Negation is a fundamental aspect of natural language, playing a critical role in communication an... more Negation is a fundamental aspect of natural language, playing a critical role in communication and comprehension. Our study assesses the negation detection performance of Generative Pretrained Transformer (GPT) models, specifically GPT-2, GPT-3, GPT-3.5, and GPT-4. We focus on the identification of negation in natural language using a zero-shot prediction approach applied to our custom xNot360 dataset. Our approach examines sentence pairs labeled to indicate whether the second sentence negates the first. Our findings expose a considerable performance disparity among the GPT models, with GPT-4 surpassing its counterparts and GPT-3.5 displaying a marked performance reduction. The overall proficiency of the GPT models in negation detection remains relatively modest, indicating that this task pushes the boundaries of their natural language understanding capabilities. We not only highlight the constraints of GPT models in handling negation but also emphasize the importance of logical reliability in high-stakes domains such as healthcare, science, and law.
We propose a new self-explainable model for Natural Language Processing (NLP) text classification... more We propose a new self-explainable model for Natural Language Processing (NLP) text classification tasks. Our approach constructs explanations concurrently with the formulation of classification predictions. To do so, we extract a rationale from the text, then use it to predict a concept of interest as the final prediction. We provide three types of explanations: 1) rationale extraction, 2) a measure of feature importance, and 3) clustering of concepts. In addition, we show how our model can be compressed without applying complicated compression techniques. We experimentally demonstrate our explainability approach on a number of well-known text classification datasets.
A mixed shop is a manufacturing infrastructure designed to process a mixture of a set of flowshop... more A mixed shop is a manufacturing infrastructure designed to process a mixture of a set of flowshop jobs and a set of open-shop jobs. Mixed shops are in general much more complex to schedule than flow-shops and open-shops, and have been studied since the 1980's. We consider the three machine proportionate mixed shop problem denoted as M 3 | prpt | C max , in which each job has equal processing times on all three machines. Koulamas and Kyparisis [European Journal of Operational Research, 243:70-74,2015] showed that the problem is solvable in polynomial time in some very special cases; for the non-solvable case, they proposed a 5/3-approximation algorithm. In this paper, we present an improved 4/3-approximation algorithm and show that this ratio of 4/3 is asymptotically tight; when the largest job is a flow-shop job, we present a fully polynomialtime approximation scheme (FPTAS). On the negative side, while the F 3 | prpt | C max problem is polynomial-time solvable, we show an interesting hardness result that adding one open-shop job to the job set makes the problem NP-hard if this open-shop job is larger than any flow-shop job. We are able to design an FPTAS for this special case too.
In this book, we review many examples of multimedia item types for testing. We also outline how g... more In this book, we review many examples of multimedia item types for testing. We also outline how games can be used to test physics concepts — discuss designing chemistry item types with interactive graphics; study how culture-specific linguistics can help inner-city kids and new immigrants learn better; suggest approaches for automatically adjusting difficulty level in interactive graphics-based questions; and propose strategies for giving partial marks for incorrect answers. We also study how to test different cognitive skills, such as music, using multimedia interfaces and also evaluate the effectiveness of our model. Methods for estimating difficulty levels of mathematical item types using Item Response Theory (IRT) will be discussed. Examples of item shells for human computer interaction and cell phones will be shown.
Invited Papers.- Nonmonotonic Reasoning in FLORA-2.- Data Integration and Answer Set Programming.... more Invited Papers.- Nonmonotonic Reasoning in FLORA-2.- Data Integration and Answer Set Programming.- Halo I: A Controlled Experiment for Large Scale Knowledge Base Development.- ASP Foundations.- Unfounded Sets for Disjunctive Logic Programs with Arbitrary Aggregates.- Loops: Relevant or Redundant?.- Approximating Answer Sets of Unitary Lifschitz-Woo Programs.- On Modular Translations and Strong Equivalence.- ASP Extensions.- Guarded Open Answer Set Programming.- External Sources of Computation for Answer Set Solvers.- Answer Sets for Propositional Theories.- Applications.- An ID-Logic Formalization of the Composition of Autonomous Databases.- On the Local Closed-World Assumption of Data-Sources.- Computing Dialectical Trees Efficiently in Possibilistic Defeasible Logic Programming.- Actions and Causations.- An Approximation of Action Theories of and Its Application to Conformant Planning.- Game-Theoretic Reasoning About Actions in Nonmonotonic Causal Theories.- Some Logical Properties of Nonmonotonic Causal Theories.- odular-?: An Elaboration Tolerant Approach to the Ramification and Qualification Problems.- Algorithms and Computation.- Platypus: A Platform for Distributed Answer Set Solving.- Solving Hard ASP Programs Efficiently.- Mode-Directed Fixed Point Computation.- Lookahead in Smodels Compared to Local Consistencies in CSP.- Foundations.- Nested Epistemic Logic Programs.- An Algebraic Account of Modularity in ID-Logic.- Default Reasoning with Preference Within Only Knowing Logic.- Semantics.- A Social Semantics for Multi-agent Systems.- Revisiting the Semantics of Interval Probabilistic Logic Programs.- Routley Semantics for Answer Sets.- The Well Supported Semantics for Multidimensional Dynamic Logic Programs.- Application Track.- Application of Smodels in Quartet Based Phylogeny Construction.- Using Answer Set Programming for a Decision Support System.- Data Integration: a Challenging ASP Application.- Abduction and Preferences in Linguistics.- Inference of Gene Relations from Microarray Data by Abduction.- System Track.- nomore: A System for Computing Preferred Answer Sets.- Integrating an Answer Set Solver into Prolog: - .- circ2dlp - Translating Circumscription into Disjunctive Logic Programming.- Pbmodels - Software to Compute Stable Models by Pseudoboolean Solvers.- KMonitor- A Tool for Monitoring Plan Execution in Action Theories.- The nomore++ System.- Smodels A - A System for Computing Answer Sets of Logic Programs with Aggregates.- A DLP System with Object-Oriented Features.- Testing Strong Equivalence of Datalog Programs - Implementation and Examples.- SELP - A System for Studying Strong Equivalence Between Logic Programs.- cmodels - SAT-Based Disjunctive Answer Set Solver.
Pacific Asia Conference on Language, Information, and Computation, Dec 1, 2009
Chinese semantic units fall into a continuum of connection tightness, ranging from very tight, no... more Chinese semantic units fall into a continuum of connection tightness, ranging from very tight, non-compositional expressions, tight compositional words, phrases, and then to loose more or less arbitrary combinations of words. We propose an approach to measure tightness connection within this continuum, based on document frequency of segmentation patterns in a reference corpus. A variety of corpora, including search engine snippets, search engine results derived from query logs, as well as standard corpora have been investigated. Our tightness ranking on 300 phrases is quite close to their manual ranking, and non-compositional compound extraction can achieve a precision as high as 94.3% on the top 1,000 4-grams extracted from the Chinese Gigaword corpus.
Dechter, R. and Y. El Fattah Topological parameters for time-space tradeoff 93–118 du Boulay, Benedict NJ Nilsson, Artificial Intelligence: A New Synthesis; T. Dean, J. Allen and Y. Aloimonos, Artificial Intelligence: Theory and Practice; D. Poole
In contrast with traditional relation extraction, which only considers a fixed set of relations, ... more In contrast with traditional relation extraction, which only considers a fixed set of relations, Open Information Extraction (Open IE) aims at extracting all types of relations from text. Because of data sparseness, Open IE systems typically ignore lexical information, and instead employ parse trees and Part-of-Speech (POS) tags. However, the same syntactic structure may correspond to different relations. In this paper, we propose to use a lexicalized tree kernel based on the word embeddings created by a neural network model. We show that the lexicalized tree kernel model surpasses the unlexicalized model. Experiments on three datasets indicate that our Open IE system performs better on the task of relation extraction than the stateof-the-art Open IE systems of Xu et al. (2013) and Mesquita et al. (2013).
From association to meaning via correspondence connections: requirements and challenges for creating knowledge from behaviour
The problem of transforming physical behaviour into symbolic knowledge is a kind of \arti cial in... more The problem of transforming physical behaviour into symbolic knowledge is a kind of \arti cial intelligence complete" problem in the sense that almost all AI research ever done seems to be somewhere relevant to the most general conception of the problem. The hypothesis sketched here claims that if any general computational framework for this problem is to emerge, it will necessarily involve the capture and deployment of domain knowledge at the information boundaries we will call correspondence connections, after Brian Smith's idea called the \correspondence continuum." In this brief position paper, we scramble through a few central ideas that provide a basis for describing the idea of correspondence connections, and try to explain the requirements and challenges for the development and exploitation of such a framework. [D R A F T { comments welcome] 1
A Constraint Optimization Framework for MultiAgent Anytime Scheduling
Lecture Notes in Computer Science, 1999
Dynamic real-time scheduling applications are characterized by rapidly changing requirements and ... more Dynamic real-time scheduling applications are characterized by rapidly changing requirements and the need for timely response. We employ an approach combining Constraint Programming and Anytime Algorithms within a Multiagent framework. Agents implement various ...
In this paper, we present a minimal connection model of abductive diagnostic reasoning. The domai... more In this paper, we present a minimal connection model of abductive diagnostic reasoning. The domain knowledge is represented by a causal network. An explanation of a set of observations is a chain of causation events. These causation events constitute a scenario where all the observations can be observed. We define the best explanation to be the most probable explanation. The underlying causal model enables us to compute the probabilities of explanations from the conditional probabilities of the participating causation events. We then present an algorithm for finding the most probable explanations. Although probabilistic inference using belief networks is NP-hard in general, our algorithm is polynomial to the number of nodes in the networks and is exponential only to the number of observations to be explained, which, in any single case, is usually small.
Tries and patricia tries are two popular data structures for storing strings. Let H n denote the ... more Tries and patricia tries are two popular data structures for storing strings. Let H n denote the height of the trie (the patricia trie, respectively) on a set of n strings. Under the uniform distribution model on the strings, it is well known that H n / log n → 2 for tries and H n / log n → 1 for patricia tries, when n approaches infinity. Nevertheless, in the worst case, the height of a trie can be unbounded and the height of a patricia trie is in (n). To better understand the practical performance of both tries and patricia tries, we investigate these two classical data structures in a smoothed analysis model. Given a set S = {s 1 , s 2 ,. .. , s n } of n binary strings, we perturb the set by adding an i.i.d. Bernoulli random noise to each bit of every string. We show that the resulting smoothed heights of the trie and the patricia trie are both in (log n).
The development of Question Answering (QA) systems has become important because it reveals resear... more The development of Question Answering (QA) systems has become important because it reveals research issues that require insight from a variety of disciplines, including Artificial Intelligence, Information Extraction, Natural Language Processing, and Psychology. Our goal here is to develop a QA approach to answer yes/no questions relevant to civil laws in legal bar exams. A bar examination is intended to determine whether a candidate is qualified to practice law in a given jurisdiction. We have found that the development of a QA system for this task provides insight into the challenges of formalizing reasoning about legal text, and about how to exploit advances in computational linguistics. We separate our QA approach into two steps. The first step is to identify legal documents relevant to the exam questions; the second step is to answer the questions by analyzing the relevant documents. In our initial approach described here, the first step has been already solved for us: the appropriate articles for each question have been identified by legal experts. So here, we focus on the second task, which can be considered as a form of Recognizing Textual Entailment (RTE), where input to the system is a question sentence and its corresponding civil law article(s), and the output is a binary answer: whether the question sentence is entailed from the article(s). We propose a hybrid method, which combines simple rules and an unsupervised learning model using deep linguistic features. We first construct a knowledge base for negation and antonym words for the legal domain. We then identify potential premise and conclusion components of input questions and documents, based on text patterns and separating commas. We further classify the questions into easy and difficult ones, and develop a two-phase method for answering yes/no questions. We answer easy questions by negation/antonym detection. For more difficult questions, we adapt an unsupervised machine learning method based on morphological, syntactic, and lexical semantic analysis on identified premises and conclusions. This provides the basis to compare the semantic correlation between a question and a legal article. Our experimental results show reasonable performance, which improves the baseline system, and outperforms an SVM-based supervised machine learning model.
In many practical situations, batching of similar jobs to avoid setups is performed while constru... more In many practical situations, batching of similar jobs to avoid setups is performed while constructing a schedule. This paper addresses the problem of non-preemptively scheduling independent jobs in a two-machine flow shop with the objective of minimizing the makespan. Jobs are grouped into batches. A sequence independent batch setup time on each machine is required before the first job is processed, and when a machine switches from processing a job in some batch to a job of another batch. Besides its practical interest, this problem is a direct generalization of the classical two-machine flow shop problem with no grouping of jobs, which can be solved optimally by Johnson's well-known algorithm. The problem under investigation is known to be NP-hard. We propose two O(n log n) time heuristic algorithms. The first heuristic, which creates a schedule with minimum total setup time by forcing all jobs in the same batch to be sequenced in adjacent positions, has a worst-case performance ratio of 3/2. By allowing each batch to be split into at most two sub-batches, a second heuristic is developed which has an improved worst-case performance ratio of 4/3.
Most existing formalizations treat belief change as a single-step process, and ignore several pro... more Most existing formalizations treat belief change as a single-step process, and ignore several problems that become important when a theory, or belief state, is revised over several steps. This paper identifies these problems, and argues for the need to retain all of the multiple possible outcomes of a belief change step, and for a framework in which the effects of a belief change step persist as long as is consistently possible. To demonstrate that such a formalization is indeed possible, we develop a framework, which uses the language of PJ-default logic (Delgrande and Jackson 1991) to represent a belief state, and which enables the effects of a belief change step to persist by propagating belief constraints. Belief change in this framework maps one belief state to another, where each belief state is a collection of theories given by the set of extensions of the PJ-default theory representing that belief state. Belief constraints do not need to be separately recorded; they are encoded as clearly identifiable components of a PJ-default theory. The framework meets the requirements for iterated belief change that we identify and satisfies most of the AGM postulates (Alchourrón, Gärdenfors, and Makinson 1985) as well.
The success of statistical machine learning from big data, especially of deep learning, has made ... more The success of statistical machine learning from big data, especially of deep learning, has made artificial intelligence (AI) very popular. Unfortunately, especially with the most successful methods, the results are very difficult to comprehend by human experts. The application of AI in areas that impact human life (e.g., agriculture, climate, forestry, health, etc.) has therefore led to an demand for trust, which can be fostered if the methods can be interpreted and thus explained to humans. The research field of explainable artificial intelligence (XAI) provides the necessary foundations and methods. Historically, XAI has focused on the development of methods to explain the decisions and internal mechanisms of complex AI systems, with much initial research concentrating on explaining how convolutional neural networks produce image classification predictions by producing visualizations which highlight what input patterns are most influential in activating hidden units, or are most ...
We describe the development of a model to detect user-level clinical depression based on a user's... more We describe the development of a model to detect user-level clinical depression based on a user's temporal social media posts. Our model uses a Depression Symptoms Detection (DSD) model, which is trained on the largest existing samples of clinician annotated tweets for clinical depression symptoms. We subsequently use our DSD model to extract clinically relevant features, e.g., depression scores and their consequent temporal patterns, as well as user posting activity patterns, e.g., quantifying their "no activity" or "silence." Furthermore, to evaluate the efficacy of these extracted features, we create three kinds of datasets including a test dataset, from two existing well-known benchmark datasets for user-level depression detection. We then provide accuracy measures based on single features, baseline features and feature ablation tests, at several different levels of temporal granularity, data distributions, and clinical depression detection related settings to draw a complete picture of the impact of these features across our created datasets. Finally, we show that, in general, only semantic oriented representation models perform well. However, clinical features may enhance overall performance provided that the training and testing distribution is similar, and there is more data in a user's timeline. Further, we show 1
Negation is a fundamental aspect of natural language, playing a critical role in communication an... more Negation is a fundamental aspect of natural language, playing a critical role in communication and comprehension. Our study assesses the negation detection performance of Generative Pretrained Transformer (GPT) models, specifically GPT-2, GPT-3, GPT-3.5, and GPT-4. We focus on the identification of negation in natural language using a zero-shot prediction approach applied to our custom xNot360 dataset. Our approach examines sentence pairs labeled to indicate whether the second sentence negates the first. Our findings expose a considerable performance disparity among the GPT models, with GPT-4 surpassing its counterparts and GPT-3.5 displaying a marked performance reduction. The overall proficiency of the GPT models in negation detection remains relatively modest, indicating that this task pushes the boundaries of their natural language understanding capabilities. We not only highlight the constraints of GPT models in handling negation but also emphasize the importance of logical reliability in high-stakes domains such as healthcare, science, and law.
We propose a new self-explainable model for Natural Language Processing (NLP) text classification... more We propose a new self-explainable model for Natural Language Processing (NLP) text classification tasks. Our approach constructs explanations concurrently with the formulation of classification predictions. To do so, we extract a rationale from the text, then use it to predict a concept of interest as the final prediction. We provide three types of explanations: 1) rationale extraction, 2) a measure of feature importance, and 3) clustering of concepts. In addition, we show how our model can be compressed without applying complicated compression techniques. We experimentally demonstrate our explainability approach on a number of well-known text classification datasets.
A mixed shop is a manufacturing infrastructure designed to process a mixture of a set of flowshop... more A mixed shop is a manufacturing infrastructure designed to process a mixture of a set of flowshop jobs and a set of open-shop jobs. Mixed shops are in general much more complex to schedule than flow-shops and open-shops, and have been studied since the 1980's. We consider the three machine proportionate mixed shop problem denoted as M 3 | prpt | C max , in which each job has equal processing times on all three machines. Koulamas and Kyparisis [European Journal of Operational Research, 243:70-74,2015] showed that the problem is solvable in polynomial time in some very special cases; for the non-solvable case, they proposed a 5/3-approximation algorithm. In this paper, we present an improved 4/3-approximation algorithm and show that this ratio of 4/3 is asymptotically tight; when the largest job is a flow-shop job, we present a fully polynomialtime approximation scheme (FPTAS). On the negative side, while the F 3 | prpt | C max problem is polynomial-time solvable, we show an interesting hardness result that adding one open-shop job to the job set makes the problem NP-hard if this open-shop job is larger than any flow-shop job. We are able to design an FPTAS for this special case too.
In this book, we review many examples of multimedia item types for testing. We also outline how g... more In this book, we review many examples of multimedia item types for testing. We also outline how games can be used to test physics concepts — discuss designing chemistry item types with interactive graphics; study how culture-specific linguistics can help inner-city kids and new immigrants learn better; suggest approaches for automatically adjusting difficulty level in interactive graphics-based questions; and propose strategies for giving partial marks for incorrect answers. We also study how to test different cognitive skills, such as music, using multimedia interfaces and also evaluate the effectiveness of our model. Methods for estimating difficulty levels of mathematical item types using Item Response Theory (IRT) will be discussed. Examples of item shells for human computer interaction and cell phones will be shown.
Invited Papers.- Nonmonotonic Reasoning in FLORA-2.- Data Integration and Answer Set Programming.... more Invited Papers.- Nonmonotonic Reasoning in FLORA-2.- Data Integration and Answer Set Programming.- Halo I: A Controlled Experiment for Large Scale Knowledge Base Development.- ASP Foundations.- Unfounded Sets for Disjunctive Logic Programs with Arbitrary Aggregates.- Loops: Relevant or Redundant?.- Approximating Answer Sets of Unitary Lifschitz-Woo Programs.- On Modular Translations and Strong Equivalence.- ASP Extensions.- Guarded Open Answer Set Programming.- External Sources of Computation for Answer Set Solvers.- Answer Sets for Propositional Theories.- Applications.- An ID-Logic Formalization of the Composition of Autonomous Databases.- On the Local Closed-World Assumption of Data-Sources.- Computing Dialectical Trees Efficiently in Possibilistic Defeasible Logic Programming.- Actions and Causations.- An Approximation of Action Theories of and Its Application to Conformant Planning.- Game-Theoretic Reasoning About Actions in Nonmonotonic Causal Theories.- Some Logical Properties of Nonmonotonic Causal Theories.- odular-?: An Elaboration Tolerant Approach to the Ramification and Qualification Problems.- Algorithms and Computation.- Platypus: A Platform for Distributed Answer Set Solving.- Solving Hard ASP Programs Efficiently.- Mode-Directed Fixed Point Computation.- Lookahead in Smodels Compared to Local Consistencies in CSP.- Foundations.- Nested Epistemic Logic Programs.- An Algebraic Account of Modularity in ID-Logic.- Default Reasoning with Preference Within Only Knowing Logic.- Semantics.- A Social Semantics for Multi-agent Systems.- Revisiting the Semantics of Interval Probabilistic Logic Programs.- Routley Semantics for Answer Sets.- The Well Supported Semantics for Multidimensional Dynamic Logic Programs.- Application Track.- Application of Smodels in Quartet Based Phylogeny Construction.- Using Answer Set Programming for a Decision Support System.- Data Integration: a Challenging ASP Application.- Abduction and Preferences in Linguistics.- Inference of Gene Relations from Microarray Data by Abduction.- System Track.- nomore: A System for Computing Preferred Answer Sets.- Integrating an Answer Set Solver into Prolog: - .- circ2dlp - Translating Circumscription into Disjunctive Logic Programming.- Pbmodels - Software to Compute Stable Models by Pseudoboolean Solvers.- KMonitor- A Tool for Monitoring Plan Execution in Action Theories.- The nomore++ System.- Smodels A - A System for Computing Answer Sets of Logic Programs with Aggregates.- A DLP System with Object-Oriented Features.- Testing Strong Equivalence of Datalog Programs - Implementation and Examples.- SELP - A System for Studying Strong Equivalence Between Logic Programs.- cmodels - SAT-Based Disjunctive Answer Set Solver.
Pacific Asia Conference on Language, Information, and Computation, Dec 1, 2009
Chinese semantic units fall into a continuum of connection tightness, ranging from very tight, no... more Chinese semantic units fall into a continuum of connection tightness, ranging from very tight, non-compositional expressions, tight compositional words, phrases, and then to loose more or less arbitrary combinations of words. We propose an approach to measure tightness connection within this continuum, based on document frequency of segmentation patterns in a reference corpus. A variety of corpora, including search engine snippets, search engine results derived from query logs, as well as standard corpora have been investigated. Our tightness ranking on 300 phrases is quite close to their manual ranking, and non-compositional compound extraction can achieve a precision as high as 94.3% on the top 1,000 4-grams extracted from the Chinese Gigaword corpus.
Dechter, R. and Y. El Fattah Topological parameters for time-space tradeoff 93–118 du Boulay, Benedict NJ Nilsson, Artificial Intelligence: A New Synthesis; T. Dean, J. Allen and Y. Aloimonos, Artificial Intelligence: Theory and Practice; D. Poole
In contrast with traditional relation extraction, which only considers a fixed set of relations, ... more In contrast with traditional relation extraction, which only considers a fixed set of relations, Open Information Extraction (Open IE) aims at extracting all types of relations from text. Because of data sparseness, Open IE systems typically ignore lexical information, and instead employ parse trees and Part-of-Speech (POS) tags. However, the same syntactic structure may correspond to different relations. In this paper, we propose to use a lexicalized tree kernel based on the word embeddings created by a neural network model. We show that the lexicalized tree kernel model surpasses the unlexicalized model. Experiments on three datasets indicate that our Open IE system performs better on the task of relation extraction than the stateof-the-art Open IE systems of Xu et al. (2013) and Mesquita et al. (2013).
From association to meaning via correspondence connections: requirements and challenges for creating knowledge from behaviour
The problem of transforming physical behaviour into symbolic knowledge is a kind of \arti cial in... more The problem of transforming physical behaviour into symbolic knowledge is a kind of \arti cial intelligence complete" problem in the sense that almost all AI research ever done seems to be somewhere relevant to the most general conception of the problem. The hypothesis sketched here claims that if any general computational framework for this problem is to emerge, it will necessarily involve the capture and deployment of domain knowledge at the information boundaries we will call correspondence connections, after Brian Smith's idea called the \correspondence continuum." In this brief position paper, we scramble through a few central ideas that provide a basis for describing the idea of correspondence connections, and try to explain the requirements and challenges for the development and exploitation of such a framework. [D R A F T { comments welcome] 1
A Constraint Optimization Framework for MultiAgent Anytime Scheduling
Lecture Notes in Computer Science, 1999
Dynamic real-time scheduling applications are characterized by rapidly changing requirements and ... more Dynamic real-time scheduling applications are characterized by rapidly changing requirements and the need for timely response. We employ an approach combining Constraint Programming and Anytime Algorithms within a Multiagent framework. Agents implement various ...
In this paper, we present a minimal connection model of abductive diagnostic reasoning. The domai... more In this paper, we present a minimal connection model of abductive diagnostic reasoning. The domain knowledge is represented by a causal network. An explanation of a set of observations is a chain of causation events. These causation events constitute a scenario where all the observations can be observed. We define the best explanation to be the most probable explanation. The underlying causal model enables us to compute the probabilities of explanations from the conditional probabilities of the participating causation events. We then present an algorithm for finding the most probable explanations. Although probabilistic inference using belief networks is NP-hard in general, our algorithm is polynomial to the number of nodes in the networks and is exponential only to the number of observations to be explained, which, in any single case, is usually small.
Tries and patricia tries are two popular data structures for storing strings. Let H n denote the ... more Tries and patricia tries are two popular data structures for storing strings. Let H n denote the height of the trie (the patricia trie, respectively) on a set of n strings. Under the uniform distribution model on the strings, it is well known that H n / log n → 2 for tries and H n / log n → 1 for patricia tries, when n approaches infinity. Nevertheless, in the worst case, the height of a trie can be unbounded and the height of a patricia trie is in (n). To better understand the practical performance of both tries and patricia tries, we investigate these two classical data structures in a smoothed analysis model. Given a set S = {s 1 , s 2 ,. .. , s n } of n binary strings, we perturb the set by adding an i.i.d. Bernoulli random noise to each bit of every string. We show that the resulting smoothed heights of the trie and the patricia trie are both in (log n).
The development of Question Answering (QA) systems has become important because it reveals resear... more The development of Question Answering (QA) systems has become important because it reveals research issues that require insight from a variety of disciplines, including Artificial Intelligence, Information Extraction, Natural Language Processing, and Psychology. Our goal here is to develop a QA approach to answer yes/no questions relevant to civil laws in legal bar exams. A bar examination is intended to determine whether a candidate is qualified to practice law in a given jurisdiction. We have found that the development of a QA system for this task provides insight into the challenges of formalizing reasoning about legal text, and about how to exploit advances in computational linguistics. We separate our QA approach into two steps. The first step is to identify legal documents relevant to the exam questions; the second step is to answer the questions by analyzing the relevant documents. In our initial approach described here, the first step has been already solved for us: the appropriate articles for each question have been identified by legal experts. So here, we focus on the second task, which can be considered as a form of Recognizing Textual Entailment (RTE), where input to the system is a question sentence and its corresponding civil law article(s), and the output is a binary answer: whether the question sentence is entailed from the article(s). We propose a hybrid method, which combines simple rules and an unsupervised learning model using deep linguistic features. We first construct a knowledge base for negation and antonym words for the legal domain. We then identify potential premise and conclusion components of input questions and documents, based on text patterns and separating commas. We further classify the questions into easy and difficult ones, and develop a two-phase method for answering yes/no questions. We answer easy questions by negation/antonym detection. For more difficult questions, we adapt an unsupervised machine learning method based on morphological, syntactic, and lexical semantic analysis on identified premises and conclusions. This provides the basis to compare the semantic correlation between a question and a legal article. Our experimental results show reasonable performance, which improves the baseline system, and outperforms an SVM-based supervised machine learning model.
In many practical situations, batching of similar jobs to avoid setups is performed while constru... more In many practical situations, batching of similar jobs to avoid setups is performed while constructing a schedule. This paper addresses the problem of non-preemptively scheduling independent jobs in a two-machine flow shop with the objective of minimizing the makespan. Jobs are grouped into batches. A sequence independent batch setup time on each machine is required before the first job is processed, and when a machine switches from processing a job in some batch to a job of another batch. Besides its practical interest, this problem is a direct generalization of the classical two-machine flow shop problem with no grouping of jobs, which can be solved optimally by Johnson's well-known algorithm. The problem under investigation is known to be NP-hard. We propose two O(n log n) time heuristic algorithms. The first heuristic, which creates a schedule with minimum total setup time by forcing all jobs in the same batch to be sequenced in adjacent positions, has a worst-case performance ratio of 3/2. By allowing each batch to be split into at most two sub-batches, a second heuristic is developed which has an improved worst-case performance ratio of 4/3.
Most existing formalizations treat belief change as a single-step process, and ignore several pro... more Most existing formalizations treat belief change as a single-step process, and ignore several problems that become important when a theory, or belief state, is revised over several steps. This paper identifies these problems, and argues for the need to retain all of the multiple possible outcomes of a belief change step, and for a framework in which the effects of a belief change step persist as long as is consistently possible. To demonstrate that such a formalization is indeed possible, we develop a framework, which uses the language of PJ-default logic (Delgrande and Jackson 1991) to represent a belief state, and which enables the effects of a belief change step to persist by propagating belief constraints. Belief change in this framework maps one belief state to another, where each belief state is a collection of theories given by the set of extensions of the PJ-default theory representing that belief state. Belief constraints do not need to be separately recorded; they are encoded as clearly identifiable components of a PJ-default theory. The framework meets the requirements for iterated belief change that we identify and satisfies most of the AGM postulates (Alchourrón, Gärdenfors, and Makinson 1985) as well.
The success of statistical machine learning from big data, especially of deep learning, has made ... more The success of statistical machine learning from big data, especially of deep learning, has made artificial intelligence (AI) very popular. Unfortunately, especially with the most successful methods, the results are very difficult to comprehend by human experts. The application of AI in areas that impact human life (e.g., agriculture, climate, forestry, health, etc.) has therefore led to an demand for trust, which can be fostered if the methods can be interpreted and thus explained to humans. The research field of explainable artificial intelligence (XAI) provides the necessary foundations and methods. Historically, XAI has focused on the development of methods to explain the decisions and internal mechanisms of complex AI systems, with much initial research concentrating on explaining how convolutional neural networks produce image classification predictions by producing visualizations which highlight what input patterns are most influential in activating hidden units, or are most ...
We describe the development of a model to detect user-level clinical depression based on a user's... more We describe the development of a model to detect user-level clinical depression based on a user's temporal social media posts. Our model uses a Depression Symptoms Detection (DSD) model, which is trained on the largest existing samples of clinician annotated tweets for clinical depression symptoms. We subsequently use our DSD model to extract clinically relevant features, e.g., depression scores and their consequent temporal patterns, as well as user posting activity patterns, e.g., quantifying their "no activity" or "silence." Furthermore, to evaluate the efficacy of these extracted features, we create three kinds of datasets including a test dataset, from two existing well-known benchmark datasets for user-level depression detection. We then provide accuracy measures based on single features, baseline features and feature ablation tests, at several different levels of temporal granularity, data distributions, and clinical depression detection related settings to draw a complete picture of the impact of these features across our created datasets. Finally, we show that, in general, only semantic oriented representation models perform well. However, clinical features may enhance overall performance provided that the training and testing distribution is similar, and there is more data in a user's timeline. Further, we show 1
Uploads
Papers by R. Goebel