harvardnlp / boxscore-data Goto Github PK

Hi,
While I read the data, I found that the first sentence describing the winner of this game, the winner & loser are mistaken. I am not sure if I understand the data correctly. Below is a example.

[Summary]: The Memphis Grizzlies ( 31 - 12 ) beat the Philadelphia 76ers ( 8 - 36 ) , 101 - 83 .
[Home_line]: "TEAM-CITY": "Philadelphia", "TEAM-PTS": "101","TEAM-NAME": "76ers"
[vis_line]: "TEAM-CITY": "Memphis", "TEAM-PTS": "83","TEAM-AST": "18", "TEAM-NAME": "Grizzlies"

Hello！Recently I have read "Challenges in Data-to-Document Generation" and it was really cool! However, I wonder if I can know how to caluculate the Relation Generation (RG). Maybe scripts or something? Because I do not know how you calculate precision of unique relations extracted from generated text that also appear in s, with different pair types (r.t)? Thanks!

Hi,
I am trying to execute this code but a dependency "nba_py" is not available.
Command executed : python scripts/run_pipeline.py

Error:
import nba_py
ImportError: No module named nba_py
can you please let us know what is missing?

Thank you

Hi, I can't find the nba_sbnation_links.txt file (required for scraping sbnation summaries).
-Thanks

Hi,

One of the summary entries in train.json in the rotowire dataset is Lorem ipsum, a typesetting placeholder: https://en.wikipedia.org/wiki/Lorem_ipsum

It is a Heat@Spurs game on "06_15_14".