Home

Week: 09/01 ~ 09/08

• Measure precision, recall, f-measure
• Write grammar rules for the selected questions
• Select questions that will be answered
• Group questions
• Get Accuracy of two query [done]

Week: 08/04 ~ 08/11

• Disambiguation of property [done]
• Fix the punctuation problem [done]
• Improve QALD-4 result
  • map the chinese property into English ontology

Week: 07/28 ~ 08/03

• write grammar for QALD-4
• [Limitation of DBpedia](Limitation of DBpedia)
• Complete working version for Chinese [done]
• Extend the Dube parser for Chinese [done]
• Extend the DRS parser for Chinese [done]
• Handle the Chinese name [done]

Week: 07/21 ~ 07/27

• 07/27
• 07/26
  • Include Chinese SPARQL End point [done]
• 07/25
• 07/24
  • Include Chinese AGISTIS [Done]
• 07/23
• 07/22
  • Include Chinese POS [Done]
• 07/21

Week: 07/14 ~ 07/20

• Make the Grammar working for Chinese [Done]
• Make the Post tagger working for Chinese [Done]

Week: 07/01 ~ 07/08

• 07/10 * Due day for ISWC Demo (extended to 07/17) [Done]
• 07/08 * Evaluate * Deploy a chinese DBpedia endpoint (link) ([Chinese DBpedia Endpoint Overview](Chinese DBpedia Endpoint Overview))
• 07/07 * First draft for ISWC Demo [Done]
• 07/06 * Generate the query with the grammar * Prepare the Chinese DBpedia [done] * Send the query to the Chinese DBpedia to get Answer [done]
• 07/05 * Make AutoSparal working for English [Done]
• 07/01 * Generate query from template [done]

Week: 06/23 ~ 06/30

• 06/30 [Disambiguation Results](disambiguation results) * Accuracy of the disambiguation component with automatic NER [Done] * Accuracy of the disambiguation component with perfect input (NER) [Done] * Label the correct entity recognition [Done] * Fix the translation errors: train [done], test [Done] * Get the correct disambiguation result manually by converting the English links to Chinese [Done]
• 06/24 * Generate simple templates with current grammar [who, when]
• 06/23 * Fix the grammar [Done]

Week: 06/03 ~ 06/10

• 06/14 Get the evaluation results for the Chinese disambiguation component [Done]
• 06/12 Extracted correct answer from the Qald4 * Issue: the output of the disambiguation result is Chinese.
• 06/10 Get Chinese benchmark * TODO: write a script to get AGDISTIS result [Done] * Issue: cannot find the label (Obama) in the file instance_type [Fixed]
• 06/09 Get AGDISTIS to disambiguate Chinese in a local machine [Done] * Issue: very single character is treated as a word for chinese, making the candidate space too big [Fixed by using WhitespaceAnalyzer instead of StandardAnalyzer]
• 06/06 Get AGDISTIS to disambiguate English in a local machine [Done] TripleIndexTest.testMultipleTermsPerField() failed [Done]

Week: 05/26 ~ 06/02

1. Get AGDISTIS to disambiguate Chinese (we can evaluate the performance on the QALD queries)

2. Create the domain-independent dictionary for generating SPARQL solution modifiers

• 06/02

• 06/01

• 05/31 Find keywords according to questions

• 05/30 Start to build the domain-independent dictionary: know the format

  BasicTemplator

  ? Is LTAG grammar only applicable to English?

  ? What is the DudeParser for?

• 05/29

• 05/28, 05/27 Get the files for Chinese (Run DBpedia Extractor: MappingExtactor) [Need mapping.xml for Chinese] instance_types_zh.ttl" mappingbased_properties_zh.ttl specific_mappingbased_properties_zh.ttl

• 05/26 [Done] Fix the dependency issue for AGDISTIS and AutoSPARQL (package org.dllearner.common.index does not exist)