MERLOT Collection
MERLOT CollectionSelected resources from MERLOT
Works (64)
-
Presentation
This video was recorded at 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), Washington 2010. The problem of optimally managing the collections process by taxation authorities is one of prime importance, not only for the revenue it brings but also as a means to administer a fair taxing system. The analogous problem of debt collections management in the private sector, such as banks and credit card companies, is also increasingly gaining attention. With the recent successes in the applications of data analytics and optimization to various business areas, the question arises to what extent such collections processes can be improved by use of leading edge data modeling and optimization techniques. In this paper, we propose and develop a novel approach to this problem based on the framework of constrained Markov Decision Process (MDP), and report on our experience in an actual deployment of a tax collections optimization system at New York State Department of Taxation and Finance (NYS DTF).
- Subjects:
- Computing, Data Science and Artificial Intelligence and Management
- Keywords:
- Markov processes -- Mathematical models Debt -- Management
- Resource Type:
- Presentation
-
Presentation
This video was recorded at 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), Washington 2010. In 2009 IBM was recognized as a finalist of the INFORMS Edelman competition for its predictive modeling initiative to improve the productivity of its global salesforce and with an estimated business impact of ~ 100 Million dollars. The first component implements some traditional propensity modeling to identify new sales opportunities and is currently used by over 13,000 sales reps. The second 'wallet estimation' component is used strategically to allocate sales resources based on validated analytical estimates of revenue opportunity. In this case study we cover the key elements leading to the success including the data integration, data mining and predictive modeling, solution delivery, human guided model validation, integration of the business process and we conclude with an assessment of the bottom-line business impact.
- Subjects:
- Computing, Data Science and Artificial Intelligence and Management
- Keywords:
- Sales management -- Data processing Management -- Data processing
- Resource Type:
- Presentation
-
Presentation
This video was recorded at European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), Athens 2011. Comparing frequency counts over texts or corpora is an important task in many applications and scientific disciplines. Given a text corpus, we want to test a hypothesis, such as "word X is frequent", "word X has become more frequent over time", or "word X is more frequent in male than in female speech". For this purpose we need a null model of word frequencies. The commonly used bag-of-words model, which corresponds to a Bernoulli process with fixed parameter, does not account for any structure present in natural languages. Using this model for word frequencies results in large numbers of words being reported as unexpectedly frequent. We address how to take into account the inherent occurrence patterns of words in significance testing of word frequencies. Based on studies of words in two large corpora, we propose two methods for modeling word frequencies that both take into account the occurrence patterns of words and go beyond the bag-of-words assumption. The first method models word frequencies based on the spatial distribution of individual words in the language. The second method is based on bootstrapping and takes into account only word frequency at the text level. The proposed methods are compared to the current gold standard in a series of experiments on both corpora. We find that words obey different spatial patterns in the language, ranging from bursty to non-bursty/uniform, independent of their frequency, showing that the traditional approach leads to many false positives.
- Subjects:
- Computing, Data Science and Artificial Intelligence and Management
- Keywords:
- Computational linguistics Text processing (Computer science) Discourse analysis -- Data processing
- Resource Type:
- Presentation
-
Presentation
This video was recorded at European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), Athens 2011. Hierarchical modeling and reasoning are fundamental in machine intelligence, and for this the two-parameter Poisson-Dirichlet Process (PDP) plays an important role. The most popular MCMC sampling algorithm for the hierarchical PDP and hierarchical Dirichlet Process is to conduct an incremental sampling based on the Chinese restaurant metaphor, which originates from the Chinese restaurant process (CRP). In this paper, with the same metaphor, we propose a new table representation for the hierarchical PDPs by introducing an auxiliary latent variable, called table indicator, to record which customer takes responsibility for starting a new table. In this way, the new representation allows full exchangeability that is an essential condition for a correct Gibbs sampling algorithm. Based on this representation, we develop a block Gibbs sampling algorithm, which can jointly sample the data item and its table contribution. We test this out on the hierarchical Dirichlet process variant of latent Dirichlet allocation (HDP-LDA) developed by Teh, Jordan, Beal and Blei. Experiment results show that the proposed algorithm outperforms their "posterior sampling by direct assignment" algorithm in both out-of-sample perplexity and convergence speed. The representation can be used with many other hierarchical PDP models.
- Subjects:
- Computing, Data Science and Artificial Intelligence
- Keywords:
- Machine learning Artificial intelligence
- Resource Type:
- Presentation
-
Presentation
This video was recorded at 5th Annual European Semantic Web Conference (ESWC), Tenerife 2008. The degree of automation in the management of the business process space of single enterprises and whole value chains is still unsatisfying. A key source of problems are representational heterogeneities between the various perspectives and the various stages in the life-cycles of business processes. Typical examples are incompatible representations of the managerial vs. the IT perspective, or the gap between normative modeling for compliance purposes and process execution log data. As early as in the 1990s, researchers have evaluated the potential of using ontologies for improving business process management in the context of the TOVE project; however, the impact of that work remained beyond initial expectations. Since 2005, there is now a renewed and growing interest in exploiting ontologies, of varying expressivity and focus, for advancing the state of the art in business process management, in particular in ERP-centric IT landscapes. The term "Semantic Business Process Management" has been suggested for the described branch of research in an early 2005 paper, which is now frequently cited as the first description of the overall vision. A flagship activity in the field is the European research project "SUPER", with more than a dozen premier industrial and academic partners, among them SAP, IDS Scheer, and IBM. In the past two years, substantial advancement has been made in investigating the theoretical and practical branches of this vision. However, the interdisciplinary nature of the topic requires a tight collaboration of researcher from multiple fields of, namely the BPM, SOA, Semantic Web, Semantic Web services, and Economics communities. There is a clear need for an annual event at which those communities meet, debate, challenge each others approaches, and eventually align their research efforts. Due to the strong involvement of Semantic Web researchers in the field, ESWC is the ideal target venue for this event. In this workshop, we want to bring together experts from the relevant communities and help reach agreement on a roadmap for SBPM research. We aim at bundling experiences and prototypes from the successful application of Semantic Web technology to BPM in various industries, like automotive, engineering, chemical and pharmaceutical, and services domains. The particular focus is on deriving reusable best-practices from such experiences, and to yield convincing showcases of semantic technology.
- Subjects:
- Computing, Data Science and Artificial Intelligence and Management
- Keywords:
- Industrial management Workflow -- Management
- Resource Type:
- Presentation
-
Presentation
This video was recorded at COIN / ACTIVE Summer School on Advanced Technologies for Knowledge Intensive Networked Organizations, Aachen 2010. Part 1. Context Computing. Context is used as a term for packaging information for a particular need. A criterion for selecting or prioritization information from a broader pool of information could be called contextual model. Search can be contextual: http://searchpoint.ijs.si. The relevance of Context in computing seems to be growing. Many application areas see an opportunity in extending its value by introducing "context sensitivity". More details do to be found in ISWC2006 Tutorial on "context sensitivity": http://videolectures.net/iswc06_athens_ga/ Part 2. Text Mining & Light Weight Semantics. Videolectures discusses the following topics: - levels of text representations - modeling the data (Support Vector Machine) - classification into large taxonomies (DMoz) - visual & contextual search (Search Point) - multilingual search - news bias, news visualization - text enrichment (Enrycher) - knowledge based summarization - question answering (AnswerArt) - Cyc knowledge base and reasoning
- Subjects:
- Computing, Data Science and Artificial Intelligence and Management
- Keywords:
- Data mining Management information systems Business -- Data Processing Information resources management
- Resource Type:
- Presentation
-
Presentation
This video was recorded at COIN / ACTIVE Summer School on Advanced Technologies for Knowledge Intensive Networked Organizations, Aachen 2010. Organized by COIN FP7 Integrated Project (http://www.coin-ip.eu/) and ACTIVE FP7 Integrated Project (http://www.active-project.eu), the summer school seeks to bring together students, scholars and researchers from industry in order to foster collaboration and interoperability through innovative software solutions and share the recent research developments from well-established researchers and educators. The main topics of the summer school are: Interoperability and collaboration models and solutions, Enterprise interoperability and collaboration services, Innovative knowledge and semantically powered technologies, Knowledge process and context modeling, Pro-active knowledge tools, Large scale analytics and reasoning tools, Business cases and real case studies. More about the event at http://coin-active-ss.ijs.si/
-
Presentation
This video was recorded at COIN / ACTIVE Summer School on Advanced Technologies for Knowledge Intensive Networked Organizations, Aachen 2010. Organized by COIN FP7 Integrated Project (http://www.coin-ip.eu/) and ACTIVE FP7 Integrated Project (http://www.active-project.eu), the summer school seeks to bring together students, scholars and researchers from industry in order to foster collaboration and interoperability through innovative software solutions and share the recent research developments from well-established researchers and educators. The main topics of the summer school are: Interoperability and collaboration models and solutions, Enterprise interoperability and collaboration services, Innovative knowledge and semantically powered technologies, Knowledge process and context modeling, Pro-active knowledge tools, Large scale analytics and reasoning tools, Business cases and real case studies. More about the event at http://coin-active-ss.ijs.si/
-
Presentation
This video was recorded at COIN / ACTIVE Summer School on Advanced Technologies for Knowledge Intensive Networked Organizations, Aachen 2010. Organized by COIN FP7 Integrated Project (http://www.coin-ip.eu/) and ACTIVE FP7 Integrated Project (http://www.active-project.eu), the summer school seeks to bring together students, scholars and researchers from industry in order to foster collaboration and interoperability through innovative software solutions and share the recent research developments from well-established researchers and educators. The main topics of the summer school are: Interoperability and collaboration models and solutions, Enterprise interoperability and collaboration services, Innovative knowledge and semantically powered technologies, Knowledge process and context modeling, Pro-active knowledge tools, Large scale analytics and reasoning tools, Business cases and real case studies. More about the event at http://coin-active-ss.ijs.si/
-
Presentation
This video was recorded at COIN / ACTIVE Summer School on Advanced Technologies for Knowledge Intensive Networked Organizations, Aachen 2010. Organized by COIN FP7 Integrated Project (http://www.coin-ip.eu/) and ACTIVE FP7 Integrated Project (http://www.active-project.eu), the summer school seeks to bring together students, scholars and researchers from industry in order to foster collaboration and interoperability through innovative software solutions and share the recent research developments from well-established researchers and educators. The main topics of the summer school are: Interoperability and collaboration models and solutions, Enterprise interoperability and collaboration services, Innovative knowledge and semantically powered technologies, Knowledge process and context modeling, Pro-active knowledge tools, Large scale analytics and reasoning tools, Business cases and real case studies. More about the event at http://coin-active-ss.ijs.si/