Συντάχθηκε 12-12-2012 16:50
από Vasilis Digalakis
Email συντάκτη: vdigalakis<στο>tuc.gr
Ενημερώθηκε:
-
Ιδιότητα: ΔΕΠ ΗΜΜΥ.
Παρασκευή 14 Δεκεμβρίου, ώρα 2:00-3:00, αμφιθέατρο κτηρίου Επιστημών
Ομιλητής: Αν. Καθ. Αλέξανδρος Ποταμιάνος, Τμήμα ΗΜΜΥ, Πολ. Κρήτης
Θέμα: Building Lexical Cognitive Networks from Web Corpora with Application to Semantic Similarity Computation and Affective Text Analysis
Περίληψη ομιλίας
We investigate language-agnostic algorithms for the construction of unsupervised distributional semantic models using web-harvested corpora. A corpus is created from web document snippets and the relevant semantic similarity statistics are encoded in a semantic network. We propose the notion of semantic neighborhoods that are defined using co-occurrence or context similarity features.Three neighborhood-based similarity metrics are proposed, motivated by the hypotheses of attributional and maximum sense similarity. The lexical networks and semantic distances are motivated by cognitive considerations (associative networks and lexical priming). The proposed metrics are evaluated against human similarity ratings achieving state-of-the-art results. The proposed semantic similarity metrics are applied to affective modeling of text. Continuous valence ratings are estimated for unseen words using the underlying assumption that semantic similarity implies affective similarity. Starting from a set of manually annotated words, a linear affective model is trained using the least mean squares algorithm followed by feature selection. We then propose linear and non-linear fusion schemes for investigating how lexical valence scores can be combined to produce sentence-level scores, as well as, extend the lexical similarity model to groups of words (compounds). Evaluation on affective text tasks (e.g., polarity recognition ) show significant performance improvement compared to the state of the art.