions using << one-class classification >> and [[ patch-based clustering algorithms ]] where one-class classifiers model the regions with relatively uniform color and texture properties , and clustering of patches aims to detect structures in the remaining regions .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "How to obtain hierarchical relations -LRB- e.g. [[ superordinate - hyponym relation ]] , << synonym relation >> -RRB- is one of the most important problems for thesaurus construction .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "Furthermore , in contrast to the approach of Dalrymple et al. -LSB- 1991 -RSB- , the treatment directly encodes the intuitive distinction between [[ full NPs ]] and the << referential elements >> that corefer with them through what we term role linking .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "We evaluate our approach on several standard datasets such as im2gps , [[ San Francisco ]] and << MediaEval2010 >> , and obtain state-of-the-art results .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "We show that the newly proposed concept-distance measures outperform traditional distributional word-distance measures in the tasks of -LRB- 1 -RRB- << ranking word pairs in order of semantic distance >> , and -LRB- 2 -RRB- [[ correcting real-word spelling errors ]] .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "The performance of SRDA is tested on [[ face recognition -LRB- singular case ]] -RRB- and << handwritten capital letter classification -LRB- nonsingular case -RRB- >> against existing algorithms .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "We explore a variety of predictive models , including [[ Na \u00a8 \u0131ve Bayes mixture models ]] and << mixtures of Markov models >> , and report empirical evidence that MINPATH finds useful shortcuts that save substantial navigational effort .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "PAKTUS supports the adaptation of the generic core to a variety of domains : JINTACCS messages , RAINFORM messages , news reports about a specific type of event , such as financial transfers or terrorist acts , etc. , by acquiring [[ sublanguage and domain-specific grammar ]] , << words >> , conceptual mappings , and discourse patterns .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "A decoding strategy that minimizes WWER based on a Minimum Bayes-Risk framework has been shown , and the reduction of errors on both [[ ASR ]] and << IR >> has been reported .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "Each of these parsing strategies exploits different types of knowledge ; and their combination provides a strong framework in which to process conjunctions , fragmentary input , and [[ ungrammatical structures ]] , as well as less << exotic , grammatically correct input >> .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "The system participated in all the tracks of the segmentation bakeoff -- PK-open , PK-closed , AS-open , [[ AS-closed ]] , << HK-open >> , HK-closed , MSR-open and MSR - closed -- and achieved the state-of-the-art performance in MSR-open , MSR-close and PK-open tracks .", "aspect": "scii"}, {"sentiment": "CONJUNCTION", "sentence": "The application of the techniques to the analysis of plant growth , to [[ ocean surface microturbulence in IR image sequences ]] , and to << sediment transport >> is demonstrated .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "The proposed << mechanism >> includes title-driven name recognition , [[ adaptive dynamic word formation ]] , identification of 2-character and 3-character Chinese names without title .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "We argue that the method is an appealing alternative - in terms of both simplicity and efficiency - to work on [[ feature selection methods ]] within << log-linear -LRB- maximum-entropy -RRB- models >> .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "The << co-occurrence pattern >> , a combination of [[ binary or local features ]] , is more discriminative than individual features and has shown its advantages in object , scene , and action recognition .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "The [[ correlations ]] are further incorporated into a << Maximum Entropy-based ranking model >> which estimates path weights from training .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "[[ Dictionary construction ]] , one of the most difficult tasks in developing a << machine translation system >> , is expensive .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "In this paper , we propose a partially-blurred-image classification and analysis framework for automatically detecting << images >> containing [[ blurred regions ]] and recognizing the blur types for those regions without needing to perform blur kernel estimation and image deblurring .", "aspect": "scii"}, {"sentiment": "PART-OF", "sentence": "We describe a new << approach >> which involves [[ clustering subcategorization frame -LRB- SCF -RRB- distributions ]] using the Information Bottleneck and nearest neighbour methods .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In comparison with earlier work , the proposed [[ method ]] covers a much wider range of verb entailment types and learns the << mapping between verbs >> with highly varied argument structures .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We trained classifiers on the speech processes extracted from the [[ alignments ]] of an << APT >> and an MPT with a canonical transcription .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The representation contains << complementary information >> to that learned from [[ supervised image datasets ]] like ImageNet .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The experimental results show that with simple and generic qualitative constraints and using only a small amount of [[ training data ]] , our << method >> can robustly and accurately estimate the BN model parameters .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The [[ parser ]] computes a compact << parse forest representation >> of the complete set of possible analyses for large treebank grammars and long input sentences .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Moreover , the models are automatically derived by << decision tree learning >> using [[ real dialogue data ]] collected by the system .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In such domains a [[ cascade of simple classifiers ]] each trained to achieve high detection rates and modest false positive rates can yield a final << detector >> with many desirable features : including high detection rates , very low false positive rates , and fast performance .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We describe a << method >> for identifying systematic patterns in translation data using [[ part-of-speech tag sequences ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Finally , we show that we can provide effective acquisition [[ techniques ]] for novel << word senses >> using a combination of online sources .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "[[ Constraints ]] provided by observed pixel colors , highlight color analysis and illumination color uniformity are employed in our << method >> to improve estimation of the underlying diffuse color .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We show that our [[ method ]] is able to produce convincing << per-object 3D reconstructions >> on one of the most challenging existing object-category detection datasets , PASCAL VOC .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Previous approaches learned << models >> based just on positions and [[ velocities ]] of the body parts while ignoring their appearance .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We go , on to describe FlexP , a [[ bottom-up pattern-matching parser ]] that we have designed and implemented to provide these << flexibilities >> for restricted natural language input to a limited-domain computer system .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Automatic image annotation is a newly developed and promising technique to provide << semantic image retrieval >> via [[ text descriptions ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "This paper presents a [[ method ]] to estimate the << sense priors of words >> drawn from a new domain , and highlights the importance of using well calibrated probabilities when performing these estimations .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Methods developed for [[ spelling correction ]] for << languages >> like English -LRB- see the review by Kukich -LRB- Kukich , 1992 -RRB- -RRB- are not readily applicable to agglutinative languages .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In this paper we introduce a modal language LT for imposing constraints on trees , and an [[ extension LT -LRB- LF -RRB- ]] for imposing << constraints on trees decorated with feature structures >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Furthermore , we introduce [[ global variables ]] in the << model >> , which can represent global properties such as translation , scale or viewpoint .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Although our experiments are focused on parsing , the [[ techniques ]] described generalize naturally to << NLP structures >> other than parse trees .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "A series of tests are described that show the power of the << error correction methodology >> when [[ stereotypic dialogue ]] occurs .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We propose a novel [[ limited-memory stochastic block BFGS update ]] for << incorporating enriched curvature information in stochastic approximation methods >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The system utilizes [[ typed feature structures ]] to control the << top-down derivation >> in a declarative way .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "This paper presents an [[ algorithm ]] for << labeling curvilinear structure >> at multiple scales in line drawings and edge images Symbolic CURVE-ELEMENT tokens residing in a spatially-indexed and scale-indexed data structure denote circular arcs fit to image data .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "[[ Optimisation of the tree ]] for << non-sequential tracking >> , which minimises the errors in temporal consistency due to both the drift and the jumps , is proposed .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We propose a << method >> of attaining such a design through a method of [[ structure-sharing ]] which avoids log -LRB- d -RRB- overheads often associated with structure-sharing of graphs without any use of costly dependency pointers .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Moreover , fluorescence 's wavelength-shifting property enables us to estimate the << shape >> of an object by applying [[ photomet-ric stereo ]] to emission-only images without suffering from specular reflection .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "To solve this problem , we are working towards the integration of [[ natural language generation ]] to augment the << interaction >>", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "To overcome this , state-of-the-art [[ structured learning methods ]] frame the << problem >> as one of large margin estimation .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "<< Finegrained CFG rules >> are automatically induced from a [[ parsed corpus ]] by training a PCFG-LA model using an EM-algorithm .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "This paper presents a [[ formal analysis ]] for a large class of words called << alternative markers >> , which includes other -LRB- than -RRB- , such -LRB- as -RRB- , and besides .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We show that [[ intrinsic image methods ]] can be used to refine an << initial , low-frequency shading estimate >> based on a global lighting reconstruction from an original texture and coarse scene geometry in order to resolve the inherent global ambiguity in shading .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We previously presented a framework for segmentation of complex scenes using multiple [[ physical hypotheses ]] for << simple image regions >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "[[ PAKTUS ]] supports the adaptation of the generic core to a variety of domains : << JINTACCS messages >> , RAINFORM messages , news reports about a specific type of event , such as financial transfers or terrorist acts , etc. , by acquiring sublanguage and domain-specific grammar , words , conceptual mappings , and discourse patterns .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "First , starting from the [[ WSJ-trained recognizer ]] , how much adaptation data -LRB- taken from the Phonebook training corpus -RRB- is necessary to achieve a reasonable << recognition >> performance in spite of the high degree of mismatch ?", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "An [[ algorithm ]] is presented for the << learning >> of such basis components .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "When << classifying high-dimensional sequence data >> , traditional methods -LRB- e.g. , [[ HMMs ]] , CRFs -RRB- may require large amounts of training data to avoid overfitting .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "First , we transform the original instances into a new data representation using << projections >> learnt from [[ pairwise constraints ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We present an image set classification algorithm based on << unsupervised clustering >> of [[ labeled training and unla-beled test data ]] where labels are only used in the stopping criterion .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "After registration , these [[ watertight models ]] with different poses are used to train a << parametric model >> in a fashion similar to the SCAPE method .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The [[ grammar ]] for this << generator >> is designed to properly generate the speaker 's intention in a telephone dialogue .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "The [[ method ]] has been successfully applied to << robust automatic speech recognition >> in reverberant environments by model selection .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Experimental results in the domain of << face detection >> show the [[ training algorithm ]] yields significant improvements in performance over conventional AdaBoost .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We propose a draft scheme of the [[ model ]] formalizing the << structure of communicative context >> in dialogue interaction .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In this paper , this selection is optimized as minimization of Bayes risk based on reward for correct information presentation and [[ penalty ]] for << redundant turns >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "To handle this problem , we propose a new [[ method ]] , called adaptive dither voting , for << robust spatial verification >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In this paper , we investigate the problem of automatically << predicting segment boundaries >> in [[ spoken multiparty dialogue ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Previous research has demonstrated the utility of << clustering >> in inducing semantic verb classes from [[ undisambiguated corpus data ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "With this simple [[ task ]] and no semantic labels , we learn a powerful << visual representation >> using a Convolutional Neural Network -LRB- CNN -RRB- .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "In our experiments , we use [[ Wallstreet Journal -LRB- WSJ -RRB- data ]] to train a << recognizer >> , which is adapted and evaluated in the Phonebook domain .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We describe three techniques for making syntactic analysis more robust -- an agenda-based scheduling parser , a [[ recovery technique ]] for << failed parses >> , and a new technique called terminal substring parsing .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Instead of performing pixel-domain super-resolution and recognition independently as two separate sequential processes , we integrate the tasks of super-resolution and recognition by directly computing a [[ maximum likelihood identity parameter vector ]] in high-resolution tensor space for << recognition >> .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "Like most existing approaches << it >> utilizes [[ clustering of word co-occurrences ]] .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "[[ HBG ]] incorporates lexical , syntactic , semantic , and structural information from the parse tree into the << disambiguation process >> in a novel way .", "aspect": "scii"}, {"sentiment": "USED-FOR", "sentence": "We derive a convex optimization problem for the task of [[ segmenting sequential data ]] , which explicitly treats presence of << outliers >> .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "In many languages , [[ morphology ]] provides better clues to a word 's category than << word order >> .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "Our results not only show that similar distinguishing speech processes were identified ; our [[ APT-based classifier ]] yielded better classification accuracy than the << MPT-based classifier >> whilst using fewer classification features .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "To validate our method , we compare [[ it ]] with the << Maximum Likelihood -LRB- ML -RRB- estimation method >> under sparse data and with the Expectation Maximization -LRB- EM -RRB- algorithm under incomplete data respectively .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "Since the significance of words differs in IR , automatic speech recognition -LRB- ASR -RRB- performance has been evaluated based on << weighted word error rate -LRB- WWER -RRB- >> , which gives a weight on errors from the viewpoint of IR , instead of [[ word error rate -LRB- WER -RRB- ]] , which treats all words uniformly .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "For this purpose , we have designed a version of [[ KL-ONE ]] which represents the epistemological level , while the new experimental language , << KL-Conc >> , represents the conceptual level .", "aspect": "scii"}, {"sentiment": "COMPARE", "sentence": "Experiments show that this [[ approach ]] is superior to a single << decision-tree classifier >> .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "We demonstrate the effectiveness of our << approach >> on several tasks involving the [[ discrimination of human gesture and motion categories ]] , as well as on a database of dynamic textures .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "With only 12 training speakers for << SI recognition >> , we achieved a 7.5 % [[ word error rate ]] on a standard grammar and test set from the DARPA Resource Management corpus .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "Finally , a cross-corpus -LRB- and cross-language -RRB- experiment reveals better [[ noise and reverberation robustness ]] for << DOCCs >> than for MFCCs .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "Through experiments with parallel corpora of a Korean and English language pairs , we show that our << paraphrasing method >> effectively extracts paraphrases with high [[ precision ]] , 94.3 % and 84.6 % respectively for Korean and English , and the translation knowledge extracted from the bilingual corpora could be generalized successfully using the paraphrases with the 12.5 % compression ratio .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "We introduce the evaluation methodology and describe how performance was assessed with this [[ methodology ]] in the most extensive empirical evaluation conducted on an << explanation system >> .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "In experiments using the [[ Penn WSJ corpus ]] , our automatically trained << model >> gave a performance of 86.6 % -LRB- F1 , sentences < 40 words -RRB- , which is comparable to that of an unlexicalized PCFG parser created using extensive manual feature selection .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "In this paper we formulate story link detection and new event detection as information retrieval task and hypothesize on the impact of [[ precision ]] and recall on both << systems >> .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "Over two distinct datasets , we find that indexing according to simple << character bigrams >> produces a [[ retrieval accuracy ]] superior to any of the tested word N-gram models .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "With performance above 97 % accuracy for [[ newspaper text ]] , << part of speech -LRB- pos -RRB- tagging >> might be considered a solved problem .", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "We also demonstrate that a subcategorization dictionary built with the system improves the [[ accuracy ]] of a << parser >> by an appreciable amount", "aspect": "scii"}, {"sentiment": "EVALUATE-FOR", "sentence": "Our << method >> achieves new state-of-the-art results on the Flickr30K and MSCOCO image-sentence datasets and shows promise on the new task of [[ phrase lo-calization ]] on the Flickr30K Entities dataset .", "aspect": "scii"}]