Results 1  10
of
10
Stochastic grammatical inference with Multinomial Tests
, 2002
"... We present a new statistical framework for stochastic grammatical inference algorithms based on a state merging strategy. We propose to use multinomial statistical tests to decide which states should be merged. This approach has three main advantages. First, since it is not based on asymptotic resul ..."
Abstract

Cited by 12 (1 self)
 Add to MetaCart
We present a new statistical framework for stochastic grammatical inference algorithms based on a state merging strategy. We propose to use multinomial statistical tests to decide which states should be merged. This approach has three main advantages. First, since it is not based on asymptotic results, small sample case can be specifically dealt with. Second, all the probabilities associated to a state are included in a single test so that statistical evidence is cumulated. Third, a statistical score is associated to each possible merging operation and can be used for bestfirst strategy. Improvement over classical stochastic grammatical inference algorithm is shown on artificial data.
FlowCube: Constructing RFID FlowCubes for MultiDimensional Analysis of Commodity Flows
 In: VLDB 2006
, 2006
"... With the advent of RFID (Radio Frequency Identification) technology, manufacturers, distributors, and retailers will be able to track the movement of individual objects throughout the supply chain. The volume of data generated by a typical RFID application will be enormous as each item will generate ..."
Abstract

Cited by 8 (1 self)
 Add to MetaCart
With the advent of RFID (Radio Frequency Identification) technology, manufacturers, distributors, and retailers will be able to track the movement of individual objects throughout the supply chain. The volume of data generated by a typical RFID application will be enormous as each item will generate a complete history of all the individual locations that it occupied at every point in time, possibly from a specific production line at a given factory, passing through multiple warehouses, and all the way to a particular checkout counter in a store. The movement trails of such RFID data form gigantic commodity flowgraph representing the locations and durations of the path stages traversed by each item. This commodity flow contains rich multidimensional information on the characteristics, trends, changes and outliers of commodity movements. In this paper, we propose a method to construct a warehouse of commodity flows, called flowcube. As in standard OLAP, the model will be composed of cuboids that aggregate item flows at a given abstraction level. The flowcube differs from the traditional data cube in two major ways. First, the measure of each cell will not be a scalar aggregate but a commodity flowgraph that captures the major movement trends and significant deviations of the items aggregated in the cell. Second, each flowgraph itself can be viewed at multiple levels by changing the level of abstraction of path stages. In this paper, we motivate the importance of the model, and present an efficient method to compute it by (1) performing simultaneous aggregation of paths to all interesting abstraction levels, (2) pruning low support path segments along the item and path stage abstraction lattices, and (3) compressing the cube by removing rarely occurring cells, and cells whose commodity flows can be inferred from higher level cells.
Grammatical Inference as a Principal Component Analysis Problem
"... One of the main problems in probabilistic grammatical inference consists in inferring a stochastic language, i.e. a probability distribution, in some class of probabilistic models, from a sample of strings independently drawn according to a fixed unknown target distribution p. Here, we consider the ..."
Abstract

Cited by 4 (1 self)
 Add to MetaCart
One of the main problems in probabilistic grammatical inference consists in inferring a stochastic language, i.e. a probability distribution, in some class of probabilistic models, from a sample of strings independently drawn according to a fixed unknown target distribution p. Here, we consider the class of rational stochastic languages composed of stochastic languages that can be computed by multiplicity automata, which can be viewed as a generalization of probabilistic automata. Rational stochastic languages p have a useful algebraic characterization: all the mappings ˙up: v → p(uv) lie in a finite dimensional vector subspace V ∗ p of the vector space R〈〈Σ〉 〉 composed of all realvalued functions defined over Σ ∗. Hence, a first step in the grammatical inference process can consist in identifying the subspace V ∗ p. In this paper, we study the possibility of using Principal Component Analysis to achieve this task. We provide an inference algorithm which computes an estimate of this space and then build a multiplicity automaton which computes an estimate of the target distribution. We prove some theoretical properties of this algorithm and we provide results from numerical simulations that confirm the relevance of our approach. 1.
Event extraction from heterogeneous news sources
 in ‘Proceedings of the AAAI 2006 Workshop on Event Extraction and Synthesis
, 2006
"... With the proliferation of news articles from thousands of different sources now available on the Web, summarization of such information is becoming increasingly important. Our research focuses on merging descriptions of news events from multiple sources, to provide a concise description that combine ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
With the proliferation of news articles from thousands of different sources now available on the Web, summarization of such information is becoming increasingly important. Our research focuses on merging descriptions of news events from multiple sources, to provide a concise description that combines the information from each source. Specifically, we describe and evaluate methods for grouping sentences in news articles that refer to the same event. The key idea is to cluster the sentences, using two novel distance metrics. The first distance metric exploits regularities in the sequential structure of events within a document. The second metric uses a TFIDFlike weighting scheme, enhanced to capture word frequencies within events even though the events themselves are not known a priori. Typical news articles contain sentences that do not describe specific events. We use machine learning methods to differentiate between sentences that describe one or more events, and those that do not. We then remove nonevent sentences before initiating the clustering process. We demonstrate that this approach achieves significant improvements in overall clustering performance.
Learning Hidden Markov Models to Fit LongTerm Dependencies
, 2005
"... this report a novel approach to the induction of the structure of Hidden Markov Models (HMMs). The notion of partially observable Markov models (POMMs) is introduced. POMMs form a particular case of HMMs where any state emits a single letter with probability one, but several states can emit the ..."
Abstract

Cited by 2 (2 self)
 Add to MetaCart
this report a novel approach to the induction of the structure of Hidden Markov Models (HMMs). The notion of partially observable Markov models (POMMs) is introduced. POMMs form a particular case of HMMs where any state emits a single letter with probability one, but several states can emit the same letter. It is shown that any HMM can be represented by an equivalent POMM. The proposed induction algorithm aims at finding a POMM fitting the dynamics of the target machine, that is to best approximate the stationary distribution and the mean first passage times observed in the sample. The induction relies on nonlinear optimization and iterative state splitting from an initial order one Markov chain. Experimental results illustrate the advantages of the proposed approach as compared to BaumWelch HMM estimation or backo# smoothed Ngrams equivalent to variable order Markov chains
Negative Feedback: The Forsaken Nature Available for Reranking
"... Reranking for Information Retrieval aims to elevate relevant feedbacks and depress negative ones in initial retrieval result list. Compared to relevance feedbackbased reranking method widely adopted in the literature, this paper proposes a new method to well use three features in known negative f ..."
Abstract
 Add to MetaCart
Reranking for Information Retrieval aims to elevate relevant feedbacks and depress negative ones in initial retrieval result list. Compared to relevance feedbackbased reranking method widely adopted in the literature, this paper proposes a new method to well use three features in known negative feedbacks to identify and depress unknown negative feedbacks. The features include: 1) the minor (lowerweighted) terms in negative feedbacks; 2) hierarchical distance (HD) among feedbacks in a hierarchical clustering tree; 3) obstinateness strength of negative feedbacks. We evaluate the method on the TDT4 corpus, which is made up of news topics and their relevant stories. And experimental results show that our new scheme substantially outperforms its counterparts. 1.
Machine Learning Approach for the Automatic Annotation of Events
"... After the beginning of the extension of current Web towards the semantics, the annotation starts to take a significant role, since it takes part to give the semantic aspect to the different types of documents. With the proliferation of news articles from thousands of different sources now available ..."
Abstract
 Add to MetaCart
After the beginning of the extension of current Web towards the semantics, the annotation starts to take a significant role, since it takes part to give the semantic aspect to the different types of documents. With the proliferation of news articles from thousands of different sources now available on the Web, summarization of such information is becoming increasingly important. We will define a methodological approach to extract the events from the news articles and to annotate them according to the principal events which they contain. Considering the large number of news source (for examples, BBC, Reuters, CNN…), every day, thousands of articles are produced in the entire world concerning a given event. This is why we have to think to automate the process of annotation of such articles.
Deriving Protocol Models from Imperfect Service Conversation Logs Hamid R. MotahariNezhad
"... Abstract—Understanding the business (interaction) protocol supported by a service is very important for both clients and service providers: It allows developers to know how to write clients that interact with a service, and it allows development tools and runtime middleware to deliver functionalitie ..."
Abstract
 Add to MetaCart
Abstract—Understanding the business (interaction) protocol supported by a service is very important for both clients and service providers: It allows developers to know how to write clients that interact with a service, and it allows development tools and runtime middleware to deliver functionalities that simplifies the service development life cycle. It also greatly facilitates the monitoring, visualization, and aggregation of interaction data. This paper presents an approach for discovering protocol models from realworld service interaction logs. It presents a novel discovery algorithm, which is widely applicable, robust to different kinds of imperfections often present in realworld service logs, and able to derive protocols of small sizes targeted for human understanding. As inferring the most precise and concise model is not always possible from imperfect service logs using purely automated method, the paper presents a novel method for userdriven refinement of the discovered protocol models. The proposed approach has been implemented and experimental results show its viability on both synthetic and realworld data sets. Index Terms—Web services, business protocols, noise handling in service logs, protocol discovery, protocol refinement. Ç 1
LARS: A learning algorithm for rewriting systems
, 2007
"... Whereas there is a number of methods and algorithms to learn regular languages, moving up the Chomsky hierarchy is proving to be a challenging task. Indeed, several theoretical barriers make the class of contextfree languages hard to learn. To tackle these barriers, we choose to change the way we r ..."
Abstract
 Add to MetaCart
Whereas there is a number of methods and algorithms to learn regular languages, moving up the Chomsky hierarchy is proving to be a challenging task. Indeed, several theoretical barriers make the class of contextfree languages hard to learn. To tackle these barriers, we choose to change the way we represent these languages. Among the formalisms that allow the definition of classes of languages, the one of stringrewriting systems (SRS) has outstanding properties. We introduce a new type of SRS’s, called Delimited SRS (DSRS), that are expressive enough to define, in a uniform way, a noteworthy and non trivial class of languages that contains all the regular languages, {a n b n: n ≥ 0}, {w ∈{a, b} ∗ : wa =wb}, the parenthesis languages of Dyck, the language of Lukasiewicz, and many others. Moreover, DSRS’s constitute an efficient (often linear) parsing device for strings, and are thus promising candidates in forthcoming applications of grammatical inference. In this paper, we pioneer the problem of their learnability. We propose a novel and sound algorithm (called LARS) which identifies a large subclass of them in polynomial time (but not data). We illustrate the execution of our algorithm through several examples, discuss the position of the class in the Chomsky hierarchy and finally raise some open questions and research directions.
Learning Behavior Models for Hybrid Timed Systems
 PROCEEDINGS OF THE TWENTYSIXTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE
, 2012
"... A tailored model of a system is the prerequisite for various analysis tasks, such as anomaly detection, fault identification, or quality assurance. This paper deals with the algorithmic learning of a system’s behavior model given a sample of observations. In particular, we consider realworld produc ..."
Abstract
 Add to MetaCart
A tailored model of a system is the prerequisite for various analysis tasks, such as anomaly detection, fault identification, or quality assurance. This paper deals with the algorithmic learning of a system’s behavior model given a sample of observations. In particular, we consider realworld production plants where the learned model must capture timing behavior, dependencies between system variables, as well as mode switches—in short: hybrid system’s characteristics. Usually, such model formation tasks are solved by human engineers, entailing the wellknown bunch of problems including knowledge acquisition, development cost, or lack of experience. Our contributions to the outlined field are as follows. (1) We present a taxonomy of learning problems related to model formation tasks. As a result, an important open learning problem for the domain of production system is identified: The learning of hybrid timed automata. (2) For this class of models, the learning algorithm HyBUTLA is presented. This algorithm is the first of its kind to solve the underlying model formation problem at scalable precision. (3) We present two case studies that illustrate the usability of this approach in realistic settings. (4) We give a proof for the learning and runtime properties of HyBUTLA.