In knowledge bases or information extraction results, differently expressed relations can be semantically similar (e.g., (X, wrote, Y) and (X,’s written work, Y)). Therefore, grouping semantically similar relations into clusters would facilitate and improve many applications, including knowledge base completion, information extraction, information retrieval, and more. This paper formulates relation clustering as a constrained tripartite graph clustering problem, presents an efficient clustering algorithm and exhibits the advantage of the constrained framework. We introduce several ways that provide side information via must-link and cannot link constraints to improve the clustering results. Different from traditional semi-supervised learning approaches, we propose to use the similarity of relation expressions and the knowledge of entity types to automatically construct the constraints for the algorithm. We show improved relation clustering results on two datasets extracted from human annotated knowledge base (i.e., Freebase) and open information extraction results (i.e., ReVerb data).

In this paper, we study Bayesian techniques for entity discovery and temporal segmentation of videos. Existing temporal video segmentation techniques are based on low-level features, and are usually suitable for discovering short, homogeneous shots rather than diverse scenes, each of which contains several such shots. We define scenes in terms of semantic entities (eg. persons). This is the first attempt at entity-driven scene discovery in videos, without using meta-data like scripts. The problem is hard because we have no explicit prior information about the entities and the scenes. However such sequential data exhibit temporal coherence in multiple ways, and this provides implicit cues. To capture these, we propose a Bayesian generative model- EntScene, that represents entities with mixture components and scenes with discrete distributions over these components. The most challenging part of this approach is the inference, as it involves complex interactions of latent variables. To this end, we propose an algorithm based on Dynamic Blocked Gibbs Sampling, that attempts to jointly learn the components and the segmentation, by progressively merging an initial set of short segments. The proposed algorithm compares favourably against suitably designed baselines on several TV-series videos. We extend the method to an unexplored problem: temporal co-segmentation of videos containing same entities.

This paper studies an interesting problem: is it possible to predict the crowd opinion about a movie before the movie is actually released? The crowd opinion is here expressed by the distribution of ratings given by a sufficient amount of people. Consequently, the pre-release crowd opinion prediction can be regarded as a Label Distribution Learning (LDL) problem. In order to solve this problem, a Label Distribution Support Vector Regressor (LDSVR) is proposed in this paper. The basic idea of LDSVR is to fit a sigmoid function to each component of the label distribution simultaneously by a multi-output support vector machine. Experimental results show that LDSVR can accurately predict peoples’s rating distribution about a movie just based on the pre-release metadata of the movie.

Most multi-label domains lack an authoritative taxonomy. Therefore, different taxonomies are commonly used in the same domain, which results in complications. Although this situation occurs frequently, there has been little study of it using a principled statistical approach. Given that (1) different taxonomies used in the same domain are generally founded on the same latent semantic space, where each possible label set in a taxonomy denotes a single semantic concept, and that (2) crowdsourcing is beneficial in identifying relationships between semantic concepts and instances at low cost, we proposed a novel probabilistic cascaded method for establishing a semantic matching function in a crowdsourcing setting that maps label sets in one (source) taxonomy to label sets in another (target) taxonomy in terms of the semantic distances between them. The established function can be used to detect the associated label set in the target taxonomy for an instance directly from its associated label set in the source taxonomy without any extra effort. Experimental results on real-world data (emotion annotations for narrative sentences) demonstrated that the proposed method can robustly establish semantic matching functions exhibiting satisfactory performance from a limited number of crowdsourced annotations.

In recent years, recommendation algorithms have become one of the most active research areas driven by the enormous industrial demands. Most of the existing recommender systems focus on topics such as movie, music, e-commerce etc., which essentially differ from the TV show recommendations due to the cold-start and temporal dynamics. Both effectiveness (effectively handling the cold-start TV shows) and efficiency (efficiently updating the model to reflect the temporal data changes) concerns have to be addressed to design real-world TV show recommendation algorithms. In this paper, we introduce a novel hybrid recommendation algorithm incorporating both collaborative user-item relationship as well as item content features. The cold-start TV shows can be correctly recommended to desired users via a so called space alignment technique. On the other hand, an online updating scheme is developed to utilize new user watching behaviors. We present experimental results on a real TV watch behavior data set to demonstrate the significant performance improvement over other state-of-the-art algorithms.

The goal of query conceptualization is to map instances in a query to concepts defined in a certain ontology or knowledge base. Queries usually do not observe the syntax of a written language, nor do they contain enough signals for statistical inference. However, the available context, i.e., the verbs related to the instances, the adjectives and attributes of the instances, do provide valuable clues to understand instances. In this paper, we first mine a variety of relations among terms from a large web corpus and map them to related concepts using a probabilistic knowledge base. Then, for a given query, we conceptualize terms in the query using a random walk based iterative algorithm. Finally, we examine our method on real data and compare it to representative previous methods. The experimental results show that our method achieves higher accuracy and efficiency in query conceptualization.

Trivia is any fact about an entity which is interesting due to its unusualness, uniqueness, unexpectedness or weirdness. In this paper, we propose a novel approach for mining entity trivia from their Wikipedia pages. Given an entity, our system extracts relevant sentences from its Wikipedia page and produces a list of sentences ranked based on their interestingness as trivia. At the heart of our system lies an interestingness ranker which learns the notion of interestingness, through a rich set of domain-independent linguistic and entity based features. Our ranking model is trained by leveraging existing user-generated trivia data available on the Web instead of creating new labeled data. We evaluated our system on movies domain and observed that the system performs significantly better than the defined baselines. A thorough qualitative analysis of the results revealed that our rich set of features indeed help in surfacing interesting trivia in the top ranks.

Probabilistic sentential decision diagrams (PSDDs) are a tractable representation of structured probability spaces, which are characterized by complex logical constraints on what constitutes a possible world. We develop general-purpose techniques for probabilistic reasoning and learning with PSDDs, allowing one to compute the probabilities of arbitrary logical formulas and to learn PSDDs from incomplete data. We illustrate the effectiveness of these techniques in the context of learning preference distributions, to which considerable work has been devoted in the past. We show, analytically and empirically, that our proposed framework is general enough to support diverse and complex data and query types. In particular, we show that it can learn maximum-likelihood models from partial rankings, pairwise preferences, and arbitrary preference constraints. Moreover, we show that it can efficiently answer many queries exactly, from expected and most likely rankings, to the probability of pairwise preferences, and diversified recommendations. This case study illustrates the effectiveness and flexibility of the developed PSDD framework as a domain-independent tool for learning and reasoning with structured probability spaces.

We describe an unconventional line of attack in our quest to teach machines how to rap battle by improvising hip hop lyrics on the fly, in which a novel recursive bilingual neural network, TRAAM, implicitly learns soft, context-dependent generalizations over the structural relationships between associated parts of challenge and response raps, while avoiding the exponential complexity costs that symbolic models would require. TRAAM learns feature vectors simultaneously using context from both the challenge and the response, such that challenge-response association patterns with similar structure tend to have similar vectors. Improvisation is modeled as a quasi-translation learning problem, where TRAAM is trained to improvise fluent and rhyming responses to challenge lyrics. The soft structural relationships learned by our TRAAM model are used to improve the probabilistic responses generated by our improvisational response component.

In this paper we introduce our haiku generator, which, in contrast to other systems, is not restricted to limited classic vocabulary sets and preserves a classic style without becoming too random and abstract because it performs a semantic integrity check using the Internet. Moreover, it is able to analyze blog entry input and, by using nouns and adjectives for web-mining, to stay on topic and still preserve kigo, traditional seasonal words used in Japanese poetry. The haiku generator utilizes grammar templates automatically generated from poems written by Japanese poets and a lexicon of 2,473 kigo words from an online haiku repository. In addition to generating haiku poems, it can output them vocally together with related sound effects and images retrieved from the WWW. Our experiments demonstrate that the proposed system generates high-quality haikus and that using content-related input and multimedia-rich output is effective for increasing users' satisfaction. We have performed impression evaluation experiments and confirmed that our method is especially useful for generating haikus with higher depth and sound-sharpness, which are two very important categories in professional evaluation of Japanese poetry. Next, haikus generated using the proposed method were evaluated by blog authors and blog readers and again, the proposed method outperformed the baseline. We also measured how the presence or absence of multimedia output influenced the evaluation. While using both vocal output and an image achieved higher scores than text alone, there were cases in which some combinations of effects were evaluated higher than all the effects used together. With our original approach to generating poetry, we wish to show the importance of new media and possibilities that are arising from the utilization of the "wisdom of (web-)crowds" in order to achieve higher standards for AI-generated art.