Archive for the ‘HP Labs’ Category

3D Printing: When and Where Does It Make Sense?

August 7, 2013 Comments off

3D Printing: When and Where Does It Make Sense?
Source: HP Labs

Twenty years ago, when Captain Jean-Luc Picard ordered: “Tea, Earl Grey, hot” it emerged in a pot from the Star Trek replicator, a machine which made everything and anything from seemingly nothing. An image was created which is so ingrained in our perception of the possible future that 3D printing is perceived by many to be todays? equivalent of the replicator. Does it make sense to print everything and anything on a 3D printer? The media and countless amateur videos suggest that the possibilities are boundless, from a cake to a door handle, from designer shoes to a washer and 3D printing will replace traditional assembly line manufacturing in the near future. Traditional manufacturing has its drawbacks, especially mass production, but it can produce high quality for an amazingly low cost. 3D printing, on the other hand, generates items within a few hours which can be customized each time they are made. However, only in a very few cases can the quality of a mass produced item be attained via 3D printing. In this paper, we discuss glass manufacturing in the UK as an example.

About these ads

Tell Me What I Don’t Know – Making the most of Social Health Forums

June 26, 2013 Comments off

Tell Me What I Don’t Know – Making the most of Social Health Forums
Source: HP Labs

We describe a novel approach to helping patients learn about their condition using social health forums. We infer a patient’s condition based on her clinical data, and assign her to a clinical state pertaining to that condition. Our system presents the patient with forum posts that are most relevant to her current state. By adjusting a LikeMe dial, the patient can modify the results from those that are most popular with patients in the same condition to the most general that apply to patients that map to other states for the condition. The posts are ranked in order of importance to the patient condition and the dial settings and the Top-N results are shown to the patient. The initial rankings of posts for different states are obtained by weights assigned by a medical expert. The patients can also give feedback with a LIKE or a DISLIKE button on whether the post is useful to them. This is incorporated into our method to dynamically change the initial weights. As the system is used in practice the initial rankings are subsumed by the feedback provided by users of the system. Our approach is general but we give results in the context of a diabetes social health forum. The algorithms are given along with preliminary results and an illustration of the user interface. We view this effort as a component of a larger patient navigation and engagement system that can help patients gain a better understanding of their condition.

A Study on Subject Matter Expertise

December 10, 2012 Comments off

A Study on Subject Matter Expertise

Source: HP Labs

We suggest a method to automatically evaluate employees’ subject matter expertise in knowledge management systems. The approach stems from several conclusions that arose in extensive research of subject matter experts and their expertise. We find that self-reported expertise is biased and constantly changing. We, furthermore, observe that automatic methods to infer expertise perform about as well as self-reported profiles. The crux of the proposed method is to combine automatic inference expertise with the benefit of the expert’s opinion, and to do so periodically so the profile remains accurate and fresh.

Privacy Management in Global Organisations

November 9, 2012 Comments off

Privacy Management in Global Organisations

Source: HP Labs

In this paper it is considered how meeting privacy requirements can be challenging for global organisations, particularly where future Internet service provision models are involved. Approaches will be explained that can be used to help address these issues, with a focus on some of the innovative solutions that the author has been involved in developing in HP Labs that are currently being used, rolled out or are the subjects of further research.

Enterprise Collective: Connecting People via Content

August 18, 2012 Comments off

Enterprise Collective: Connecting People via Content
Source: HP Labs

We describe an application for rapidly and optimally responding to enterprise opportunities and challenges, by leveraging the tacit knowledge in an enterprise, via identifying the right subject matter expert(s). Enterprise Collective is a web application that automatically discovers experts and their expertise via semantic analysis of their work products (e.g., e- mails, patents, papers, reports, presentations, and blogs). The key feature of Enterprise Collective is being “passive”; where the employees do not fill out or maintain forms or profiles. The application provides an interactive user interface that hides the underlying complexity. Enterprise Collective can benefit any business user, without extensive training or any analytical background. The application leverages the Expert-Expertise, Expert-Documents, and Expertise-Documents relationships, and subsequently permits navigation within this knowledge space. Enterprise Collective uses technologies for semantic analysis of work products and relevance computation using graph flow. A semi-automatic taxonomy generator is used to extract expertise from documents. The “authority” of each expert in relation to an expertise is computed via the nature of the work product and frequency of references. To demonstrate the benefit of Enterprise Collective in large organization, we describe a case study.

A Live Comparison of Methods for Personalized Article Recommendation at

August 12, 2012 Comments off

A Live Comparison of Methods for Personalized Article Recommendation at
Source: HP Labs

We present the results of a multi-phase study to optimize strategies for generating personalized article recommendations at the web site. In the first phase we compared the performance of a variety of recommendation methods on historical data. In the second phase we deployed a live system at for five months on a sample of 82,000 users, each randomly assigned to one of 20 methods. We analyze the live results both in terms of click- through rate (CTR) and user session lengths. The method with the best CTR was a hybrid of collaborative-filtering and a content-based method that leverages Wikipedia-based concept features, post- processed by a novel Bayesian remapping technique that we introduce. It both statistically significantly beat decayed popularity and increased CTR by 37%.

How to Track Your Data: Rule-Based Data Provenance Tracing Algorithms

July 26, 2012 Comments off

How to Track Your Data: Rule-Based Data Provenance Tracing Algorithms
Source: HP Labs

As cloud computing and virtualization technologies become mainstream, the need to be able to track data has grown in importance. Having the ability to track data from its creation to its current state or its end state will enable the full transparency and accountability in cloud computing environments. In this paper, we showcase a novel technique for tracking end-to-end data provenance, a meta-data describing the derivation history of data. This breakthrough is crucial as it enhances trust and security for complex computer systems and communication networks. By analyzing and utilizing provenance, it is possible to detect various data leakage threats and alert data administrators and owners; thereby addressing the increasing needs of trust and security for customers’ data. We also present our rule-based data provenance tracing algorithms, which trace data provenance to detect actual operations that have been performed on files, especially those under the threat of leaking customers’ data. We implemented the cloud data provenance algorithms into an existing software with a rule correlation engine, show the performance of the algorithms in detecting various data leakage threats, and discuss technically its capabilities and limitations.


Get every new post delivered to your Inbox.

Join 919 other followers