Systems that find music recordings based on hummed or sung, melodic input are called Query-By-Humming (QBH) systems. Such systems employ search keys that are more similar to a cappella singing than the original recordings. Successful deployed systems use human computation to create these search keys: hand-entered MIDI melodies or recordings of a cappella singing. Tunebot is one such system. In this paper, we compare search results using keys built from two automated melody extraction system to those gathered using two populations of humans: local paid singers and Amazon Turk workers.

In this paper, we propose an innovative approach for the development of social collaboration argumentation systems. These systems enable a community to collaboratively create answers to questions where many possible answers, or nuanced perspectives on a single answer, can be posited. We examine the emergence of critical reasoning via crowdsourced structured discussions, which are built upon a graph-theoretic framework populated by atomic argumentation components. Finally, we address the design of the online community to best facilitate this interaction. Our main contribution is the rationale and design of the system, which can easily be extended to build a general eLearning framework.

In a separate study, we were interested in understanding people's Q&A habits on Twitter. Finding questions within Twitter turned out to be a difficult challenge, so we considered applying some traditional NLP approaches to the problem. On the one hand, Twitter is full of idiosyncrasies, which make processing it difficult. On the other it is very restricted in length and tends to employ simple syntactic constructions, which could help the performance of NLP processing. In order to find out the viability of NLP and Twitter, we built a pipeline of tools to work specifically with Twitter input for the task of finding questions in tweets. This work is still preliminary, but in this paper we discuss the techniques we used and the lessons we learned.

Malfunctioning HVAC equipment in commercial buildings wastes between 15% and 30% of energy. Many diagnosis approaches tackle this problem, but they either suffer from a lack of detailed fault information or a lack of adaptability to different buildings and equipment. Clearly, especially in the light of an ever increasing amount of sensor data that is available in heavily metered smart buildings, easily adaptable self learning in-depth diagnosis approaches are needed. This paper addresses the challenges of developing such approaches and describes the contribution artificial intelligence techniques like transfer learning, ontologies, knowledge representation or diagnosis can make in overcoming these challenges.

Mobile robots are increasingly being used in the real-world due to the availability of high-fidelity sensors and sophisticated information processing algorithms. A key challenge to the widespread deployment of robots is the ability to accurately sense the environment and collaborate towards a common objective. Probabilistic sequential decision-making methods can be used to address this challenge because they encapsulate the partial observability and non-determinism of robot domains. However, such formulations soon become intractable for domains with complex state spaces that require real-time operation. Our prior work enabled a mobile robot to use hierarchical partially observable Markov decision processes (POMDPs) to automatically tailor visual sensing and information processing to the task at hand. This paper introduces adaptive observation functions and policy re-weighting in a three-layered POMDP hierarchy to enable reliable and efficient visual processing in dynamic domains. In addition, each robot merges its beliefs with those communicated by teammates, to enable a team of robots to collaborate robustly. All algorithms are evaluated in simulated domains and on physical robots tasked with locating target objects in indoor environments.

We describe work on representing and using a rich notion ofcontext that goes beyond current networking applications focusingmostly on location. Our context model includes locationand surroundings, the presence of people and devices,inferred activities and the roles people fill in them. A keyelement of our work is the use of collaborative informationsharing where devices share and integrate knowledge abouttheir context. This introduces a requirement that users canset appropriate levels of privacy to protect the personal informationbeing collected and the inferences that can be drawnfrom it. We use Semantic Web technologies to model contextand to specify high-level, declarative policies specifying informationsharing constraints. The policies involve attributesof the subject (i.e., information recipient), target (i.e., the information)and their dynamic context (e.g., are the parties copresent).We discuss our ongoing work on context representationand inference and present a model for protecting andcontrolling the sharing of private data in context-aware mobileapplications.

The use of computer mediated communication has resulted in a new form of written text--Microtext--which is very different from well-written text. Tweets and SMS messages, which have limited length and may contain misspellings, slang, or abbreviations, are two typical examples of microtext. Microtext poses new challenges to standard natural language processing tools which are usually designed for well-written text. The objective of this work is to normalize microtext, in order to produce text that could be suitable for further treatment. We propose a normalization approach based on the source channel model, which incorporates four factors, namely an orthographic factor, a phonetic factor, a contextual factor and acronym expansion. Experiments show that our approach can normalize Twitter messages reasonably well, and it outperforms existing algorithms on a public SMS data set.

We build a mechanism to form an ontology of entities which improves a relevance of matching and searching microtext. Ontology construction starts from the seed entities and mines the web for new entities associated with them. To form these new entities, machine learning of syntactic parse trees (syntactic generalization) is applied to form commonalities between various search results for existing entities on the web. Ontology and syntactic generalization are applied to relevance improvement in search and text similarity assessment in commercial setting; evaluation results show substantial contribution of both sources to microtext processing.

End-user programming raises the possibility that the people who know best what a software system should do will be able to customize, remedy original programming defects and adapt systems as requirements change. As computing increasingly enters the home and workplace, the need for such tools is high, but state of practice approaches offer very limited capability. We describe the Interactive Bootstrapped Learning (iBL) system which allows users to modify code by interactive teaching similar to human instruction. It builds on an earlier system focused on exploring how machine learning can be used to compensate for limited instructional content. iBL provides an end-to-end solution in which user-iBL dialog gradually refines a hypothesis about what transformation to a target code base will best achieve user intent. The approach integrates elements of many AI technologies including machine learning, dialog management, AI planning and automated model construction.

Abduction is widely used in the task of plan recognition, since it can be viewed as the task of finding the best explanation for a set of observations. The major drawback of abduction is its computational complexity. The task of abductive reasoning quickly becomes intractable as the background knowledge is increased. Recent efforts in the field of computational linguistics have enriched computational resources for commonsense reasoning. The enriched knowledge base facilitates exploring practical plan recognition models in an open-domain. Therefore, it is essential to develop an efficient framework for such large-scale processing. In this paper, we propose an efficient implementation of Weighted abduction. Our framework transforms the problem of explanation finding in Weighted abduction into a linear programming problem. Our experiments showed that our approach efficiently solved problems of plan recognition and outperforms state-of-the-art tool for Weighted abduction.