Tasks that require modeling of both language and visual information, such as image captioning, have become very popular in recent years. Most state-of-the-art approaches make use of image representations obtained from a deep neural network, which are used to generate language information in a variety of ways with end-to-end neural-network-based models. However, it is not clear how different image representations contribute to language generation tasks. In this paper, we probe the representational contribution of the image features in an end-to-end neural modeling framework and study the properties of different types of image representations. We focus on two popular vision to language problems: The task of image captioning and the task of multimodal machine translation. Our analysis provides interesting insights into the representational properties and suggests that end-to-end approaches implicitly learn a visual-semantic subspace and exploit the subspace to generate captions.

We present an approach to combining distributional semantic representations induced from text corpora with manually constructed lexical semantic networks. While both kinds of semantic resources are available with high lexical coverage, our aligned resource combines the domain specificity and availability of contextual information from distributional models with the conciseness and high quality of manually crafted lexical networks. We start with a distributional representation of induced senses of vocabulary terms, which are accompanied with rich context information given by related lexical items. We then automatically disambiguate such representations to obtain a full-fledged proto-conceptualization, i.e. a typed graph of induced word senses. In a final step, this proto-conceptualization is aligned to a lexical ontology, resulting in a hybrid aligned resource. Moreover, unmapped induced senses are associated with a semantic type in order to connect them to the core resource. Manual evaluations against ground-truth judgments for different stages of our method as well as an extrinsic evaluation on a knowledge-based Word Sense Disambiguation benchmark all indicate the high quality of the new hybrid resource. Additionally, we show the benefits of enriching top-down lexical knowledge resources with bottom-up distributional information from text for addressing high-end knowledge acquisition tasks such as cleaning hypernym graphs and learning taxonomies from scratch.

This paper describes a question answering framework that can answer student questions given in natural language. We suggest a methodology that makes use of reliable resources only, provides the answer in the form of a multi-document summary for both factoid and open-ended questions, and produces an answer also from foreign resources by translating into the native language. The resources are compiled using a question database in the selected domains based on reliability and coverage metrics. A question is parsed using a dependency parser, important parts are extracted by rule-based and statistical methods, the question is converted into a representation, and a query is built. Documents relevant to the query are retrieved from the set of resources. The documents are summarized and the answers to the question together with other relevant information about the topic of the question are shown to the user. A summary answer from the foreign resources is also built by the translation of the input question and the retrieved documents. The proposed approach was applied to the Turkish language and it was tested with several experiments and a pilot study. The experiments have shown that the summaries returned include the answer for about 50–60 percent of the questions. The data bank built for factoid and open-ended questions in the two domains covered is made publicly available.

Summarizing content contributed by individuals can be challenging, because people make different lexical choices even when describing the same events. However, there remains a significant need to summarize such content. Examples include the student responses to post-class reflective questions, product reviews, and news articles published by different news agencies related to the same events. High lexical diversity of these documents hinders the system’s ability to effectively identify salient content and reduce summary redundancy. In this paper, we overcome this issue by introducing an integer linear programming-based summarization framework. It incorporates a low-rank approximation to the sentence-word cooccurrence matrix to intrinsically group semantically similar lexical items. We conduct extensive experiments on datasets of student responses, product reviews, and news documents. Our approach compares favorably to a number of extractive baselines as well as a neural abstractive summarization system. The paper finally sheds light on when and why the proposed framework is effective at summarizing content with high lexical variety.

The extraction of templates such as ‘regard X as Y’ from a set of related phrases requires the identification of their internal structures. This paper presents an unsupervised approach for extracting templates on-the-fly from only tagged text by using a novel relaxed variant of the Sequence Binary Decision Diagram (SeqBDD). A SeqBDD can compress a set of sequences into a graphical structure equivalent to a minimal deterministic finite state automata, but more compact and better suited to the task of template extraction. The main contribution of this paper is a relaxed form of the SeqBDD construction algorithm that enables it to form general representations from a small amount of data. The process of compression of shared structures in the text during Relaxed SeqBDD construction, naturally induces the templates we wish to extract. Experiments show that the method is capable of high-quality extraction on tasks based on verb+preposition templates from corpora and phrasal templates from short messages from social media.

This paper addresses the feasibility of cross-lingual parsing with Universal Dependencies (UD) between Romance languages, analyzing its performance when compared to the use of manually annotated resources of the target languages. Several experiments take into account factors such as the lexical distance between the source and target varieties, the impact of delexicalization, the combination of different source treebanks or the adaptation of resources to the target language, among others. The results of these evaluations show that the direct application of a parser from one Romance language to another reaches similar labeled attachment score (LAS) values to those obtained with a manual annotation of about 3,000 tokens in the target language, and unlabeled attachment score (UAS) results equivalent to the use of around 7,000 tokens, depending on the case. These numbers can noticeably increase by performing a focused selection of the source treebanks. Furthermore, the removal of the words in the training corpus (delexicalization) is not useful in most cases of cross-lingual parsing of Romance languages. The lessons learned with the performed experiments were used to build a new UD treebank for Galician, with 1,000 sentences manually corrected after an automatic cross-lingual annotation. Several evaluations in this new resource show that a cross-lingual parser built with the best combination and adaptation of the source treebanks performs better (77 percent LAS and 82 percent UAS) than using more than 16,000 (for LAS results) and more than 20,000 (UAS) manually labeled tokens of Galician.

Research on topic segmentation has recently focused on segmenting documents by taking advantage of documents covering the same topics. In order to properly evaluate such approaches, a dataset of related documents is needed. However, existing datasets are limited in the number of related documents per domain. In addition, most of the available datasets do not consider documents from different media sources (PowerPoints, videos, etc.), which pose specific challenges to segmentation. We fill this gap with the MUltimedia SEgmentation Dataset (MUSED), a collection of documents manually segmented, from different media sources, in seven different domains, with an average of twenty related documents per domain. In this paper, we describe the process of building MUSED. A multi-annotator study is carried out to determine if it is possible to observe agreement among human judges and characterize their disagreement patterns. In addition, we use MUSED to compare the state-of-the-art topic segmentation techniques, including the ones that take advantage of related documents. Moreover, we study the impact of having documents from different media sources in the dataset. To the best of our knowledge, MUSED is the first dataset that allows a straightforward evaluation of both single- and multiple-documents topic segmentation techniques, as well as to study how these behave in the presence of documents from different media sources. Results show that some techniques are, indeed, sensitive to different media sources, and also that current multi-document segmentation models do not outperform previous models, pointing to a research line that needs to be boosted.

In this paper, we study the problem of aspect-based sentiment analysis. Our model simultaneously extracts aspect-specific opinion expressions and determines the rating for each aspect in reviews. Previous works have mainly focused on the problem of opinion phrase extraction and aspect rating prediction in a pipelined manner and are not able to capture the dependencies of aspect opinion expression on aspect rating and vice-versa. They are also unable to discover aspect-specific opinion expressions and their associated rating scores. We present a joint modelling approach to extract aspect-specific sentiment expression and aspect rating prediction simultaneously. This paper proposes a novel LDA–CRF hybrid model which employs discriminative conditional random field component for phrase extraction, a regression component for rating prediction and a generative component for grouping aspect–sentiment expressions (aspect-specific opinion expressions) into coherent topics. To show the effectiveness of our approach, we evaluate the performance of the model on both task: (i) aspect-specific opinion expressions and (ii) rating prediction on the dataset of hotel and restaurant reviews from TripAdvisor.com. Experimental results show that both task potentially reinforce each other and joint modeling outperformed state-of-the-art baselines for each individual tasks.

A growing body of recent work focuses on the challenging problem of scene understanding using a variety of cross-modal methods which fuse techniques from image and text processing. In this paper, we develop representations for the semantics of scenes by explicitly encoding the objects detected in them and their spatial relations. We represent image content via two well-known types of tree representations, namely constituents and dependencies. Our representations are created deterministically, can be applied to any image dataset irrespective of the task at hand, and are amenable to standard NLP tools developed for tree-based structures. We show that we can apply syntax-based SMT and tree kernel methods in order to build models for image description generation and image-based retrieval. Experimental results on real-world images demonstrate the effectiveness of the framework.

It is amazing that I have written for more than a year on Emerging Trends without mentioning China’s investments in Artificial Intelligence (AI). Now that I have moved to Baidu, I would like to take this opportunity to share some of my personal observations with what’s happening in China over the past 25 years. The top universities in China have always been very good, but they are better today than they were 25 years ago, and they are on a trajectory to become the biggest and the best in the world. China is investing big time in what we do, both in the private sector and the public sector. Kai-Fu Lee is bullish on his investments in AI and China. There is a bold government plan for AI with specific milestones for parity with the West in 2020, major breakthroughs by 2025 and the envy of the world by 2030.

It seems like there’s yet another cloud-based text analytics Application Programming Interface (API) on the market every few weeks. If you’re interested in building an application using these kinds of services, how do you decide which API to go for? In the previous Industry Watch post, we looked at the text analytics APIs from the behemoths in the cloud software world: Amazon, Google, IBM and Microsoft. In this post, we survey sixteen APIs offered by smaller players in the market.

When a recurrent neural network (RNN) language model is used for caption generation, the image information can be fed to the neural network either by directly incorporating it in the RNN – conditioning the language model by ‘injecting’ image features – or in a layer following the RNN – conditioning the language model by ‘merging’ image features. While both options are attested in the literature, there is as yet no systematic comparison between the two. In this paper, we empirically show that it is not especially detrimental to performance whether one architecture is used or another. The merge architecture does have practical advantages, as conditioning by merging allows the RNN’s hidden state vector to shrink in size by up to four times. Our results suggest that the visual and linguistic modalities for caption generation need not be jointly encoded by the RNN as that yields large, memory-intensive models with few tangible advantages in performance; rather, the multimodal integration should be delayed to a subsequent stage.

A diacritic is a mark placed near or through a character to alter its original phonetic or orthographic value. Many languages around the world use diacritics in their orthography, whatever the writing system the orthography is based on. In many languages, diacritics are ignored either by convention or as a matter of convenience. For users who are not familiar with the text domain, the absence of diacritics within text has been known to cause mild to serious readability and comprehension problems. However, the absence of diacritics in text causes near-intractable problems for natural language processing systems. This situation has led to extensive research on diacritization. Several techniques have been applied to address diacritic restoration (or diacritization) but the existing surveys of techniques have been restricted to some languages and hence left gaps for practitioners to fill. Our survey examined diacritization from the angle of resources deployed and various formulation employed for diacritization. It was concluded by recommending that (a) any proposed technique for diacritization should consider the language features and the purpose served by diacritics, (b) that evaluation metrics needed to be more rigorously defined for easy comparison of performance of models.

If you’re in the market for an off-the-shelf text analytics API, you have a lot of options. You can choose to go with a major player in the software world, for whom each AI-related service is just another entry in their vast catalogues of tools, or you can go for a smaller provider that focusses on text analytics as their core business. In this first of two related posts, we look at what the most prominent software giants have to offer today.

Lots of companies are offering lots of APIs. Reviews are not always as constructive as they could be. Some reviews encourage unproductive work on checkbox features that no one wants. It makes no sense to do the wrong thing badly. Constructive reviews should help focus priorities on what matters. Users care more about a great box opening experience than small improvements in word error rate and BLEU, popular metrics for speech and translation. In 15 minutes or less, can we teach potential users something new (and fun) that does something useful, such as how to translate PowerPoint between English and Chinese, preserving many of the features that are important to PowerPoint such as graphics and animations? See Appendix for the solution.

Charles Wayne restarted funding in speech and language in the mid-1980s after a funding winter brought on by Pierce’s glamour-and-deceit criticisms in the ALPAC report and ‘Whither Speech Recognition’. Wayne introduced a new glamour-and-deceit-proof idea, an emphasis on evaluation. No other sort of program could have been funded at the time, at least in America. One could argue that Wayne has been so successful that the program no longer needs him to continue on. These days, shared tasks and leaderboards have become common place in speech and language (and vision and machine learning) research. That said, I am concerned that the community may not appreciate what it has got until it’s gone. Wayne has been doing much more than merely running competitions, but he did what he did in such a subtle Columbo-like way. Going forward, government funding is being eclipsed by consumer markets. Those of us with research to sell need to find more and more ways to be relevant to potential sponsors given this new world order.