Events are one important type of information throughout text. Event extraction is an information ... more Events are one important type of information throughout text. Event extraction is an information extraction (IE) task that involves identifying entities and objects (mainly noun phrases) that represent important roles in events of a particular type. However, the extraction performance of current event extraction systems is limited because they mainly consider local context (mostly isolated sentences) when making each extraction decision. My research aims to improve both coverage and accuracy of event extraction performance by explicitly identifying event contexts before extracting individual facts. First, I introduce new event extraction architectures that incorporate discourse information across a document to seek out and validate pieces of event descriptions within the document. TIER is a multilayered event extraction architecture that performs text analysis at multiple granularities to progressively "zoom in" on relevant event information. LINKER is a unified discourse-guided approach that includes a structured sentence classifier to sequentially read a story and determine which sentences contain event information based on both the local and preceding contexts. Experimental results on two distinct event domains show that compared to previous event extraction systems, TIER can find more event information while maintaining a good extraction accuracy, and LINKER can further improve extraction accuracy. Finding documents that describe a specific type of event is also highly challenging because of the wide variety and ambiguity of event expressions. In this dissertation, I present the multifaceted event recognition approach that uses event defining characteristics (facets), in addition to event expressions, to effectively resolve the complexity of event descriptions. I also present a novel bootstrapping algorithm to automatically learn event expressions as well as facets of events, which requires minimal human supervision. Experimental results show that the multifaceted event recognition approach can effectively identify documents that describe a particular type of event and make event extraction systems more precise.
Findings of the Association for Computational Linguistics: EMNLP 2021, 2021
We present an actor-critic framework to induce subtopical structures in a news article for news d... more We present an actor-critic framework to induce subtopical structures in a news article for news discourse profiling. The model uses multiple critics that act according to known subtopic structures while the actor aims to outperform them. The content structures constitute sentences that represent latent subtopic boundaries. Then, we introduce a hierarchical neural network that uses the identified subtopic boundary sentences to model multi-level interaction between sentences, subtopics, and the document. Experimental results and analyses on the NewsDiscourse corpus show that the actor model learns to effectively segment a document into subtopics and improves the performance of the hierarchical model on the news discourse profiling task 1 .
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, 2021
Journalists usually organize and present the contents of a news article following a welldefined s... more Journalists usually organize and present the contents of a news article following a welldefined structure. In this paper, we propose a novel joint model for structure-based news genre classification that simultaneously identifies one of four commonly used news structures (including Inverted Pyramid and three other structures) for a news article as well as recognizes a sequence of news elements within the article that define the corresponding news structure. Experiments show that the joint model consistently outperforms its variants that perform two tasks independently, which supports our motivation that preserving the two-way dependencies and constraints between a type of news structure and its sequence of news elements enables the model to better predict both of them. Although being not perfect, the system predicted news structure type and news elements have improved the performance of text summarization when incorporated into a recent neural network system.
Proceedings of the Annual Hawaii International Conference on System Sciences, 2019
The objective of this paper is to propose and test a graph-based approach for detection of critic... more The objective of this paper is to propose and test a graph-based approach for detection of critical infrastructure disruptions in social media data in disasters. Understanding the situation and disruptive events of critical infrastructure is essential to effective disaster response and recovery of communities. The potential of social media data for situation awareness during disasters has been highlighted in recent studies. However, the application of social sensing in detecting disruptions of critical infrastructure is limited because existing approaches cannot provide complete and nonambiguous situational information about critical infrastructure. Therefore, to address this methodological gap, we developed a graph-based approach including data filtering, burst time-frame detection, content similarity and graph analysis. A case study of Hurricane Harvey in 2017 in Houston was conducted to illustrate the application of the proposed approach. The findings highlighted the temporal patterns of critical infrastructure events that occurred in disasters including disruptive events and their adverse impacts on communities. The findings also provided insights for better understanding critical infrastructure interdependencies in disasters. From the practical perspective, the proposed methodology study can improve the ability of community members, first responders and decision makers to detect and respond to infrastructure disruptions in disasters.
Identifying events of a specific type is a challenging task as events in texts are described in n... more Identifying events of a specific type is a challenging task as events in texts are described in numerous and diverse ways. Aiming to resolve high complexities of event descriptions, previous work (Huang and Riloff, 2013) proposes multi-faceted event recognition and a bootstrapping method to automatically acquire both event facet phrases and event expressions from unannotated texts. However, to ensure high quality of learned phrases, this method is constrained to only learn phrases that match certain syntactic structures. In this paper, we propose a bilingual structure projection algorithm that explores linguistic divergences between two languages (Chinese and English) and mines new phrases with new syntactic structures, which have been ignored in the previous work. Experiments show that our approach can successfully find novel event phrases and structures, e.g., phrases headed by nouns. Furthermore, the newly mined phrases are capable of recognizing additional event descriptions and...
In the wake of a polarizing election, social media is laden with hateful content. To address vari... more In the wake of a polarizing election, social media is laden with hateful content. To address various limitations of supervised hate speech classification methods including corpus bias and huge cost of annotation, we propose a weakly supervised two-path bootstrapping approach for an online hate speech detection model leveraging large-scale unlabeled data. This system significantly outperforms hate speech detection systems that are trained in a supervised manner using manually annotated data. Applying this model on a large quantity of tweets collected before, after, and on election day reveals motivations and patterns of inflammatory language.
In this paper, we describe TAMU's system submitted to the TAC KBP 2017 event nugget detection... more In this paper, we describe TAMU's system submitted to the TAC KBP 2017 event nugget detection and coreference resolution task. Our system builds on the statistical and empirical observations made on training and development data. We found that modifiers of event nuggets tend to have unique syntactic distribution. Their parts-of-speech tags and dependency relations provides them essential characteristics that are useful in identifying their span and also defining their types and realis status. We further found that the joint modeling of event span detection and realis status identification performs better than the individual models for both tasks. Our simple system designed using minimal features achieved the micro-average F1 scores of 57.72, 44.27 and 42.47 for event span detection, type identification and realis status classification tasks respectively. Also, our system achieved the CoNLL F1 score of 27.20 in event coreference resolution task.
Python type inference is challenging in practice. Due to its dynamic properties and extensive dep... more Python type inference is challenging in practice. Due to its dynamic properties and extensive dependencies on third-party libraries without type annotations, the performance of traditional static analysis techniques is limited. Although semantics in source code can help manifest intended usage for variables (thus help infer types), they are usually ignored by existing tools. In this paper, we propose PYInfer, an end-to-end learning-based type inference tool that automatically generates type annotations for Python variables. The key insight is that contextual code semantics is critical in inferring the type for a variable. For each use of a variable, we collect a few tokens within its contextual scope, and design a neural network to predict its type. One challenge is that it is difficult to collect a high-quality human-labeled training dataset for this purpose. To address this issue, we apply an existing static analyzer to generate the ground truth for variables in source code. Our m...
Focusing on the task of identifying event temporal status, we find that events directly or indire... more Focusing on the task of identifying event temporal status, we find that events directly or indirectly governing the target event in a dependency tree are most important contexts. Therefore, we extract dependency chains containing context events and use them as input in neural network models, which consistently outperform previous models using local context words as input. Visualization verifies that the dependency chain representation can effectively capture the context events which are closely related to the target event and play key roles in predicting event temporal status.
Event information is usually scattered across multiple sentences within a document. The local sen... more Event information is usually scattered across multiple sentences within a document. The local sentence-level event extractors often yield many noisy event role filler extractions in the absence of a broader view of the document-level context. Filtering spurious extractions and aggregating event information in a document remains a challenging problem. Following the observation that a document has several relevant event regions densely populated with event role fillers, we build graphs with candidate role filler extractions enriched by sentential embeddings as nodes, and use graph attention networks to identify event regions in a document and aggregate event information. We characterize edges between candidate extractions in a graph into rich vector representations to facilitate event region identification. The experimental results on two datasets of two languages show that our approach yields new state-of-the-art performance for the challenging event extraction task.
Most supervised word sense disambiguation (WSD) systems build word-specific classifiers by levera... more Most supervised word sense disambiguation (WSD) systems build word-specific classifiers by leveraging labeled data. However, when using word-specific classifiers, the sparseness of annotations leads to inferior sense disambiguation performance on less frequently seen words. To combat data sparsity, we propose to learn a single model that derives sense representations and meanwhile enforces congruence between a word instance and its right sense by using both sense-annotated data and lexical resources. The model is shared across words that allows utilizing sense correlations across words, and therefore helps to transfer common disambiguation rules from annotation-rich words to annotation-lean words. Empirical evaluation on benchmark datasets shows that the proposed shared model outperforms the equivalent classifier-based models by 1.7%, 2.5% and 3.8% in F1-score when using GloVe, ELMo and BERT word embeddings respectively.
Proceedings of the 21st ACM Symposium on Document Engineering, 2021
Multiscale design is the widely practiced use of space and scale to visually explore and articula... more Multiscale design is the widely practiced use of space and scale to visually explore and articulate relationships. Free-form web curation (FFWC) is an approach to supporting multiscale design, involving creative strategies of collecting content, assembling it to juxtapose and organize, sketching, writing, shifting perspective to navigate, and exhibiting to share and collaborate. Our long term goal is to support design students with automatic, on demand feedback. We introduce a spatial clustering technique for recognizing multiscale design characteristics-scales and clusters-in FFWC documents. We perform quantitative evaluation to establish baseline performance. We contribute to human-centered AI by advancing fundamental human aspirations, through automatic recognizers of creative design, e.g., for representing and communicating abstract ideas. We develop implications, (1) for supporting people using content recognition in creative contexts, such as design education; (2) for overcoming design fixation with human-centered AI; and (3) for recognizing multiscale design characteristics. CCS CONCEPTS • Human-centered computing; • Applied computing → Education; • Computing methodologies → Cluster analysis;
In this paper, we mainly explore the effectiveness of two kernelbased methods, the convolution tr... more In this paper, we mainly explore the effectiveness of two kernelbased methods, the convolution tree kernel and the shortest path dependency kernel, for Chinese relation extraction based on ACE 2007 corpus. For the convolution kernel, the performances of different parse tree spans involved in it for relation extraction are studied. Then, experiments with composite kernels, which are a combination of the convolution kernel and feature-based kernels are presented in order to discuss the complementary effects between tree kernel and flat kernels. For the shortest path dependency kernel, we improve it by replacing the strict same length requirement with finding the longest common subsequences between two shortest dependency paths. Experiments show kernel-based methods are effective for Chinese relation extraction.
The paper presents our work in the opinion pilot task in NTCIR6 in Chinese. In extracting opinion... more The paper presents our work in the opinion pilot task in NTCIR6 in Chinese. In extracting opinion holders, we applied Conditional Random Field (CRF) model to find the opinion holders as a sequential labeling task, while in determining the subjectivity and the polarity, we adopted a simple empirical algorithms based on the sentimental dictionary to discriminate the subjective sentences from the objective ones and suggest their polarities. Besides the features used in the CRF model and the detailed specification in the machine learning system, the evaluation results and the error analysis will also be presented.
Identifying documents that describe a specific type of event is challenging due to the high compl... more Identifying documents that describe a specific type of event is challenging due to the high complexity and variety of event descriptions. We propose a multi-faceted event recognition approach, which identifies documents about an event using event phrases as well as defining characteristics of the event. Our research focuses on civil unrest events and learns civil unrest expressions as well as phrases corresponding to potential agents and reasons for civil unrest. We present a bootstrapping algorithm that automatically acquires event phrases, agent terms, and purpose (reason) phrases from unannotated texts. We use the bootstrapped dictionaries to identify civil unrest documents and show that multi-faceted event recognition can yield high accuracy.
This paper mainly describes a Chinese named entity recognition (NER) system NER@ISCAS, which inte... more This paper mainly describes a Chinese named entity recognition (NER) system NER@ISCAS, which integrates text, partof-speech and a small-vocabularycharacter-lists feature and heristic postprocess rules for MSRA NER open track under the framework of Conditional Random Fields (CRFs) model.
This research explores the idea of inducing domain-specific semantic class taggers using only a d... more This research explores the idea of inducing domain-specific semantic class taggers using only a domain-specific text collection and seed words. The learning process begins by inducing a classifier that only has access to contextual features, forcing it to generalize beyond the seeds. The contextual classifier then labels new instances, to expand and diversify the training set. Next, a cross-category bootstrapping process simultaneously trains a suite of classifiers for multiple semantic classes. The positive instances for one class are used as negative instances for the others in an iterative bootstrapping cycle. We also explore a one-semantic-class-per-discourse heuristic, and use the classifiers to dynamically create semantic features. We evaluate our approach by inducing six semantic taggers from a collection of veterinary medicine message board posts.
RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning, 2017
Capabilities of detecting temporal relations between two events can benefit many applications. Mo... more Capabilities of detecting temporal relations between two events can benefit many applications. Most of existing temporal relation classifiers were trained in a supervised manner. Instead, we explore the observation that regular event pairs show a consistent temporal relation despite of their various contexts, and these rich contexts can be used to train a contextual temporal relation classifier, which can further recognize new temporal relation contexts and identify new regular event pairs. We focus on detecting after and before temporal relations and design a weakly supervised learning approach that extracts thousands of regular event pairs and learns a contextual temporal relation classifier simultaneously. Evaluation shows that the acquired regular event pairs are of high quality and contain rich commonsense knowledge and domain specific knowledge. In addition, the weakly supervised trained temporal relation classifier achieves comparable performance with the state-of-the-art supervised systems.
The goal of our research is to improve event extraction by learning to identify secondary role fi... more The goal of our research is to improve event extraction by learning to identify secondary role filler contexts in the absence of event keywords. We propose a multi-layered event extraction architecture that progressively "zooms in" on relevant information. Our extraction model includes a document genre classifier to recognize event narratives, two types of sentence classifiers, and noun phrase classifiers to extract role fillers. These modules are organized as a pipeline to gradually zero in on event-related information. We present results on the MUC-4 event extraction data set and show that this model performs better than previous systems.
RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning
In the wake of a polarizing election, the cyber world is laden with hate speech. Context accompan... more In the wake of a polarizing election, the cyber world is laden with hate speech. Context accompanying a hate speech text is useful for identifying hate speech, which however has been largely overlooked in existing datasets and hate speech detection models. In this paper, we provide an annotated corpus of hate speech with context information well kept. Then we propose two types of hate speech detection models that incorporate context information, a logistic regression model with context features and a neural network model with learning components for context. Our evaluation shows that both models outperform a strong baseline by around 3% to 4% in F1 score and combining these two models further improve the performance by another 7% in F1 score.
Events are one important type of information throughout text. Event extraction is an information ... more Events are one important type of information throughout text. Event extraction is an information extraction (IE) task that involves identifying entities and objects (mainly noun phrases) that represent important roles in events of a particular type. However, the extraction performance of current event extraction systems is limited because they mainly consider local context (mostly isolated sentences) when making each extraction decision. My research aims to improve both coverage and accuracy of event extraction performance by explicitly identifying event contexts before extracting individual facts. First, I introduce new event extraction architectures that incorporate discourse information across a document to seek out and validate pieces of event descriptions within the document. TIER is a multilayered event extraction architecture that performs text analysis at multiple granularities to progressively "zoom in" on relevant event information. LINKER is a unified discourse-guided approach that includes a structured sentence classifier to sequentially read a story and determine which sentences contain event information based on both the local and preceding contexts. Experimental results on two distinct event domains show that compared to previous event extraction systems, TIER can find more event information while maintaining a good extraction accuracy, and LINKER can further improve extraction accuracy. Finding documents that describe a specific type of event is also highly challenging because of the wide variety and ambiguity of event expressions. In this dissertation, I present the multifaceted event recognition approach that uses event defining characteristics (facets), in addition to event expressions, to effectively resolve the complexity of event descriptions. I also present a novel bootstrapping algorithm to automatically learn event expressions as well as facets of events, which requires minimal human supervision. Experimental results show that the multifaceted event recognition approach can effectively identify documents that describe a particular type of event and make event extraction systems more precise.
Findings of the Association for Computational Linguistics: EMNLP 2021, 2021
We present an actor-critic framework to induce subtopical structures in a news article for news d... more We present an actor-critic framework to induce subtopical structures in a news article for news discourse profiling. The model uses multiple critics that act according to known subtopic structures while the actor aims to outperform them. The content structures constitute sentences that represent latent subtopic boundaries. Then, we introduce a hierarchical neural network that uses the identified subtopic boundary sentences to model multi-level interaction between sentences, subtopics, and the document. Experimental results and analyses on the NewsDiscourse corpus show that the actor model learns to effectively segment a document into subtopics and improves the performance of the hierarchical model on the news discourse profiling task 1 .
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021, 2021
Journalists usually organize and present the contents of a news article following a welldefined s... more Journalists usually organize and present the contents of a news article following a welldefined structure. In this paper, we propose a novel joint model for structure-based news genre classification that simultaneously identifies one of four commonly used news structures (including Inverted Pyramid and three other structures) for a news article as well as recognizes a sequence of news elements within the article that define the corresponding news structure. Experiments show that the joint model consistently outperforms its variants that perform two tasks independently, which supports our motivation that preserving the two-way dependencies and constraints between a type of news structure and its sequence of news elements enables the model to better predict both of them. Although being not perfect, the system predicted news structure type and news elements have improved the performance of text summarization when incorporated into a recent neural network system.
Proceedings of the Annual Hawaii International Conference on System Sciences, 2019
The objective of this paper is to propose and test a graph-based approach for detection of critic... more The objective of this paper is to propose and test a graph-based approach for detection of critical infrastructure disruptions in social media data in disasters. Understanding the situation and disruptive events of critical infrastructure is essential to effective disaster response and recovery of communities. The potential of social media data for situation awareness during disasters has been highlighted in recent studies. However, the application of social sensing in detecting disruptions of critical infrastructure is limited because existing approaches cannot provide complete and nonambiguous situational information about critical infrastructure. Therefore, to address this methodological gap, we developed a graph-based approach including data filtering, burst time-frame detection, content similarity and graph analysis. A case study of Hurricane Harvey in 2017 in Houston was conducted to illustrate the application of the proposed approach. The findings highlighted the temporal patterns of critical infrastructure events that occurred in disasters including disruptive events and their adverse impacts on communities. The findings also provided insights for better understanding critical infrastructure interdependencies in disasters. From the practical perspective, the proposed methodology study can improve the ability of community members, first responders and decision makers to detect and respond to infrastructure disruptions in disasters.
Identifying events of a specific type is a challenging task as events in texts are described in n... more Identifying events of a specific type is a challenging task as events in texts are described in numerous and diverse ways. Aiming to resolve high complexities of event descriptions, previous work (Huang and Riloff, 2013) proposes multi-faceted event recognition and a bootstrapping method to automatically acquire both event facet phrases and event expressions from unannotated texts. However, to ensure high quality of learned phrases, this method is constrained to only learn phrases that match certain syntactic structures. In this paper, we propose a bilingual structure projection algorithm that explores linguistic divergences between two languages (Chinese and English) and mines new phrases with new syntactic structures, which have been ignored in the previous work. Experiments show that our approach can successfully find novel event phrases and structures, e.g., phrases headed by nouns. Furthermore, the newly mined phrases are capable of recognizing additional event descriptions and...
In the wake of a polarizing election, social media is laden with hateful content. To address vari... more In the wake of a polarizing election, social media is laden with hateful content. To address various limitations of supervised hate speech classification methods including corpus bias and huge cost of annotation, we propose a weakly supervised two-path bootstrapping approach for an online hate speech detection model leveraging large-scale unlabeled data. This system significantly outperforms hate speech detection systems that are trained in a supervised manner using manually annotated data. Applying this model on a large quantity of tweets collected before, after, and on election day reveals motivations and patterns of inflammatory language.
In this paper, we describe TAMU's system submitted to the TAC KBP 2017 event nugget detection... more In this paper, we describe TAMU's system submitted to the TAC KBP 2017 event nugget detection and coreference resolution task. Our system builds on the statistical and empirical observations made on training and development data. We found that modifiers of event nuggets tend to have unique syntactic distribution. Their parts-of-speech tags and dependency relations provides them essential characteristics that are useful in identifying their span and also defining their types and realis status. We further found that the joint modeling of event span detection and realis status identification performs better than the individual models for both tasks. Our simple system designed using minimal features achieved the micro-average F1 scores of 57.72, 44.27 and 42.47 for event span detection, type identification and realis status classification tasks respectively. Also, our system achieved the CoNLL F1 score of 27.20 in event coreference resolution task.
Python type inference is challenging in practice. Due to its dynamic properties and extensive dep... more Python type inference is challenging in practice. Due to its dynamic properties and extensive dependencies on third-party libraries without type annotations, the performance of traditional static analysis techniques is limited. Although semantics in source code can help manifest intended usage for variables (thus help infer types), they are usually ignored by existing tools. In this paper, we propose PYInfer, an end-to-end learning-based type inference tool that automatically generates type annotations for Python variables. The key insight is that contextual code semantics is critical in inferring the type for a variable. For each use of a variable, we collect a few tokens within its contextual scope, and design a neural network to predict its type. One challenge is that it is difficult to collect a high-quality human-labeled training dataset for this purpose. To address this issue, we apply an existing static analyzer to generate the ground truth for variables in source code. Our m...
Focusing on the task of identifying event temporal status, we find that events directly or indire... more Focusing on the task of identifying event temporal status, we find that events directly or indirectly governing the target event in a dependency tree are most important contexts. Therefore, we extract dependency chains containing context events and use them as input in neural network models, which consistently outperform previous models using local context words as input. Visualization verifies that the dependency chain representation can effectively capture the context events which are closely related to the target event and play key roles in predicting event temporal status.
Event information is usually scattered across multiple sentences within a document. The local sen... more Event information is usually scattered across multiple sentences within a document. The local sentence-level event extractors often yield many noisy event role filler extractions in the absence of a broader view of the document-level context. Filtering spurious extractions and aggregating event information in a document remains a challenging problem. Following the observation that a document has several relevant event regions densely populated with event role fillers, we build graphs with candidate role filler extractions enriched by sentential embeddings as nodes, and use graph attention networks to identify event regions in a document and aggregate event information. We characterize edges between candidate extractions in a graph into rich vector representations to facilitate event region identification. The experimental results on two datasets of two languages show that our approach yields new state-of-the-art performance for the challenging event extraction task.
Most supervised word sense disambiguation (WSD) systems build word-specific classifiers by levera... more Most supervised word sense disambiguation (WSD) systems build word-specific classifiers by leveraging labeled data. However, when using word-specific classifiers, the sparseness of annotations leads to inferior sense disambiguation performance on less frequently seen words. To combat data sparsity, we propose to learn a single model that derives sense representations and meanwhile enforces congruence between a word instance and its right sense by using both sense-annotated data and lexical resources. The model is shared across words that allows utilizing sense correlations across words, and therefore helps to transfer common disambiguation rules from annotation-rich words to annotation-lean words. Empirical evaluation on benchmark datasets shows that the proposed shared model outperforms the equivalent classifier-based models by 1.7%, 2.5% and 3.8% in F1-score when using GloVe, ELMo and BERT word embeddings respectively.
Proceedings of the 21st ACM Symposium on Document Engineering, 2021
Multiscale design is the widely practiced use of space and scale to visually explore and articula... more Multiscale design is the widely practiced use of space and scale to visually explore and articulate relationships. Free-form web curation (FFWC) is an approach to supporting multiscale design, involving creative strategies of collecting content, assembling it to juxtapose and organize, sketching, writing, shifting perspective to navigate, and exhibiting to share and collaborate. Our long term goal is to support design students with automatic, on demand feedback. We introduce a spatial clustering technique for recognizing multiscale design characteristics-scales and clusters-in FFWC documents. We perform quantitative evaluation to establish baseline performance. We contribute to human-centered AI by advancing fundamental human aspirations, through automatic recognizers of creative design, e.g., for representing and communicating abstract ideas. We develop implications, (1) for supporting people using content recognition in creative contexts, such as design education; (2) for overcoming design fixation with human-centered AI; and (3) for recognizing multiscale design characteristics. CCS CONCEPTS • Human-centered computing; • Applied computing → Education; • Computing methodologies → Cluster analysis;
In this paper, we mainly explore the effectiveness of two kernelbased methods, the convolution tr... more In this paper, we mainly explore the effectiveness of two kernelbased methods, the convolution tree kernel and the shortest path dependency kernel, for Chinese relation extraction based on ACE 2007 corpus. For the convolution kernel, the performances of different parse tree spans involved in it for relation extraction are studied. Then, experiments with composite kernels, which are a combination of the convolution kernel and feature-based kernels are presented in order to discuss the complementary effects between tree kernel and flat kernels. For the shortest path dependency kernel, we improve it by replacing the strict same length requirement with finding the longest common subsequences between two shortest dependency paths. Experiments show kernel-based methods are effective for Chinese relation extraction.
The paper presents our work in the opinion pilot task in NTCIR6 in Chinese. In extracting opinion... more The paper presents our work in the opinion pilot task in NTCIR6 in Chinese. In extracting opinion holders, we applied Conditional Random Field (CRF) model to find the opinion holders as a sequential labeling task, while in determining the subjectivity and the polarity, we adopted a simple empirical algorithms based on the sentimental dictionary to discriminate the subjective sentences from the objective ones and suggest their polarities. Besides the features used in the CRF model and the detailed specification in the machine learning system, the evaluation results and the error analysis will also be presented.
Identifying documents that describe a specific type of event is challenging due to the high compl... more Identifying documents that describe a specific type of event is challenging due to the high complexity and variety of event descriptions. We propose a multi-faceted event recognition approach, which identifies documents about an event using event phrases as well as defining characteristics of the event. Our research focuses on civil unrest events and learns civil unrest expressions as well as phrases corresponding to potential agents and reasons for civil unrest. We present a bootstrapping algorithm that automatically acquires event phrases, agent terms, and purpose (reason) phrases from unannotated texts. We use the bootstrapped dictionaries to identify civil unrest documents and show that multi-faceted event recognition can yield high accuracy.
This paper mainly describes a Chinese named entity recognition (NER) system NER@ISCAS, which inte... more This paper mainly describes a Chinese named entity recognition (NER) system NER@ISCAS, which integrates text, partof-speech and a small-vocabularycharacter-lists feature and heristic postprocess rules for MSRA NER open track under the framework of Conditional Random Fields (CRFs) model.
This research explores the idea of inducing domain-specific semantic class taggers using only a d... more This research explores the idea of inducing domain-specific semantic class taggers using only a domain-specific text collection and seed words. The learning process begins by inducing a classifier that only has access to contextual features, forcing it to generalize beyond the seeds. The contextual classifier then labels new instances, to expand and diversify the training set. Next, a cross-category bootstrapping process simultaneously trains a suite of classifiers for multiple semantic classes. The positive instances for one class are used as negative instances for the others in an iterative bootstrapping cycle. We also explore a one-semantic-class-per-discourse heuristic, and use the classifiers to dynamically create semantic features. We evaluate our approach by inducing six semantic taggers from a collection of veterinary medicine message board posts.
RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning, 2017
Capabilities of detecting temporal relations between two events can benefit many applications. Mo... more Capabilities of detecting temporal relations between two events can benefit many applications. Most of existing temporal relation classifiers were trained in a supervised manner. Instead, we explore the observation that regular event pairs show a consistent temporal relation despite of their various contexts, and these rich contexts can be used to train a contextual temporal relation classifier, which can further recognize new temporal relation contexts and identify new regular event pairs. We focus on detecting after and before temporal relations and design a weakly supervised learning approach that extracts thousands of regular event pairs and learns a contextual temporal relation classifier simultaneously. Evaluation shows that the acquired regular event pairs are of high quality and contain rich commonsense knowledge and domain specific knowledge. In addition, the weakly supervised trained temporal relation classifier achieves comparable performance with the state-of-the-art supervised systems.
The goal of our research is to improve event extraction by learning to identify secondary role fi... more The goal of our research is to improve event extraction by learning to identify secondary role filler contexts in the absence of event keywords. We propose a multi-layered event extraction architecture that progressively "zooms in" on relevant information. Our extraction model includes a document genre classifier to recognize event narratives, two types of sentence classifiers, and noun phrase classifiers to extract role fillers. These modules are organized as a pipeline to gradually zero in on event-related information. We present results on the MUC-4 event extraction data set and show that this model performs better than previous systems.
RANLP 2017 - Recent Advances in Natural Language Processing Meet Deep Learning
In the wake of a polarizing election, the cyber world is laden with hate speech. Context accompan... more In the wake of a polarizing election, the cyber world is laden with hate speech. Context accompanying a hate speech text is useful for identifying hate speech, which however has been largely overlooked in existing datasets and hate speech detection models. In this paper, we provide an annotated corpus of hate speech with context information well kept. Then we propose two types of hate speech detection models that incorporate context information, a logistic regression model with context features and a neural network model with learning components for context. Our evaluation shows that both models outperform a strong baseline by around 3% to 4% in F1 score and combining these two models further improve the performance by another 7% in F1 score.
Uploads
Papers by Ruihong Huang