Tatjana Scheffler

Matching points of interest from different social networking sites

Valuable user-generated information about locations (points of interest, POIs) is stored in various online social media platforms. Merging the data associated with one POI is hard because the platforms lack common identifiers. In addition, user-generated data is commonly faulty or contradictory. Here we present an approach matching POIs from Qype and Facebook Places to their counterparts in OpenStreetMap. The algorithm uses different similarity measures taking the geographic distance of POIs into account as well as the string similarity of selected metadata fields, showing good results.

Binding of Anaphors in LTAG

This paper presents an LTAG account for binding of reflexives and reciprocals in English. For these anaphors, a multi-component lexical entry is proposed, whose first component is a degenerate NP-tree that adjoins into the anaphors binder. This establishes the local structural relationship needed to ensure coreference and agreement. The analysis also allows a parallel treatment of reflexives and reciprocals, which is desirable because their behavior is very similar. In order to account for non-local binding phenomena, as in raising and ECM cases, we employ flexible composition, constrained by a subject intervention constraint between the two components of the anaphors lexical entry. Finally, the paper discusses further data such as extraction and picture-NP examples.

The Metagrammar Goes Multilingual: A Cross-Linguistic Look at the V2-Phenomenon

We present an initial investigation into the use of a metagrammar for explicitly sharing abstract grammatical specifications among languages. We define a single class hierarchy for a metagrammar which allows us to automatically generate grammars for different languages from a single compact metagrammar hierarchy. We use as our linguistic example the verb-second phenomenon, which shows considerable variation while retaining a basic property, namely the fact that the verb can appear in one of two positions in the clause.

Dialog Act Annotation for Twitter Conversations

We present a dialog act annotation for German Twitter conversations. In this paper, we describe our annotation effort of a corpus of German Twitter conversations using a full schema of 57 dialog acts, with a moderate inter-annotator agreement of multi- = 0.56 for three untrained annotators. This translates to an agreement of 0.76 for a minimal set of 10 broad dialog acts, comparable to previous work. Based on multiple annotations, we construct a merged gold standard, backing off to broader categories when needed. We draw conclusions wrt. the structure of Twitter conversations and the problems they pose for dialog act characterization.

User simulation for the evaluation of bus information systems

In this paper, we describe our contribution to the Spoken Dialog Challenge. We set up a user simulation using the large Lets Go corpus as resource to build our models. Automatic calls were made to all four dialog systems in the SDC, bus information systems that cover the schedule of Pittsburgh, PA. We discuss in detail the architecture and required setup for our system-independent user simulation and report the results and challenges we faced. We also report initial evaluation results.

SpeechEval: A Domain-Independent User Simulation Platform for Spoken Dialog System Evaluation

In this paper, we present the SpeechEval user simulation platform for spoken dialog system (SDS) evaluation. Our user simulation realizes a complete end-to-end user test system with spoken input and output that calls speech-based dialog systems via telephone, and is intended to be used in the development and evaluation of real-life SDSs. We give an overview of the system architecture and present the results of applying our approach to two different kinds of systems in two domains.

Semi-automatic creation of resources for spoken dialog systems

The increasing number of spoken dialog systems calls for efficient approaches for their development and testing. Our goal is the minimization of hand-crafted resources to maximize the portability of this evaluation environment across spoken dialog systems and domains. In this paper we discuss the user simulation technique which allows us to learn general user strategies from a new corpus. We present this corpus, the VOICE Awards human-machine dialog corpus, and show how it is used to semi-automatically extract the resources and knowledge bases necessary in spoken dialog systems, e.g., the ASR grammar, the dialog classifier, the templates for generation, etc.

OPT: Oslo—Potsdam—Teesside Pipelining Rules, Rankers, and Classifier Ensembles for Shallow Discourse Parsing

The OPT submission to the Shared Task of the 2016 Conference on Natural Language Learning (CoNLL) implements a ‘classic’ pipeline architecture, combining binary classification of (candidate) explicit connectives, heuristic rules for non-explicit discourse relations, ranking and ‘editing’ of syntactic constituents for argument identification, and an ensemble of classifiers to assign discourse senses. With an end-toend performance of 27.77 F1 on the English ‘blind’ test data, our system advances the previous state of the art (Wang & Lan, 2015) by close to four F1 points, with particularly good results for the argument identification sub-tasks. OPT system results appear more competitive on the new, ‘blind’ test data than on the ‘test’ and ‘development’ sections of the Penn Discourse Treebank (PDTB; Prasad et al., 2008), which may indicate reduced over-fitting to specific properties of the venerable Wall Street Journal (WSJ) text underlying the PDTB.

Human-machine corpus analysis for generation and interaction with spoken dialog systems

This paper describes a new approach to language generation for simulated users based on the construction of flexible templates extracted from a corpus. In our opinion a realistic user simulation on the speech level is based on two parts: user behavior and language generation. In this work we mainly concentrate on the language generation for simulated user interaction with spoken dialog systems (SDS). The presented approach could be used as part of a user simulation for intensive end-to-end system tests and evaluations and for testing purposes of the speech recognition and natural language understanding modules of an SDS.