Verb Semantics and Lexical Selection

By Zhibiao Wu, Martha Palmer. 1994

This paper will focus on the semantic representation of verbs in computer systems and its impact on lexical selection problems in machine translation (MT). Two groups of English and Chinese verbs are examined to show that lexical selection must be based on interpretation of the sentences as well as selection restrictions placed on the verb arguments. A novel representation scheme is suggested, and is compared to representations with selection restrictions used in transfer-based MT. We see our approach as closely aligned with knowledge-based MT approaches (KBMT), and as a separate component that could be incorporated into existing systems. Examples and experimental results will show that, using this scheme, inexact matches can achieve correct lexical selection.

Subfields: Computational Linguistics & Semantics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Linguistics & Semantics

Citations in the corpus

NExT-QA: Next Phase of Question-Answering to Explaining Temporal Actions (2021, methodology): “WUPS computes the Wu-Palmer similarity [48] of the words based on their depths in WordNet [33].”
Stacked Attention Networks for Image Question Answering (2015, methodology): “n problem since most of answers are single words. We evaluate the model using classiﬁcation accuracy as reported in [1,21,19]. The reference models also report the Wu-Palmer similarity (WUPS) measure [29]. The WUPS measure calculates the similarity between two words based on their longest common sub”
Hierarchical Question-Image Co-Attention for Visual Question Answering (2016, methodology): “kes 70%, 7%, 17%, and 6% of the whole dataset, respectively. All answers in this data set are single word. We report the results both on classiﬁcation accuracy and Wu-Palmer similarity (WUPS) measure [20] in Table 2. 4.2 Setup We use torch [4] to develop our model. We use the Rmsprop optimizer [19] ”
A Multi-World Approach to Question Answering about Real-World Scenes based on Uncertain Input (2014, methodology): “ 3, and a set membership measure , with properties (x2X) = 1 if x2X, (x2X) = max y2X (x= y) and (x= y) 2[0;1], in Eq. 4 with equality whenever = 1. For we use a variant of Wu-Palmer similarity [23, 24]. WUP(a;b) calculates similarity based on the depth of two words aand bin the taxonomy[25, 26], and”
Exploring Models and Data for Image Question Answering (2015, methodology): “To evaluate model performance, we used the plain answer accuracy as well as the Wu-Palmer similarity (WUPS) measure (Wu & Palmer, 1994; Malinowski & Fritz, 2014b).”
Are You Talking to a Machine? Dataset and Methods for Multilingual Image Question (2015, methodology): “Under this setting, it is plausible to use automatic evaluation metrics that measure the single word similarity, such as Wu-Palmer similarity measure (WUPS) [41].”
SemEval-2014 Task 10: Multilingual Semantic Textual Similarity (2014, methodology): “The clas-siﬁer uses a combination of features, such as those derived from traditional knowledge-based ((Lea-cock and Chodorow, 1998; Wu and Palmer, 1994; Lin, 1998), and others) and corpus-based metrics (LSA (Landauer et al., 1997)), paired with lexical features (such as Dice-Similarity,…”
SimLex-999: Evaluating Semantic Models With (Genuine) Similarity Estimation (2014, methodology): “WupSim has been shown to correlate well with human judgements on the similarity-focused RG dataset (Wu and Palmer, 1994).”
Ask Me Anything: Free-Form Visual Question Answering Based on Knowledge from External Sources (2015, methodology): “The WUPS calculates the similarity between two words based on the similarity between their common subsequence in the taxonomy tree.”
Image Question Answering Using Convolutional Neural Network with Dynamic Parameter Prediction (2015, methodology): “It uses thresholded Wu-Palmer similarity [28] based on WordNet [9] taxonomy to compute the similarity between words.”
ABC-CNN: An Attention Based Convolutional Neural Network for Visual Question Answering (2015, methodology): “WUP similarity measures the similarity of two words based on the depth of their lowest common ancestor in the taxonomy tree [29].”
Learning Knowledge Graphs for Question Answering through Conversational Dialog (2015, methodology): “To build WORDNET, we draw an edge between every pair of Wordnet concepts (ws, wq) for which the WuPalmer Similarity (WUP) (Wu and Palmer, 1994) of the first sense in each concept’s synset exceeds 0.”
FVQA: Fact-Based Visual Question Answering (2016, methodology): “Additionally, we also report the results based on WuPalmer Similarity (WUPS) [74].”
What Value Do Explicit High Level Concepts Have in Vision to Language Problems? (2015, methodology): “Besides the accuracy value (the proportion of correct answered testing questions to the total testing questions), the Wu-Palmer similarity (WUPS) [52] is also used to measure the performance of different models.”
Label Refinery: Improving ImageNet Classification through Label Progression (2018, methodology): “omic relationship between the categories). Since ImageNet categories are drawn from WordNet, we can use taxonomy-based distances to rene the labels. We experiment with using the Wu-Palmer similarity [34] of the WordNet [35] categories to rene the category labels. Table 5 compares rening labels at th”
Ask Your Neurons: A Deep Learning Approach to Visual Question Answering (2016, methodology): “To embrace the aforementioned ambiguities, Malinowski and Fritz (2014a) suggest using a thresholded taxonomy-based Wu-Palmer similarity (Wu and Palmer, 1994) for μ.”