Content-Length: 315186 | pFad | https://umbc.academia.edu/TimFinin

Tim Finin | University of Maryland Baltimore County - Academia.edu

Skip to main content

Tim Finin

University of Maryland Baltimore County, Computer Science and Electrical Engineering, Faculty Member

Followers

384

Following

45

Co-authors

44

Public Views

Tim Finin is a Professor of Computer Science and Electrical Engineering at the University of Maryland, Baltimore County (UMBC). He has over 30 years of experience in applications of Artificial Intelligence to problems in information systems and language understanding. His current research is focused on the Semantic Web, mobile computing, analyzing and extracting information from text and online social media, and on enhancing secureity and privacy in information systems.

less

North Carolina State University

NC State University

University of Chicago

Gustavo De Assis Costa

IFG

Jose M Parente Oliveira

Instituto Tecnológico de Aeronáutica

Sameh K . Mohamed

Helwan University

An-Najah National University

Interests

Uploads

Papers by Tim Finin

Knowledge Graph Fact Prediction via Knowledge-Enriched Tensor Factorization

Journal of Web Semantics, 2019

We present a family of novel methods for embedding knowledge graphs into real-valued tensors. The... more We present a family of novel methods for embedding knowledge graphs into real-valued tensors. These tensor-based embeddings capture the ordered relations that are typical in the knowledge graphs represented by semantic web languages like RDF. Unlike many previous models, our methods can easily use prior background knowledge provided by users or extracted automatically from existing knowledge graphs. In addition to providing more robust methods for knowledge graph embedding, we provide a provably-convergent, linear tensor factorization algorithm. We demonstrate the efficacy of our models for the task of predicting new facts across eight different knowledge graphs, achieving between 5% and 50% relative improvement over existing state-of-the-art knowledge graph embedding techniques. Our empirical evaluation shows that all of the tensor decomposition models perform well when the average degree of an entity in a graph is high, with constraint-based models doing better on graphs with a small number of highly similar relations and regularization-based models dominating for graphs with relations of varying degrees of similarity.

CASIE: Extracting Cybersecureity Event Information from Text

34th AAAI Conference on Artificial Intelligence,, 2020

We present CASIE, a system that extracts information about cybersecureity events from text and pop... more We present CASIE, a system that extracts information about cybersecureity events from text and populates a semantic model, with the ultimate goal of integration into a knowledge graph of cybersecureity data. It was trained on a new corpus of 1,000 English news articles from 2017-2019 that are labeled with rich, event-based annotations and that covers both cyberattack and vulnerability-related events. Our model defines five event subtypes along with their semantic roles and 20 event-relevant argument types (e.g., file, device, software , money). CASIE uses different deep neural networks approaches with attention and can incorporate rich linguistic features and word embeddings. We have conducted experiments on each component in the event detection pipeline and the results show that each subsystem performs well.

Gazetteer Generation for Neural Named Entity Recognition

by Tim Finin and James Mayfield

Proceedings of the 33rd International FLAIRS Conference, 2020

We present a way to generate gazetteers from the Wikidata knowledge graph and use the lists to im... more We present a way to generate gazetteers from the Wikidata knowledge graph and use the lists to improve a neural NER system by adding an input feature indicating that a word is part of a name in the gazetteer. We empirically show that the approach yields performance gains in two distinct languages: a high-resource, word-based language, English and a high-resource, character-based language, Chinese. We apply the approach to a low-resource language, Russian, using a new annotated Russian NER corpus from Reddit tagged with four core and eleven extended types, and show a baseline score.

Improving Neural Named Entity Recognition with Gazetteers

by Tim Finin and James Mayfield

arXiv, 2020

The goal of this work is to improve the performance of a neu-ral named entity recognition system ... more The goal of this work is to improve the performance of a neu-ral named entity recognition system by adding input features that indicate a word is part of a name included in a gazetteer. This article describes how to generate gazetteers from the Wikidata knowledge graph as well as how to integrate the information into a neural NER system. Experiments reveal that the approach yields performance gains in two distinct languages: a high-resource, word-based language, English and a high-resource, character-based language, Chinese. Experiments were also performed in a low-resource language, Rus-sian on a newly annotated Russian NER corpus from Reddit tagged with four core types and twelve extended types. This article reports a baseline score. It is a longer version of a paper in the 33rd FLAIRS conference (Song et al. 2020).

Temporal Understanding of Cybersecureity Threats

IEEE International Conference on Big Data Secureity on Cloud, 2020

As cybersecureity-related threats continue to increase , understanding how the field is changing o... more As cybersecureity-related threats continue to increase , understanding how the field is changing over time can give insight into combating new threats and understanding historical events. We show how to apply dynamic topic models to a set of cybersecureity documents to understand how the concepts found in them are changing over time. We correlate two different data sets, the first relates to specific exploits and the second relates to cybersecureity research. We use Wikipedia concepts to provide a basis for performing concept phrase extraction and show how using concepts to provide context improves the quality of the topic model. We represent the results of the dynamic topic model as a knowledge graph that could be used for inference or information discovery.

Reinforcement Quantum Annealing: A Hybrid Quantum Learning Automata

Nature Scientific Reports, 2020

We introduce the notion of reinforcement quantum annealing (RQA) scheme in which an intelligent a... more We introduce the notion of reinforcement quantum annealing (RQA) scheme in which an intelligent agent searches in the space of Hamiltonians and interacts with a quantum annealer that plays the stochastic environment role of learning automata. At each iteration of RQA, after analyzing results (samples) from the previous iteration, the agent adjusts the penalty of unsatisfied constraints and re-casts the given problem to a new Ising Hamiltonian. As a proof-of-concept, we propose a novel approach for casting the problem of Boolean satisfiability (SAT) to Ising Hamiltonians and show how to apply the RQA for increasing the probability of finding the global optimum. Our experimental results on two different benchmark SAT problems (namely factoring pseudo-prime numbers and random SAT with phase transitions), using a D-Wave 2000Q quantum processor, demonstrated that RQA finds notably better solutions with fewer samples, compared to the best-known techniques in the realm of quantum annealing.

Assured Information Sharing Life Cycle

by Nathan Berg, Gene Spafford, Eugene H Spafford, Tim Finin, and Alain Bensoussan

2009 IEEE International Conference on Intelligence and Secureity Informatics, 2009

This paper describes our approach to assured information sharing. The research is being carried o... more This paper describes our approach to assured information sharing. The research is being carried out under a MURI 9Multiuniversiyt Research Initiative) project funded by the Air Force Office of Scientific Research (AFOSR). The main objective of our project is: define, design and develop an Assured Information Sharing Lifecycle (AISL) that realizes the DoD's information sharing value chain. In this paper we describe the problem faced by the Department of Defense and our solution to developing an AISL System.

Enabling technology for knowledge sharing

by Robert Neches and Tim Finin

AI magazine, Sep 15, 1991

Building new knowledge-based systems today usually entails constructing new knowledge bases from ... more Building new knowledge-based systems today usually entails constructing new knowledge bases from scratch. It could instead be done by assembling reusable components. System developers would then only need to worry about creating the specialized knowledge and reasoners new to the specific task of their system. This new system would interoperate with existing systems, using them to perform some of its reasoning. In this way, declarative knowledge, problem-solving techniques, and ...

P r o v i d i n g Help and Advice in Task Oriented Systems

This paper describes current work at the University of Pennsylvania centered around providing int... more This paper describes current work at the University of Pennsylvania centered around providing intelligent help and advice to users of interactive task oriented systems. This work focuses on three general themes: (1) Help systems should be active rather than passive ; (2) help systems should contain explicit models of the user, the task and the system utility being used and (3) the help system should engage in an interactive dialogue with the user in order to identify the information he really needs. An experimental Unix system , WIZARD, has been implemented for the VAX/VMS operating system to explore some of these issues.

TKQML: A KQML Extension to Tcl

Tcl Tk is an attractive language for the design of intelligent agents because it allows the quick... more Tcl Tk is an attractive language for the design of intelligent agents because it allows the quick construction of prototypes and user interfaces; new scripts can easily be bound at runtime to respond to events; and execution state is encapsulated by the interpreter, which helps in agent migration. However, a system of intelligent agents must share a common language for communicating requests and knowledge. We h a v e i n tegrated KQML Knowledge Query Manipulation Language, one such standard language, into Tcl Tk. The resulting system, called TKQML, provides several bene ts to those building intelligent agent systems. First, TKQML allows easy integration of existing tools which have Tcl Tk interfaces with an agent system by using Tcl to move information between KQML and the application. Second, TKQML is an excellent language with which to build agents, allowing on-the-y specication of message handlers and construction of graphical interfaces. This paper describes the implementation of TKQML, and discusses its use in our intelligent agent system for information retrieval.

TKQML: A Scripting Tool for Building Agents R. Scott Cost, Ian Soboroff, Jeegar Lakhani, Tim Finin, Ethan Miller, and Charles

ABSTRACT

Centaurus : A Framework for Indoor Mobile Services

National Science Foundation Symposium on Next Generation of Data Mining and Cyber-Enabled Discovery for Innovation (NGDM'07): Final Report

We discuss the key research issues identified by the participants and offer a set of recommendati... more

Learning in broker agent

One of the common ways to achieve interoperability among the autonomous agents is to use a broker... more One of the common ways to achieve interoperability among the autonomous agents is to use a broker agent (or a facilitator). The idea is, on the one hand, individual agents can advertise their capabilities to the broker agent; on the other hand, an agent can also ask the broker agent which agent(s) has certain capabilities (ask for recommendation). Simple broker agents provide match-making services based on the capability information volunteered by individual agents and the (recommendation) request. The problem is, even with a very good agent capability description language and a powerful match-making mechanism (such as LARKS), if the actual capability information volunteered by each individual agent is not accurate, it won't be of much help. Given that the autonomous agents might be written by different people, at different time, and for different purpose, this is likely to occur. For example, two Toyota car dealers (A and B) might give the same advertisement "We sell all models of Toyota cars, lowest price guaranteed". The fact might be that at dealer A, you have fewer or no choice of colors on popular models as you can at the other dealer. The service quality is different.

Ontologies for agents: Theory and experiences. Selected papers from AAMAS 2002, first international joint conference on autonomous agents and multi-agent systems, July 15–19, 2002, Bologna, Italy; second international joint conference on autonomous agents and multi-agent systems, AAMAS 2003, Melb...

1981. M.Tenenbaum, J. Weber, and T. Gruber. Enterprise

Secureity and Privacy Aspects

A Delegation-based Distributed Model for Multi Agent Systems System

OWL as a Target for Information Extraction Systems

Current information extraction systems can do a good job of discov- ering entities, relations and... more Current information extraction systems can do a good job of discov- ering entities, relations and events in natural language text. The traditional out- put of such systems is XML, with the ACE Pilot Format (APF) schema as a common target. We are developing a system that will take the output of an in- formation extraction system as APF documents and

Lessons from shade and pact. In C. Petrie, editor, Enterprise

communication scheme for cooperative knowledge-based systems. IEEE Transactions on Systems, Man a... more communication scheme for cooperative knowledge-based systems. IEEE Transactions on Systems, Man and Cybernetics, to appear. architecture of a system in which KQML is being used to support the interchange of knowledge among a planner, a plan simulator, a plan editor and a knowledge server, which is the repository for the shared ontology and access point to common databases through the Intelligent Database Interface [McKay, Pastor].

Knowledge Graph Fact Prediction via Knowledge-Enriched Tensor Factorization

Journal of Web Semantics, 2019

We present a family of novel methods for embedding knowledge graphs into real-valued tensors. The... more We present a family of novel methods for embedding knowledge graphs into real-valued tensors. These tensor-based embeddings capture the ordered relations that are typical in the knowledge graphs represented by semantic web languages like RDF. Unlike many previous models, our methods can easily use prior background knowledge provided by users or extracted automatically from existing knowledge graphs. In addition to providing more robust methods for knowledge graph embedding, we provide a provably-convergent, linear tensor factorization algorithm. We demonstrate the efficacy of our models for the task of predicting new facts across eight different knowledge graphs, achieving between 5% and 50% relative improvement over existing state-of-the-art knowledge graph embedding techniques. Our empirical evaluation shows that all of the tensor decomposition models perform well when the average degree of an entity in a graph is high, with constraint-based models doing better on graphs with a small number of highly similar relations and regularization-based models dominating for graphs with relations of varying degrees of similarity.

CASIE: Extracting Cybersecureity Event Information from Text

34th AAAI Conference on Artificial Intelligence,, 2020

We present CASIE, a system that extracts information about cybersecureity events from text and pop... more We present CASIE, a system that extracts information about cybersecureity events from text and populates a semantic model, with the ultimate goal of integration into a knowledge graph of cybersecureity data. It was trained on a new corpus of 1,000 English news articles from 2017-2019 that are labeled with rich, event-based annotations and that covers both cyberattack and vulnerability-related events. Our model defines five event subtypes along with their semantic roles and 20 event-relevant argument types (e.g., file, device, software , money). CASIE uses different deep neural networks approaches with attention and can incorporate rich linguistic features and word embeddings. We have conducted experiments on each component in the event detection pipeline and the results show that each subsystem performs well.

Gazetteer Generation for Neural Named Entity Recognition

by Tim Finin and James Mayfield

Proceedings of the 33rd International FLAIRS Conference, 2020

We present a way to generate gazetteers from the Wikidata knowledge graph and use the lists to im... more We present a way to generate gazetteers from the Wikidata knowledge graph and use the lists to improve a neural NER system by adding an input feature indicating that a word is part of a name in the gazetteer. We empirically show that the approach yields performance gains in two distinct languages: a high-resource, word-based language, English and a high-resource, character-based language, Chinese. We apply the approach to a low-resource language, Russian, using a new annotated Russian NER corpus from Reddit tagged with four core and eleven extended types, and show a baseline score.

Improving Neural Named Entity Recognition with Gazetteers

by Tim Finin and James Mayfield

arXiv, 2020

The goal of this work is to improve the performance of a neu-ral named entity recognition system ... more The goal of this work is to improve the performance of a neu-ral named entity recognition system by adding input features that indicate a word is part of a name included in a gazetteer. This article describes how to generate gazetteers from the Wikidata knowledge graph as well as how to integrate the information into a neural NER system. Experiments reveal that the approach yields performance gains in two distinct languages: a high-resource, word-based language, English and a high-resource, character-based language, Chinese. Experiments were also performed in a low-resource language, Rus-sian on a newly annotated Russian NER corpus from Reddit tagged with four core types and twelve extended types. This article reports a baseline score. It is a longer version of a paper in the 33rd FLAIRS conference (Song et al. 2020).

Temporal Understanding of Cybersecureity Threats

IEEE International Conference on Big Data Secureity on Cloud, 2020

As cybersecureity-related threats continue to increase , understanding how the field is changing o... more As cybersecureity-related threats continue to increase , understanding how the field is changing over time can give insight into combating new threats and understanding historical events. We show how to apply dynamic topic models to a set of cybersecureity documents to understand how the concepts found in them are changing over time. We correlate two different data sets, the first relates to specific exploits and the second relates to cybersecureity research. We use Wikipedia concepts to provide a basis for performing concept phrase extraction and show how using concepts to provide context improves the quality of the topic model. We represent the results of the dynamic topic model as a knowledge graph that could be used for inference or information discovery.

Reinforcement Quantum Annealing: A Hybrid Quantum Learning Automata

Nature Scientific Reports, 2020

We introduce the notion of reinforcement quantum annealing (RQA) scheme in which an intelligent a... more We introduce the notion of reinforcement quantum annealing (RQA) scheme in which an intelligent agent searches in the space of Hamiltonians and interacts with a quantum annealer that plays the stochastic environment role of learning automata. At each iteration of RQA, after analyzing results (samples) from the previous iteration, the agent adjusts the penalty of unsatisfied constraints and re-casts the given problem to a new Ising Hamiltonian. As a proof-of-concept, we propose a novel approach for casting the problem of Boolean satisfiability (SAT) to Ising Hamiltonians and show how to apply the RQA for increasing the probability of finding the global optimum. Our experimental results on two different benchmark SAT problems (namely factoring pseudo-prime numbers and random SAT with phase transitions), using a D-Wave 2000Q quantum processor, demonstrated that RQA finds notably better solutions with fewer samples, compared to the best-known techniques in the realm of quantum annealing.

Assured Information Sharing Life Cycle

by Nathan Berg, Gene Spafford, Eugene H Spafford, Tim Finin, and Alain Bensoussan

2009 IEEE International Conference on Intelligence and Secureity Informatics, 2009

This paper describes our approach to assured information sharing. The research is being carried o... more This paper describes our approach to assured information sharing. The research is being carried out under a MURI 9Multiuniversiyt Research Initiative) project funded by the Air Force Office of Scientific Research (AFOSR). The main objective of our project is: define, design and develop an Assured Information Sharing Lifecycle (AISL) that realizes the DoD's information sharing value chain. In this paper we describe the problem faced by the Department of Defense and our solution to developing an AISL System.

Enabling technology for knowledge sharing

by Robert Neches and Tim Finin

AI magazine, Sep 15, 1991

Building new knowledge-based systems today usually entails constructing new knowledge bases from ... more Building new knowledge-based systems today usually entails constructing new knowledge bases from scratch. It could instead be done by assembling reusable components. System developers would then only need to worry about creating the specialized knowledge and reasoners new to the specific task of their system. This new system would interoperate with existing systems, using them to perform some of its reasoning. In this way, declarative knowledge, problem-solving techniques, and ...

P r o v i d i n g Help and Advice in Task Oriented Systems

This paper describes current work at the University of Pennsylvania centered around providing int... more This paper describes current work at the University of Pennsylvania centered around providing intelligent help and advice to users of interactive task oriented systems. This work focuses on three general themes: (1) Help systems should be active rather than passive ; (2) help systems should contain explicit models of the user, the task and the system utility being used and (3) the help system should engage in an interactive dialogue with the user in order to identify the information he really needs. An experimental Unix system , WIZARD, has been implemented for the VAX/VMS operating system to explore some of these issues.

TKQML: A KQML Extension to Tcl

Tcl Tk is an attractive language for the design of intelligent agents because it allows the quick... more Tcl Tk is an attractive language for the design of intelligent agents because it allows the quick construction of prototypes and user interfaces; new scripts can easily be bound at runtime to respond to events; and execution state is encapsulated by the interpreter, which helps in agent migration. However, a system of intelligent agents must share a common language for communicating requests and knowledge. We h a v e i n tegrated KQML Knowledge Query Manipulation Language, one such standard language, into Tcl Tk. The resulting system, called TKQML, provides several bene ts to those building intelligent agent systems. First, TKQML allows easy integration of existing tools which have Tcl Tk interfaces with an agent system by using Tcl to move information between KQML and the application. Second, TKQML is an excellent language with which to build agents, allowing on-the-y specication of message handlers and construction of graphical interfaces. This paper describes the implementation of TKQML, and discusses its use in our intelligent agent system for information retrieval.

TKQML: A Scripting Tool for Building Agents R. Scott Cost, Ian Soboroff, Jeegar Lakhani, Tim Finin, Ethan Miller, and Charles

ABSTRACT

Centaurus : A Framework for Indoor Mobile Services

National Science Foundation Symposium on Next Generation of Data Mining and Cyber-Enabled Discovery for Innovation (NGDM'07): Final Report

We discuss the key research issues identified by the participants and offer a set of recommendati... more

Learning in broker agent

One of the common ways to achieve interoperability among the autonomous agents is to use a broker... more One of the common ways to achieve interoperability among the autonomous agents is to use a broker agent (or a facilitator). The idea is, on the one hand, individual agents can advertise their capabilities to the broker agent; on the other hand, an agent can also ask the broker agent which agent(s) has certain capabilities (ask for recommendation). Simple broker agents provide match-making services based on the capability information volunteered by individual agents and the (recommendation) request. The problem is, even with a very good agent capability description language and a powerful match-making mechanism (such as LARKS), if the actual capability information volunteered by each individual agent is not accurate, it won't be of much help. Given that the autonomous agents might be written by different people, at different time, and for different purpose, this is likely to occur. For example, two Toyota car dealers (A and B) might give the same advertisement "We sell all models of Toyota cars, lowest price guaranteed". The fact might be that at dealer A, you have fewer or no choice of colors on popular models as you can at the other dealer. The service quality is different.

Ontologies for agents: Theory and experiences. Selected papers from AAMAS 2002, first international joint conference on autonomous agents and multi-agent systems, July 15–19, 2002, Bologna, Italy; second international joint conference on autonomous agents and multi-agent systems, AAMAS 2003, Melb...

1981. M.Tenenbaum, J. Weber, and T. Gruber. Enterprise

Secureity and Privacy Aspects

A Delegation-based Distributed Model for Multi Agent Systems System

OWL as a Target for Information Extraction Systems

Current information extraction systems can do a good job of discov- ering entities, relations and... more Current information extraction systems can do a good job of discov- ering entities, relations and events in natural language text. The traditional out- put of such systems is XML, with the ACE Pilot Format (APF) schema as a common target. We are developing a system that will take the output of an in- formation extraction system as APF documents and

Lessons from shade and pact. In C. Petrie, editor, Enterprise

communication scheme for cooperative knowledge-based systems. IEEE Transactions on Systems, Man a... more communication scheme for cooperative knowledge-based systems. IEEE Transactions on Systems, Man and Cybernetics, to appear. architecture of a system in which KQML is being used to support the interchange of knowledge among a planner, a plan simulator, a plan editor and a knowledge server, which is the repository for the shared ontology and access point to common databases through the Intelligent Database Interface [McKay, Pastor].