Amazon.com

About: Amazon.com is a company organization based out in Seattle, Washington, United States. It is known for research contribution in the topics: Service (business) & Service provider. The organization has 13363 authors who have published 17317 publications receiving 266589 citations.

Refining search queries by the suggestion of correlated terms from prior searches

Abstract: A search engine is disclosed which suggests related terms to the user to allow the user to refine a search. The related terms are generated using query term correlation data which reflects the frequencies with which specific terms have previously appeared within the same query. The correlation data is generated and stored in a look-up table using an off-line process which parses a query log file. The table is regenerated periodically from the most recent query submissions (e.g., the last two weeks of query submissions), and thus strongly reflects the current preferences of users. Each related term is presented to the user via a respective hyperlink which can be selected by the user to submit a modified query. In one embodiment, the related terms are added to and selected from the table so as to guarantee that the modified queries will not produce a NULL query result.

Bilateral Multi-Perspective Matching for Natural Language Sentences

Abstract: Natural language sentence matching is a fundamental technology for a variety of tasks. Previous approaches either match sentences from a single direction or only apply single granular (word-by-word or sentence-by-sentence) matching. In this work, we propose a bilateral multi-perspective matching (BiMPM) model under the "matching-aggregation" framework. Given two sentences $P$ and $Q$, our model first encodes them with a BiLSTM encoder. Next, we match the two encoded sentences in two directions $P \rightarrow Q$ and $P \leftarrow Q$. In each matching direction, each time step of one sentence is matched against all time-steps of the other sentence from multiple perspectives. Then, another BiLSTM layer is utilized to aggregate the matching results into a fix-length matching vector. Finally, based on the matching vector, the decision is made through a fully connected layer. We evaluate our model on three tasks: paraphrase identification, natural language inference and answer sentence selection. Experimental results on standard benchmark datasets show that our model achieves the state-of-the-art performance on all tasks.

Temporal Segment Networks for Action Recognition in Videos

Abstract: We present a general and flexible video-level framework for learning action models in videos. This method, called temporal segment network (TSN), aims to model long-range temporal structure with a new segment-based sampling and aggregation scheme. This unique design enables the TSN framework to efficiently learn action models by using the whole video. The learned models could be easily deployed for action recognition in both trimmed and untrimmed videos with simple average pooling and multi-scale temporal window integration, respectively. We also study a series of good practices for the implementation of the TSN framework given limited training samples. Our approach obtains the state-the-of-art performance on five challenging action recognition benchmarks: HMDB51 (71.0 percent), UCF101 (94.9 percent), THUMOS14 (80.1 percent), ActivityNet v1.2 (89.6 percent), and Kinetics400 (75.7 percent). In addition, using the proposed RGB difference as a simple motion representation, our method can still achieve competitive accuracy on UCF101 (91.0 percent) while running at 340 FPS. Furthermore, based on the proposed TSN framework, we won the video classification track at the ActivityNet challenge 2016 among 24 teams.

User interface and methods for recommending items to users

Abstract: An improved user interface and method are provided for presenting recommendations to a user when the user adds an item to a shopping cart. In response to the shopping cart add event, a page generation process generates and returns a page that includes a recommendations portion and a condensed view of the shopping cart. The recommendations portion preferably includes multiple recommendation sections, each of which displays a different respective set of recommended items selected according to a different respective recommendation or selection algorithm (e.g., recommendations based on shopping cart contents, recommendations based on purchase history, etc.). The condensed shopping cart view preferably lacks controls for editing the shopping cart, and lacks certain types of product information, making more screen real estate available for the display of the recommendations content. A link to a full shopping cart page allows the user to edit the shopping cart and view expanded product descriptions.

FEVER: a large-scale dataset for Fact Extraction and VERification

Abstract: In this paper we introduce a new publicly available dataset for verification against textual sources, FEVER: Fact Extraction and VERification. It consists of 185,445 claims generated by altering sentences extracted from Wikipedia and subsequently verified without knowledge of the sentence they were derived from. The claims are classified as SUPPORTED, REFUTED or NOTENOUGHINFO by annotators achieving 0.6841 in Fleiss κ. For the first two classes, the annotators also recorded the sentence(s) forming the necessary evidence for their judgment. To characterize the challenge of the dataset presented, we develop a pipeline approach and compare it to suitably designed oracles. The best accuracy we achieve on labeling a claim accompanied by the correct evidence is 31.87%, while if we ignore the evidence we achieve 50.91%. Thus we believe that FEVER is a challenging testbed that will help stimulate progress on claim verification against textual sources.