Ioannis Konstas

On social networks and collaborative recommendation

Social network systems, like last.fm, play a significant role in Web 2.0, containing large amounts of multimedia-enriched data that are enhanced both by explicit user-provided annotations and implicit aggregated feedback describing the personal preferences of each user. It is also a common tendency for these systems to encourage the creation of virtual networks among their users by allowing them to establish bonds of friendship and thus provide a novel and direct medium for the exchange of data. We investigate the role of these additional relationships in developing a track recommendation system. Taking into account both the social annotation and friendships inherent in the social graph established among users, items and tags, we created a collaborative recommendation system that effectively adapts to the personal information needs of each user. We adopt the generic framework of Random Walk with Restarts in order to provide with a more natural and efficient way to represent social networks. In this work we collected a representative enough portion of the music social network last.fm, capturing explicitly expressed bonds of friendship of the user as well as social tags. We performed a series of comparison experiments between the Random Walk with Restarts model and a user-based collaborative filtering method using the Pearson Correlation similarity. The results show that the graph model system benefits from the additional information embedded in social knowledge. In addition, the graph model outperforms the standard collaborative filtering method.

Categorising social tags to improve folksonomy-based recommendations

In social tagging systems, users have different purposes when they annotate items. Tags not only depict the content of the annotated items, for example by listing the objects that appear in a photo, or express contextual information about the items, for example by providing the location or the time in which a photo was taken, but also describe subjective qualities and opinions about the items, or can be related to organisational aspects, such as self-references and personal tasks. Current folksonomy-based search and recommendation models exploit the social tag space as a whole to retrieve those items relevant to a tag-based query or user profile, and do not take into consideration the purposes of tags. We hypothesise that a significant percentage of tags are noisy for content retrieval, and believe that the distinction of the personal intentions underlying the tags may be beneficial to improve the accuracy of search and recommendation processes. We present a mechanism to automatically filter and classify raw tags in a set of purpose-oriented categories. Our approach finds the underlying meanings (concepts) of the tags, mapping them to semantic entities belonging to external knowledge bases, namely WordNet and Wikipedia, through the exploitation of ontologies created within the W3C Linking Open Data initiative. The obtained concepts are then transformed into semantic classes that can be uniquely assigned to content- and context-based categories. The identification of subjective and organisational tags is based on natural language processing heuristics. We collected a representative dataset from Flickr social tagging system, and conducted an empirical study to categorise real tagging data, and evaluate whether the resultant tags categories really benefit a recommendation model using the Random Walk with Restarts method. The results show that content- and context-based tags are considered superior to subjective and organisational tags, achieving equivalent performance to using the whole tag space.

Summarizing Source Code using a Neural Attention Model

High quality source code is often paired with high level summaries of the computation it performs, for example in code documentation or in descriptions posted in online forums. Such summaries are extremely useful for applications such as code search but are expensive to manually author, hence only done for a small fraction of all code that is produced. In this paper, we present the first completely datadriven approach for generating high level summaries of source code. Our model, CODE-NN , uses Long Short Term Memory (LSTM) networks with attention to produce sentences that describe C# code snippets and SQL queries. CODE-NN is trained on a new corpus that is automatically collected from StackOverflow, which we release. Experiments demonstrate strong performance on two tasks: (1) code summarization, where we establish the first end-to-end learning results and outperform strong baselines, and (2) code retrieval, where our learned model improves the state of the art on a recently introduced C# benchmark by a large margin.

Learning a Neural Semantic Parser from User Feedback.

We present an approach to rapidly and easily build natural language interfaces to databases for new domains, whose performance improves over time based on user feedback, and requires minimal intervention. To achieve this, we adapt neural sequence models to map utterances directly to SQL with its full expressivity, bypassing any intermediate meaning representations. These models are immediately deployed online to solicit feedback from real users to flag incorrect queries. Finally, the popularity of SQL facilitates gathering annotations for incorrect predictions using the crowd, which is directly used to improve our models. This complete feedback loop, without intermediate representations or database specific engineering, opens up new ways of building high quality semantic parsers. Experiments suggest that this approach can be deployed quickly for any new target domain, as we show by learning a semantic parser for an online academic database from scratch.

Neural AMR: Sequence-to-Sequence Models for Parsing and Generation

Sequence-to-sequence models have shown strong performance across a broad range of applications. However, their application to parsing and generating text usingAbstract Meaning Representation (AMR)has been limited, due to the relatively limited amount of labeled data and the non-sequential nature of the AMR graphs. We present a novel training procedure that can lift this limitation using millions of unlabeled sentences and careful preprocessing of the AMR graphs. For AMR parsing, our model achieves competitive results of 62.1SMATCH, the current best score reported without significant use of external semantic resources. For AMR generation, our model establishes a new state-of-the-art performance of BLEU 33.8. We present extensive ablative and qualitative analysis including strong evidence that sequence-based AMR models are robust against ordering variations of graph-to-sequence conversions.

The Effect of Different Writing Tasks on Linguistic Style: A Case Study of the ROC Story Cloze Task

A writers style depends not just on personal traits but also on her intent and mental state. In this paper, we show how variants of the same writing task can lead to measurable differences in writing style. We present a case study based on the story cloze task (Mostafazadeh et al., 2016a), where annotators were assigned similar writing tasks with different constraints: (1) writing an entire story, (2) adding a story ending for a given story context, and (3) adding an incoherent ending to a story. We show that a simple linear classifier informed by stylistic features is able to successfully distinguish among the three cases, without even looking at the story context. In addition, combining our stylistic features with language model predictions reaches state of the art performance on the story cloze challenge. Our results demonstrate that different task framings can dramatically affect the way people write.

Story Cloze Task: UW NLP System

This paper describes University of Washington NLP’s submission for the Linking Models of Lexical, Sentential and Discourse-level Semantics (LSDSem 2017) shared task—the Story Cloze Task. Our system is a linear classifier with a variety of features, including both the scores of a neural language model and style features. We report 75.2% accuracy on the task. A further discussion of our results can be found in Schwartz et al. (2017).

Incremental Semantic Role Labeling with Tree Adjoining Grammar

We introduce the task of incremental semantic role labeling (iSRL), in which semantic roles are assigned to incomplete input (sentence prefixes). iSRL is the semantic equivalent of incremental parsing, and is useful for language modeling, sentence completion, machine translation, and psycholinguistic modeling. We propose an iSRL system that combines an incremental TAG parser with a semantically enriched lexicon, a role propagation algorithm, and a cascade of classifiers. Our approach achieves an SRL Fscore of 78.38% on the standard CoNLL 2009 dataset. It substantially outperforms a strong baseline that combines gold-standard syntactic dependencies with heuristic role assignment, as well as a baseline based on Nivre’s incremental dependency parser.

Semantic Role Labeling Improves Incremental Parsing

Incremental parsing is the task of assigning a syntactic structure to an input sentence as it unfolds word by word. Incremental parsing is more difficult than fullsentence parsing, as incomplete input increases ambiguity. Intuitively, an incremental parser that has access to semantic information should be able to reduce ambiguity by ruling out semantically implausible analyses, even for incomplete input. In this paper, we test this hypothesis by combining an incremental TAG parser with an incremental semantic role labeler in a discriminative framework. We show a substantial improvement in parsing performance compared to the baseline parser, both in full-sentence F-score and in incremental F-score.

Modeling facial expressions and peripheral physiological signals to predict topical relevance

By analyzing explicit & implicit feedback information retrieval systems can determine topical relevance and tailor search criteria to the users needs. In this paper we investigate whether it is possible to infer what is relevant by observing user affective behaviour. The sensory data employed range between facial expressions and peripheral physiological signals. We extract a set of features from the signals and analyze the data using classification methods, such as SVM and KNN. The results of our initial evaluation indicate that prediction of relevance is possible, to a certain extent, and implicit feedback models can benefit from taking into account user affective behavior.