Advanced Topics in Knowledge Bases

Block seminar, 7 ECTS credits, winter semester 2018–19

Basic Information

  • Type: Block seminar
  • Lecturer: Simon Razniewski
  • Credits: 7 ECTS credits
  • Registration: Sign up under this link until September 28
  • Dates: 3 introductory meetings on October 23, 25 and 30, two days block seminar in winter 2019 (probably February, to be agreed among the participants).

The seminar is a block seminar and will take place on two consecutive days in winter 2019. There will also be three meetings at the beginning of the semester, for which participation is mandatory.

News

  • July 4 -- first version of the course website online
  • July 24 -- registration deadline fixed to September 28

Topics

A detailed list is below

Representation, collection and extraction of general knowledge in knowledge bases (KBs) is at the core of many AI applications. In this seminar, we cover a range of topics around KBs, in particular factual KBs (e.g., Wikidata, YAGO, DBPedia), common-sense KBs (e.g., science knowledge, howto-knowledge, script knowledge) and non-textual KBs (e.g., ImageNet). We explore how these KBs are constructed and how they are used in various applications such as question answering (QA), story/script prediction and biography generation. We also explore learning new facts over KBs.

Registration

  • The number of participants is limited
  • To apply for registration, please fill the form under this link. You will receive an email confirmation
  • Places will be allocated based on background match (courses taken) and motivation

Schedule

  • September 28 -- Registration deadline
  • October 3 -- Notification of placement
  • October 23 -- Kick-off meeting and introduction to knowledge bases (participation is mandatory) [pdf]
    • Time: 10:00-11:30
    • Place: MPII building (E1 4), room 24 (ground floor)
      • Explanation of the structure and organization of the seminar
      • Introduction to knowledge bases
  • October 25 -- Introduction to knowledge bases (ctnd.) and topic presentation
    • Time: 10:00-11:30
    • Place: MPII building (E1 4), room 24 (ground floor)
      • Introduction to knowledge bases (ctnd.)
      • Presentation of the topics
  • October 30 -- "How to prepare and present a seminar talk" (participation is mandatory)
    • Time: 10:00-11:30
    • Place: MPII building (E1 4), room 24 (ground floor)
      • As this is a block seminar, it is particularly crucial that the students' presentations are of high quality. This lecture aims at preparing the participants in such a way that their slides and presentations will be of high quality.
      • Topic assignment: students prepare a preferred ranked list of proposed topics, we will do the live assignment based on a randomly decided order.
  • November 30 -- students send a suggestion of the outline of their seminar paper, including an itemization of the planned content for each section.
  • December 5/6 -- students have in-person feedback meetings
  • January 31, 2019 -- students submit their final seminar paper.
  • TBD -- students send preliminary slides
  • TBD -- students send their final slides which they will use in the block seminar
  • TBD -- Block seminar, Day 1
  • TBD -- Block seminar, Day 2

Rules and Grading

  • Participation in the introductory meetings, the "How to prepare and present a seminar talk" lecture, and both days of the block seminar is mandatory.
  • Students will be assigned a particular topic and have to submit a seminar paper (template will be provided) and give a presentation (25 minutes + 10 minutes for discussion) over the topic.
  • Grading will be based on:
    • the report
    • the presentation
    • knowledge on the subject (as evidenced in the discussion after the presentation)
    • activity in the discussions
    • ability to stick to deadlines
  • Attention: According to the study regulations, you are only allowed to withdraw from the seminar within three weeks after the kick-off meeting, i.e., until TBD. Later withdrawal counts as "failed".

Topics

  1. A WordNet-based image KB: ImageNet
    • ImageNet: A large-scale hierarchical image database (Deng et al., CVPR 2009) [pdf]
    • WordNet: A lexical database for English (Miller, CACM 1995) [pdf]
  2. Spatial common sense from text and images
    • Acquiring Common Sense Spatial Knowledge through Implicit Spatial Templates, Collell et al., AAAI 2018 [pdf]
    • Automatic Extraction of Commonsense LocatedNear Knowledge, Xu et al., ACL 2018 [pdf]
  3. Schema learning
    • Are All People Married?: Determining Obligatory Attributes in Knowledge Bases, Lajus and Suchanek, WWW 2018 [pdf]
    • Recoin: Relative Completeness in Wikidata, Balaraman et al., Wiki workshop at WWW 2018 [pdf]
  4. Change prediction and updating
    • "A Spousal Relation Begins with a Deletion of engage and Ends with an Addition of divorce": Learning State Changing Verbs from Wikipedia Revision History, Wijaya et al., EMNLP 2015 [pdf]
    • How to Keep a Knowledge Base Synchronized with Its Encyclopedia Source, Liang et al, IJCAI 2017 [pdf]
  5. Named entity recognition and coreference resolution
    • Joint Coreference Resolution and Named-Entity Linking with Multi-pass Sieves, Hajishirzi et al., EMNLP 2013 [pdf]
    • Improving Coreference Resolution by Learning Entity-Level Distributed Representations, Clark and Manning, ACL 2016 [pdf]
  6. Fact ranking and explanation
    • The unusual suspects: Deep learning based mining of interesting entity trivia from knowledge graphs (Fatma et al., AAAI 2017) [pdf]
    •  Tell Me Why Is It So? Explaining Knowledge Graph Relationships by Finding Descriptive Support Passages, Bathia et al., ISWC 2018 [pdf]
  7. Recall assessment
    • Demand-Weighted Completeness Prediction for a Knowledge Base, Hopkinson et al., NAACL 2018 [pdf]
    • What Knowledge is Needed to Solve the RTE5 Textual Entailment Challenge?, Clark, ArXiv 2018 [pdf]
    • Domain-targeted, high precision knowledge extraction, Dalvi et al., TACL 2017 (Section 5.1) [pdf]
  8. Open information extraction
    • Minie: minimizing facts in open information extraction, Gashteovski et al., EMNLP 2017. [pdf]
    • Open Language Learning for Information Extraction, Mausam et al., EMNLP 2012 [pdf]
  9. State Changes

    • What Happened? Leveraging VerbNet to Predict the Effects of Actions in Procedural Text, Clark et al., ArXiv 2018 [pdf]
    • Tracking State Changes in Procedural Text: a Challenge Dataset and Models for Process Paragraph Comprehension, Dalvi et al., NAACL 2018 [pdf]

More to come...