Automatic evaluation of text for its innovative quality has been necessitated by the growing trend to organize open innovation contests by different organizations. Such online/offline contests are known to fuel major business benefits to many industries. However, open contests result in a huge number of documents of which only a few may contain potentially interesting and relevant ideas. Usually these entries are manually reviewed and scored by multiple experts. But manual evaluation process not only require a lot of time and effort but are also prone to erroneous judgments due to inter-annotator disagreements. To counter this issue, in this paper, we have proposed a new approach towards detecting novelty or innovativeness of textual ideas from a given collection of ideas. The proposed approach uses information theoretic measures and term relevance to domain to compute document level innovativeness score. We have evaluated the performance of the proposed approach with a real world collection of innovative ideas which were manually scored by experts. We have compared the performance of our proposed model with some of the commonly used baseline approaches that rely on distributional semantics and geometric distances. The result shows that the proposed method outperform the existing baseline models.

Robust intelligent systems may leverage knowledge about the world to cope with a variety of contexts.While automatic knowledge extraction algorithms have been successfully used to build knowledge bases in English,little progress has been made in extracting non-alphabetic languages, e.g. Chinese.This paper identifies the key challenge in instance and pattern extraction for Chinese and presents the Coupled Chinese Pattern Learner that utilizes part-of-speech tagging and language-dependent grammar rules for generalized matching in the Chinese never-ending language learner framework for large-scale knowledge extraction from online documents.Experiments showed that the proposed system is scalable and achieves a precision of 79.9% in learning categories after a small number of iterations.

In this paper we describe the Unified Cybersecurity Ontology (UCO) that is intended to support information integration and cyber situational awareness in cybersecurity systems. The ontology incorporates and integratesheterogeneous data and knowledge schemas from different cybersecurity systems and most commonly usedcybersecurity standards for information sharing and exchange. The UCO ontology has also been mapped to anumber of existing cybersecurity ontologies as well asconcepts in the Linked Open Data cloud (Berners-Lee,Bizer, and Heath 2009). Similar to DBpedia (Auer etal. 2007) which serves as the core for general knowledge in Linked Open Data cloud, we envision UCO toserve as the core for cybersecurity domain, which wouldevolve and grow with the passage of time with additional cybersecurity data sets as they become available.We also present a prototype system and concrete usecases supported by the UCO ontology. To the best of ourknowledge, this is the first cybersecurity ontology thathas been mapped to general world ontologies to support broader and diverse security use cases. We comparethe resulting ontology with previous efforts, discuss itsstrengths and limitations, and describe potential futurework directions.

In intensive care units (ICU), electrocardiogram (ECG) waveforms show diverse variationsunder different patients' physical conditions.In general, physicians can diagnose patients efficientlyby detecting any disorder of heart rate or rhythm and any change in the morphological pattern of ECG data,which contain underlying semantics.To help physicians better analyze ECG data in a fairly short time,it is essential to develop a novel method for mining semantics from ECG patterns.This paper is the very first time to characterize ECG patterns by using Prefix Scalable Pattern Tree (PSP-Tree).Comparing with similar currently existing methods, PSP-Tree can mine significant semantics,such as scalability, temporality and hierarchy over ECG patterns.We conduct extensive experiments on real ECG data set which are obtained from PhysioBank Community and Beijing No.3 People Hospital.The experiment results show that our method performs more feasibly and effectively than other related work.

We propose a new model for learning bilingual word representations from non-parallel document-aligned data. Following the recent advances in word representation learning, our model learns dense real-valued word vectors, that is, bilingual word embeddings (BWEs). Unlike prior work on inducing BWEs which heavily relied on parallel sentence-aligned corpora and/or readily available translation resources such as dictionaries, the article reveals that BWEs may be learned solely on the basis of document-aligned comparable data without any additional lexical resources nor syntactic information. We present a comparison of our approach with previous state-of-the-art models for learning bilingual word representations from comparable data that rely on the framework of multilingual probabilistic topic modeling (MuPTM), as well as with distributional local context-counting models. We demonstrate the utility of the induced BWEs in two semantic tasks: (1) bilingual lexicon extraction, (2) suggesting word translations in context for polysemous words. Our simple yet effective BWE-based models significantly outperform the MuPTM-based and context-counting representation models from comparable data as well as prior BWE-based models, and acquire the best reported results on both tasks for all three tested language pairs.

The paper focuses on finding the m best solutions to combinatorial optimization problems using best-first or depth-first branch and bound search. Specifically, we present a new algorithm m-A*, extending the well-known A* to the m-best task, and for the first time prove that all its desirable properties, including soundness, completeness and optimal efficiency, are maintained. Since best-first algorithms require extensive memory, we also extend the memory-efficient depth-first branch and bound to the m-best task. We adapt both algorithms to optimization tasks over graphical models (e.g., Weighted CSP and MPE in Bayesian networks), provide complexity analysis and an empirical evaluation. Our experiments confirm theory that the best-first approach is largely superior when memory is available, but depth-first branch and bound is more robust. We also show that our algorithms are competitive with related schemes recently developed for the m-best task.

Hypothetical “value learning” AIs learn human values and then try to act according to those values. The design of such AIs, however, is hampered by the fact that there exists no satisfactory definition of what exactly human values are. After arguing that the standard concept of preference is insufficient as a definition, I draw on reinforcement learning theory, emotion research, and moral psychology to offer an alternative definition. In this definition, human values are conceptualized as mental representations that encode the brain’s value function (in the reinforcement learning sense) by being imbued with a context-sensitive affective gloss. I finish with a discussion of the implications that this hypothesis has on the design of value learners.

Hashtags are useful in many applications, such as tweet classification, clustering, searching, indexing and social network analysis. This study seeks to recommend relevant Twitter hashtags for health-related keywords based on distributed language representations, generated by the state-of-the-art Deep Learning technology. The word embeddings are built from billions of tweet words without supervision. To the best of our knowledge, this is the first study of applying distributed language representations to recommending hashtags for keywords. The experiment showed that this approach outperformed the baseline approach that is based on keyword and hashtag co-occurrence in tweets.

This study investigates the public health intelligence utility of Yik Yak, a social media platform that allows users to anonymously post and view messages within precise geographic locations. Our dataset contains 122,179 “yaks” collected from 120 college campuses across the United States during 2015. We first present an exploratory analysis of the topics commonly discussed in Yik Yak, clarifying the health issues for which this may serve as a source of information. We then present an in-depth content analysis of data describing substance use, an important public health issue that is not often discussed in public social media, but commonly discussed on Yik Yak under the cloak of anonymity.

It is well known that the title of an article impacts how well it is discovered by potential readers and read. With both people and search engines, acting on behalf of people, accessing papers from digital libraries, it is important that the paper titles should promote discovery. In this paper, we investigate the characteristics of titles of AI papers and then propose au- tomatic ways to augment them so that they can be better in- dexed and discovered by users. A user study with researchers shows that they overwhelmingly prefer the augmented titles over the originals for being more helpful.